From 3c04f76d409d649e43e984469ecc7a3d631b065d Mon Sep 17 00:00:00 2001 From: vwmaus Date: Fri, 17 Mar 2017 14:07:13 +0100 Subject: [PATCH] Update vignette to last JSS submission version --- inst/doc/applying_twdtw.R | 149 +++++++------- inst/doc/applying_twdtw.Rmd | 368 ++++++++++++++++------------------- inst/doc/applying_twdtw.pdf | Bin 680563 -> 695622 bytes man/twdtwApply.Rd | 2 +- vignettes/applying_twdtw.Rmd | 368 ++++++++++++++++------------------- vignettes/references.bib | 73 +++++++ 6 files changed, 487 insertions(+), 473 deletions(-) diff --git a/inst/doc/applying_twdtw.R b/inst/doc/applying_twdtw.R index 75c64c8..62b3fe7 100644 --- a/inst/doc/applying_twdtw.R +++ b/inst/doc/applying_twdtw.R @@ -5,6 +5,7 @@ opts_chunk$set( message = FALSE, error = FALSE, results = "hide", + cache.path = "./cache/", cache = FALSE, comment = "" ) @@ -27,15 +28,16 @@ if (other_user) { ## ---- echo=FALSE, eval = TRUE, cache = FALSE----------------------------- # Install dtwSat package -#install.packages("dtwSat_0.2.0.9000.tar.gz") +#install.packages("dtwSat_0.2.2.9000.tar.gz", repos = NULL) ## ---- echo=FALSE, eval = TRUE, cache = FALSE----------------------------- library(dtwSat) library(ggplot2) library(scales) -library(reshape2) +library(Hmisc) new_theme = theme_get() +new_theme$text$family = "Helvetica" new_theme$text$size = 8 old_theme = theme_set(new_theme) @@ -55,19 +57,19 @@ df_dist$y = 1.8 plotMatches(mat, attr="evi", k=n) + ylab("Time series Pattern") + geom_text(data=df_dist, mapping = aes_string(x='to', y='y', label='label'), - size = 2) + + size = 2, family="Helvetica") + theme(legend.position="none") -## ---- echo = TRUE, eval = TRUE, results = 'markup'----------------------- -ts = twdtwTimeSeries(MOD13Q1.ts, labels="Time series") -patterns_ts = twdtwTimeSeries(MOD13Q1.patterns.list) -MOD13Q1.ts.labels - ## ---- echo = TRUE, eval = TRUE, results = 'markup'----------------------- library(dtwSat) names(MOD13Q1.patterns.list) head(MOD13Q1.ts, n = 2) +## ---- echo = TRUE, eval = TRUE, results = 'markup'----------------------- +ts = twdtwTimeSeries(MOD13Q1.ts, labels="Time series") +patterns_ts = twdtwTimeSeries(MOD13Q1.patterns.list) +patterns_ts + ## ----example-timeseries, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/3, fig.align='center', fig.cap='Example of time series based on MODIS product MOD13Q1 \\citep{Friedl:2010}. The labels of the phenological cycle are shown in the plot.', fig.pos='!h'---- plot(ts, type = "timeseries") + annotate(geom = "text", x = MOD13Q1.ts.labels$from+90, y = 0.98, @@ -103,9 +105,9 @@ df_weight = melt(df_weight, id.vars = "Difference") names(df_weight)[-1] = c("Functions","Weight") ggplot(df_weight, aes_string(x="Difference", y="Weight", group="Functions", linetype="Functions")) + geom_line() + xlab("Time difference (days)") + - theme(text = element_text(size = 10), - plot.title = element_text(size = 10, face="bold"), - axis.title = element_text(size = 10), + theme(text = element_text(size = 10, family="Helvetica"), + plot.title = element_text(size = 10, family="Helvetica", face="bold"), + axis.title = element_text(size = 10, family="Helvetica"), legend.position = c(.3,.85), legend.background = element_rect(fill="transparent")) + scale_linetype(guide_legend(title = "")) @@ -115,7 +117,7 @@ plot(matches, type="matches", patterns.labels="Soybean", k=4) ## ----alignments-all-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.5, fig.align='center', fig.cap=c('Alignments and dissimilarity measures of the patterns "soybean", "cotton", and "maize" to the subintervals of the long-term time series using a logistic time-weight. The solid black line is the EVI time series, and the coloured lines are the alignments of the patterns that have dissimilarity measure lower than three.'), fig.pos='!h'---- plot(matches, type="alignments", attr = "evi", threshold = 3.0) -## ----time-series-classification, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.5, fig.align='center', fig.cap=c('Classification of each 6 months periods of the time series using results of the TWDTW analysis with logistic time-weight. The solid lines are the attributes of the time series, the background colours indicate the classification of the periods.'), fig.pos='!h'---- +## ----time-series-classification, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.8, fig.align='center', fig.cap=c('Classification of each 6 months periods of the time series using results of the TWDTW analysis with logistic time-weight. The solid lines are the attributes of the time series, the background colours indicate the classification of the periods.'), fig.pos='!ht'---- ts_classification = twdtwClassify(x = matches, from = as.Date("2009-09-01"), to = as.Date("2013-09-01"), by = "6 month", overlap = 0.5) @@ -135,28 +137,50 @@ ndvi = brick(paste(data_folder,"ndvi.tif", sep = "/")) day_of_year = brick(paste(data_folder,"doy.tif", sep = "/")) dates = scan(paste(data_folder,"timeline", sep = "/"), what = "dates") +## ---- echo = TRUE, eval = TRUE------------------------------------------- +raster_timeseries = twdtwRaster(blue, red, nir, mir, evi, ndvi, + timeline = dates, doy = day_of_year) + ## ---- echo = TRUE, eval = TRUE, results = 'markup'----------------------- field_samples = read.csv(paste(data_folder,"samples.csv", sep = "/")) -head(field_samples, 2) +head(field_samples, 5) table(field_samples[["label"]]) proj_str = scan(paste(data_folder,"samples_projection", sep = "/"), what = "character") proj_str -## ---- echo = TRUE, eval = TRUE------------------------------------------- -raster_timeseries = twdtwRaster(blue, red, nir, mir, evi, ndvi, - timeline = dates, doy = day_of_year) - ## ---- echo = TRUE, eval = TRUE, results = 'markup'----------------------- field_samples_ts = getTimeSeries(raster_timeseries, y = field_samples, proj4string = proj_str) field_samples_ts +## ---- echo = TRUE, eval = FALSE, results = 'markup'---------------------- +# set.seed(1) +# log_fun = logisticWeight(alpha=-0.1, beta=50) +# cross_validation = twdtwCrossValidate(field_samples_ts, times=100, p=0.1, +# freq = 8, formula = y ~ s(x, bs="cc"), weight.fun = log_fun) + +## ---- echo = FALSE, eval = TRUE------------------------------------------ +load(system.file("lucc_MT/cross_validation.RData", package = "dtwSat")) + +## ----plot-accuracy, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_width/2, fig.align='center', fig.cap='User\'s and producer\'s accuracy of the TWDTW cross-validation using 100 resampling-with-replacement. The plot shows the 95\\% confidence interval of the mean.', fig.pos='!ht'---- +plot(cross_validation, conf.int=.95) + +## ---- echo = FALSE, eval = TRUE, results = 'asis'------------------------ +twdtwXtable(cross_validation, conf.int=.95, digits = 2, caption="\\label{tab:cross-validation} User\'s and producer\'s accuracy of the TWDTW cross-validation using 100 resampling-with-replacement. The table shows the standard deviation ($\\sigma$) and the 95\\% confidence interval (ci) of the mean ($\\mu$).'", comment = FALSE, caption.placement = "bottom", table.placement="!ht") + +## ---- echo = TRUE, eval = TRUE------------------------------------------- +library(caret) +set.seed(1) +I = unlist(createDataPartition(field_samples[,"label"], p = 0.1)) +training_ts = subset(field_samples_ts, I) +validation_samples = field_samples[-I,] + ## ---- echo = TRUE, eval = TRUE------------------------------------------- temporal_patterns = - createPatterns(field_samples_ts, freq = 8, formula = y ~ s(x)) + createPatterns(training_ts, freq = 8, formula = y ~ s(x)) -## ----temporal-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_width/1.5, fig.align='center', fig.pos='!h', fig.cap='Temporal patterns of forest, cotton-fallow, soybean-cotton, soybean-maize, and soybean-millet based on the ground truth samples.'---- +## ----temporal-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_width/1.5, fig.align='center', fig.pos='!h', fig.cap='Temporal patterns of Forest, Cotton-fallow, Soybean-cotton, Soybean-maize, and Soybean-millet based on the ground truth samples.'---- plot(temporal_patterns, type = "patterns") + theme(legend.position = c(.8,.25)) @@ -165,79 +189,40 @@ log_fun = logisticWeight(alpha=-0.1, beta=50) twdtw_dist = twdtwApply(x = raster_timeseries, y = temporal_patterns, overlap = 0.5, weight.fun = log_fun, overwrite=TRUE, format="GTiff") -## ----plot-dissmilarity2013, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Illustration of the TWDTW dissimilarity from each temporal pattern in 2013. The blue areas are more similar to the pattern and the red areas are less similar to the pattern.', fig.pos='!h'---- -plot(x = twdtw_dist, type="distance", time.levels = 6) +## ----plot-dissmilarity2008, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Illustration of the TWDTW dissimilarity from each temporal pattern in 2008. The blue areas are more similar to the pattern and the red areas are less similar to the pattern.', fig.pos='!ht'---- +plot(x = twdtw_dist, type="distance") -## ---- echo = TRUE, eval = TRUE------------------------------------------- -land_use_maps = twdtwClassify(twdtw_dist, format="GTiff", overwrite=TRUE) +## ---- echo = TRUE, eval = TRUE, results = 'markup'----------------------- +land_cover_maps = twdtwClassify(twdtw_dist, format="GTiff", overwrite=TRUE) -## ----plot-map, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Land use maps for each year from 2008 to 2013.', fig.pos='!h'---- -plot(x = land_use_maps, type = "maps") +## ----plot-map, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Land cover maps for each year from 2008 to 2013.', fig.pos='!ht'---- +plot(x = land_cover_maps, type = "maps") -## ----plot-area, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Percentage of area for each land use class from 2008 to 2013.', fig.pos='!h'---- -plot(x = land_use_maps, type = "area") +## ----plot-area, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Percentage of area for each land cover class from 2008 to 2013.', fig.pos='!ht'---- +plot(x = land_cover_maps, type = "area") ## ----plot-change, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Gains and losses in area from the other classes. The $y$ axis shows the actual class; the positive direction of $x$ axis shows the gains and the negative direction of $x$ axis shows the losses of the classes indicated in $y$. The colors indicate from/to which classes the gains/losses belong.', fig.pos='!h'---- -plot(x = land_use_maps, type = "changes") +plot(x = land_cover_maps, type = "changes") -## ----plot-dissmilarity, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='TWDTW dissimilarity measure for each pixel over each classified period. The blue areas have high confidence and the red areas have low confidence in the classification.', fig.pos='!h'---- -plot(x = land_use_maps, type="distance") +## ----plot-dissmilarity, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='TWDTW dissimilarity measure for each pixel over each classified period. The blue areas have high confidence and the red areas have low confidence in the classification.', fig.pos='!ht'---- +plot(x = land_cover_maps, type="distance") -## ---- echo = TRUE, eval = FALSE------------------------------------------ -# set.seed(1) -# partitions = splitDataset(field_samples_ts, p=0.1, times=100, -# freq = 8, formula = y ~ s(x, bs="cc")) +## ---- echo = TRUE, eval = TRUE------------------------------------------- +maps_assessment = twdtwAssess(land_cover_maps, y = validation_samples, + proj4string = proj_str, conf.int=.95) -## ---- echo = TRUE, eval = FALSE, results = 'markup'---------------------- -# log_fun = logisticWeight(alpha=-0.1, beta=50) -# twdtw_res = lapply(partitions, function(x){ -# res = twdtwApply(x = x$ts, y = x$patterns, weight.fun = log_fun, n=1) -# twdtwClassify(x = res) -# }) -# assessment = twdtwAssess(twdtw_res) -# head(assessment, 5) +## ---- echo = FALSE, eval = TRUE, results = 'asis'------------------------ +twdtwXtable(maps_assessment, table.type="errormatrix", digits = 0, rotate.col = TRUE, caption="\\label{tab:map-error-matrix}Error matrix of the map classification based on TWDTW analysis. The area is in the map unit, in this case $m^2$. $w$ is the proportion of area mapped for each class.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") -## ---- echo = FALSE, eval = TRUE------------------------------------------ -load(system.file("lucc_MT/cross_validation.RData", package = "dtwSat")) +## ----plot-map-incorrect-samples, echo = FALSE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Incorrect classified samples.', fig.pos='!ht'---- +plot(x = maps_assessment, type="map", samples="incorrect") -## ----plot-accuracy, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_width/2, fig.align='center', fig.cap='User\'s Accuracy (UA) and Producer\'s Accuracy (PA) of the TWDTW method for land cover classification. The plot shows the averages and their confidence interval for 99\\%.', fig.pos='!h'---- -df = melt(assessment[,-1], id="label") -df$variable = factor(df$variable, levels = c("UA", "PA"), labels = c("User's Accuracy", "Producer's Accuracy")) -ggplot(df, aes(x=label, y=value)) + - stat_summary(fun.data="mean_cl_boot", fun.args=list(conf.int = .99), - width=0.5, geom="crossbar", size=0.1, fill = "gray") + - geom_point(size=0.2) + facet_grid(. ~ variable) + - scale_y_continuous(limits = c(0,1), labels = percent, breaks = seq(0,1,.2)) + - xlab("") + ylab("Accuracy") + coord_flip() +## ---- echo = FALSE, eval = TRUE, results = 'asis'------------------------ +twdtwXtable(maps_assessment, table.type="accuracy", show.prop = TRUE, digits = 2, rotate.col = TRUE, caption="\\label{tab:map-accuracy}Accuracy and error matrix in proportion of area of the classified map.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") ## ---- echo = FALSE, eval = TRUE, results = 'asis'------------------------ -assess_mean = aggregate(assessment[, c("UA","PA")], list(assessment$label), mean) -assess_sd = aggregate(assessment[, c("UA","PA")], list(assessment$label), sd) -l_names = levels(assessment$label) -names(l_names) = l_names -ic_ua = t(sapply(l_names, function(i) 100*mean_cl_boot(x = assessment$UA[assessment$label==i], conf.int = .99)))[,-1] -ic_pa = t(sapply(l_names, function(i) 100*mean_cl_boot(x = assessment$PA[assessment$label==i], conf.int = .99)))[,-1] - -assess_table = data.frame( - Class = assess_mean$Group.1, - - MUA = sprintf("%.2f", round(100*assess_mean$UA,2)), - SDUA = sprintf("(%.2f)", round(100*assess_sd$UA,2)), - CIUA = sprintf("[%.2f-%.2f]", round(as.numeric(ic_ua[,1]),2), round(as.numeric(ic_ua[,2]),2)), - - MPA = sprintf("%.2f", round(100*assess_mean$PA,2)), - SDPA = sprintf("(%.2f)", round(100*assess_sd$PA,2)), - CIPA = sprintf("[%.2f-%.2f]", round(as.numeric(ic_pa[,1]),2), round(as.numeric(ic_pa[,2]),2)) - ) - -x_assess = xtable::xtable(assess_table, - format = tab_format, digits = 2, label = "tab:assessment", alig=c("l","c","c","c","c","c","c","c"), - caption="User\'s and Producer\'s Accuracy of the land use classification based on TWDTW analysis. $\\mu$ is the average accuracy, $\\sigma$ the standard deviation, and CI is the confidence interval of 99\\% using 100 resampling-with-replacement.") - -addtorow = list() -addtorow$pos = list(0) -addtorow$command = paste("Class & \\multicolumn{3}{c}{User's Accuracy (UA) \\%} & \\multicolumn{3}{c}{Producer's Accuracy (PA)\\%}\\\\", paste(c("","$\\mu$","$\\sigma$","CI","$\\mu$","$\\sigma$","CI"), collapse="&"), "\\\\", collapse = "") - -xtable::print.xtable(x_assess, add.to.row=addtorow, include.colnames = FALSE, include.rownames = FALSE, - comment = FALSE, caption.placement = "bottom") +twdtwXtable(maps_assessment, table.type="area", digits = 0, rotate.col = TRUE, caption="\\label{tab:map-adjusted-area}Mapped and adjusted, accumulated over the whole period, i.e. the sum from the sum of the maps from 2008 to 2013. The area is in the map unit, in this case $m^2$.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") + +## ----plot-area-and-uncertainty, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.7, fig.align='center', fig.cap='Mapped and adjusted, accumulated over the whole period, i.e. the sum from the sum of the maps from 2008 to 2013. The area is in the map unit, in this case $m^2$.', fig.pos='!ht'---- +plot(x = maps_assessment, type="area", perc=FALSE) diff --git a/inst/doc/applying_twdtw.Rmd b/inst/doc/applying_twdtw.Rmd index 4fbe645..7c8ae09 100644 --- a/inst/doc/applying_twdtw.Rmd +++ b/inst/doc/applying_twdtw.Rmd @@ -1,17 +1,17 @@ --- author: - name: Victor Maus - affiliation: INPE - affiliation2: University of Münster + affiliation: University of Münster + affiliation2: INPE, IIASA address: > - Image Processing Division\newline - National Institute for Space Research\newline - Av. dos Astronautas, 1758\newline - 12227-010 São José dos Campos, Brazil + Ecosystems Services and Management Program\newline + International Institute for Applied Systems Analysis\newline + Schlossplatz 1\newline + A-2361 Laxenburg, Austria email: vwmaus1@gmail.com - url: \url{www.dpi.inpe.br/maus} - Telephone: +55/12/3208-7330 - - name: "Gilberto Câmara" + url: \url{www.iiasa.ac.at/staff/maus} + Telephone: +43/2236-807-550 + - name: Gilberto Câmara affiliation: INPE affiliation2: University of Münster - name: Marius Appel @@ -19,27 +19,25 @@ author: - name: Edzer Pebesma affiliation: University of Münster title: - formatted: "\\pkg{dtwSat}: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in \\proglang{R}$^1$" + formatted: "\\pkg{dtwSat}: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in \\proglang{R}" # If you use tex in the formatted title, also supply version without plain: "dtwSat: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in R" # For running headers, if needed short: "\\pkg{dtwSat}: Time-Weighted Dynamic Time Warping" abstract: > - The opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. However, few of the proposed methods for remote sensing time series analysis are available as open source software. In this paper we present the \proglang{R} package \pkg{dtwSat}. This package provides an implementation of the Time-Weighted Dynamic Time Warping method for land use and land cover mapping using sequence of multi-band satellite images. Methods based on dynamic time warping are flexible to handle irregular sampling and out-of-phase time series, and they have achieved significant results in time series analysis. \pkg{dtwSat} is available from the Comprehensive R Archive Network and contributes to making methods for satellite time series analysis available to a larger audience. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and evaluating the results. + The opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. However, few of the proposed methods for remote sensing time series analysis are available as open source software. In this paper we present the \proglang{R} package \pkg{dtwSat}. This package provides an implementation of the Time-Weighted Dynamic Time Warping method for land cover mapping using sequence of multi-band satellite images. Methods based on dynamic time warping are flexible to handle irregular sampling and out-of-phase time series, and they have achieved significant results in time series analysis. \pkg{dtwSat} is available from the Comprehensive R Archive Network and contributes to making methods for satellite time series analysis available to a larger audience. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and assessing the results. keywords: # at least one keyword must be supplied - formatted: [dynamic programming, MODIS time series, land use changes, crop monitoring] - plain: [dynamic programming, MODIS time series, land use changes, crop monitoring] + formatted: [dynamic programming, MODIS time series, land cover changes, crop monitoring] + plain: [dynamic programming, MODIS time series, land cover changes, crop monitoring] preamble: > \usepackage{amsmath} \usepackage{array} \usepackage{caption} \usepackage{subcaption} \usepackage{float} - \usepackage{framed} - \usepackage{listings} - \usepackage{siunitx} - \usepackage{latexsym} + \usepackage{microtype} + \setlength{\tabcolsep}{4pt} documentclass: nojss classoption: shortnames output: @@ -51,7 +49,7 @@ vignette: > %\VignetteIndexEntry{dtwSat: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in R} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8} - %\VignetteDepends{rticles, dtwSat, ggplot2, scales, Hmisc, reshape2, xtable, tikzDevice} + %\VignetteDepends{rticles, dtwSat, ggplot2, scales, Hmisc, reshape2, tikzDevice} --- @@ -62,6 +60,7 @@ opts_chunk$set( message = FALSE, error = FALSE, results = "hide", + cache.path = "./cache/", cache = FALSE, comment = "" ) @@ -83,11 +82,9 @@ if (other_user) { #options(tikzDocumentDeclaration = "\\documentclass{jss}\\usepackage{siunitx}\\usepackage{latexsym}") ``` - - ```{r , echo=FALSE, eval = TRUE, cache = FALSE} # Install dtwSat package -#install.packages("dtwSat_0.2.0.9000.tar.gz") +#install.packages("dtwSat_0.2.2.9000.tar.gz", repos = NULL) ``` @@ -95,9 +92,10 @@ if (other_user) { library(dtwSat) library(ggplot2) library(scales) -library(reshape2) +library(Hmisc) new_theme = theme_get() +new_theme$text$family = "Helvetica" new_theme$text$size = 8 old_theme = theme_set(new_theme) @@ -106,30 +104,28 @@ page_width = 5.590551#in 14.2#cm page_height = 9.173228#in 23.3#cm ``` -\footnotetext[1]{This vignette is based on the paper: MAUS, V.; CAMARA, G.; APPEL, M.; PEBESMA, E. dtwSat: Time-Weighted Dynamic Time Warping for satellite image time series analysis in R. Submitted to the Journal of Statistical Software.} - \sloppy # Introduction - -Remote sensing images are the most widely used data source for measuring land use and land cover change (LUCC). In many areas, remote sensing images are the only data available for this purpose [@Lambin:2006; @Fritz:2013]. Recently, the opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. These methods include multi-image compositing [@Griffiths:2013], detecting forest disturbance and recovery [@Kennedy:2010; @Zhu:2012; @DeVries:2015], crop classification [@Xiao:2005; @Wardlow:2007; @Petitjean:2012; @Maus:2016], planted forest mapping [@Maire:2014], crop expansion and intensification [@Galford:2008; @Sakamoto:2009], detecting trend and seasonal changes [@Lunetta:2006; @Verbesselt:2010; @Verbesselt:2010a; @Verbesselt:2012], and extracting seasonality metrics from satellite time series [@Jonsson:2002; @Jonsson:2004]. Given the open availability of large image data sets, the research community on Earth Observation would get much benefit from methods that are openly available, reproducible and comparable. However, few of the proposed methods for remote sensing time series analysis are available as open source software, the main exception being the BFAST and BFAST-monitor algorithms for change detection [@Verbesselt:2010; @Verbesselt:2010a]. This paper is a contribution to making methods for satellite time series analysis available to a larger audience. +Remote sensing images are the most widely used data source for measuring land use and land cover change (LUCC). In many areas, remote sensing images are the only data available for this purpose [@Lambin:2006; @Fritz:2013]. Recently, the opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. These methods include multi-image compositing [@Griffiths:2013], detecting forest disturbance and recovery [@Kennedy:2010; @Zhu:2012; @DeVries:2015], crop classification [@Xiao:2005; @Wardlow:2007; @Petitjean:2012; @Maus:2016], planted forest mapping [@Maire:2014], crop expansion and intensification [@Galford:2008; @Sakamoto:2009], detecting trend and seasonal changes [@Lunetta:2006; @Verbesselt:2010; @Verbesselt:2010a; @Verbesselt:2012], and extracting seasonality metrics from satellite time series [@Jonsson:2002; @Jonsson:2004]. Given the open availability of large image data sets, the Earth Observation community would get much benefit from methods that are openly available, reproducible and comparable. However, few of the proposed methods for remote sensing time series analysis are available as open source software, the main exception being the BFAST and BFAST-monitor algorithms for change detection [@Verbesselt:2010; @Verbesselt:2010a]. This paper is a contribution to making methods for satellite time series analysis available to a larger audience. In this paper we describe the \pkg{dtwSat} package, written in \proglang{R} [@R:2015] and \proglang{Fortran} programming languages, and available from the Comprehensive R Archive Network at [http://CRAN.R-project.org/package=dtwSat](http://CRAN.R-project.org/package=dtwSat). The package provides an implementation of Time-Weighted Dynamic Time Warping (TWDTW) [@Maus:2016] for satellite image time series analysis. -The TWDTW method is an adaptation of the well-known dynamic time warping (DTW) method for time series analysis [@Velichko:1970; @Sakoe:1971; @Sakoe:1978; @Rabiner:1993; @Berndt:1994; @Keogh:2005; @Muller:2007] for land use and land cover classification. The standard DTW compares a temporal signature of a known event (*e.g.*, a person's speech) with an unknown time series. It finds all possible alignments between two time series and provides a dissimilarity measure [@Rabiner:1993]. In contrast to standard DTW, the TWDTW method is sensitive to seasonal changes of natural and cultivated vegetation types. It also considers inter-annual climatic and seasonal variability. In a tropical forest area, the method has achieved a high accuracy for mapping classes of single cropping, double cropping, forest, and pasture [@Maus:2016]. +The TWDTW method is an adaptation of the well-known dynamic time warping (DTW) method for time series analysis [@Velichko:1970; @Sakoe:1971; @Sakoe:1978; @Rabiner:1993; @Berndt:1994; @Keogh:2005; @Muller:2007] for land cover classification. The standard DTW compares a temporal signature of a known event (*e.g.*, a person's speech) with an unknown time series. It finds all possible alignments between two time series and provides a dissimilarity measure [@Rabiner:1993]. In contrast to standard DTW, the TWDTW method is sensitive to seasonal changes of natural and cultivated vegetation types. It also considers inter-annual climatic and seasonal variability. In a tropical forest area, the method has achieved a high accuracy for mapping classes of single cropping, double cropping, forest, and pasture [@Maus:2016]. + +We chose \proglang{R} because it is an open source software that offers a large number of reliable packages. The \pkg{dtwSat} package builds upon on a number of graphical and statistical tools in \proglang{R}: \pkg{dtw} [@Giorgino:2009], \pkg{proxy} [@Meyer:2015], \pkg{zoo} [@Zeileis:2005], \pkg{mgcv} [@Wood:2000; @Wood:2003; @Wood:2004; @Wood:2006; @Wood:2011], \pkg{sp} [@Pebesma:2005; @Bivand:2013], \pkg{raster} [@Hijmans:2015], \pkg{caret} [@Kuhn:2016], and \pkg{ggplot2} [@Wickham:2009]. Other \proglang{R} packages that are related and useful for remote sensing and land cover analysis include \pkg{landsat} [@Goslee:2011], \pkg{rgdal} [@Bivand:2015], \pkg{spacetime} [@Pebesma:2012; @Bivand:2013], \pkg{bfast} [@Verbesselt:2010; @Verbesselt:2010a], \pkg{bfastmonitor} [@Verbesselt:2011], \pkg{bfastSpatial} [@Dutrieux:2014], \pkg{MODISTools} [@Tuck:2014], \pkg{maptools} [@Bivand:2015], and \pkg{lucc} [@Moulds:2015]. Using existing packages as building blocks, software developers in \proglang{R} save a lot of time and can concentrate on their intended goals. -We chose \proglang{R} because it is an open source software that offers a large number of reliable packages. The \pkg{dtwSat} package builds upon on a number of graphical and statistical tools in \proglang{R}: \pkg{dtw} [@Giorgino:2009], \pkg{proxy} [@Meyer:2015], \pkg{zoo} [@Zeileis:2005], \pkg{mgcv} [@Wood:2000; @Wood:2003; @Wood:2004; @Wood:2006; @Wood:2011], \pkg{sp} [@Pebesma:2005; @Bivand:2013], \pkg{raster} [@Hijmans:2015], \pkg{caret} [@Kuhn:2016], and \pkg{ggplot2} [@Wickham:2009]. Other \proglang{R} packages that are related and useful for remote sensing and land use analysis include \pkg{landsat} [@Goslee:2011], \pkg{rgdal} [@Bivand:2015], \pkg{spacetime} [@Pebesma:2012; @Bivand:2013], \pkg{bfast} [@Verbesselt:2010; @Verbesselt:2010a], \pkg{bfastmonitor} [@Verbesselt:2011], \pkg{bfastSpatial} [@Dutrieux:2014], \pkg{MODISTools} [@Tuck:2014], \pkg{maptools} [@Bivand:2015], and \pkg{lucc} [@Moulds:2015]. Using existing packages as building blocks, software developers in \proglang{R} save a lot of time and can concentrate on their intended goals. +There is already an \proglang{R} package that implements the standard DTW method for time series analysis: the \pkg{dtw} package [@Giorgino:2009]. In the \pkg{dtwSat} package, we focus on the specific case of satellite image time series analysis. The analysis method implemented in \pkg{dtwSat} package extends that of the \pkg{dtw} package; it adjusts the standard DTW method to account for the seasonality of different types of land cover. Our aim is to support the full cycle of land cover classification, from selecting sample patterns to visualising and assessing the final result. -There is already an \proglang{R} package that implements the standard DTW method for time series analysis: the \pkg{dtw} package [@Giorgino:2009]. In the \pkg{dtwSat} package, we focus on the specific case of satellite image time series analysis. The analysis method implemented in \pkg{dtwSat} package extends that of the \pkg{dtw} package; it adjusts the standard DTW method to account for the seasonality of different types of land cover. Our aim is to support the full cycle of land use and land cover classification, from selecting sample patterns to visualising and evaluating the final result. +This paper focuses on the motivation and guidance for using the TWDTW method for remote sensing applications. The full description of the method is available in a paper published by the lead author [@Maus:2016]. In what follows, the \autoref{the-time-weighted-dynamic-time-warping-method} describes the application of TWDTW [@Maus:2016] for satellite time series analysis. The \autoref{dtwsat-package-overview} gives an overview of the \pkg{dtwSat} package. Then, \autoref{classifying-a-time-series} focuses on the analysis of a single time series and shows some visualisation methods. We then present an example of a complete land cover change analysis for a study area in Mato Grosso, Brazil in \autoref{producing-a-land-cover-map}. -This paper focuses on the motivation and guidance for using the TWDTW method for remote sensing applications. The full description of the method is available in a paper published by the lead author [@Maus:2016]. In what follows, Section \ref{dtwsat-package-overview} gives an overview of the \pkg{dtwSat} package. The Section \ref{the-time-weighted-dynamic-time-warping-method} describes the application of TWDTW [@Maus:2016] for satellite time series analysis. Then, Section \ref{classifying-a-time-series} focuses on the analysis of a single time series and shows some visualisation methods. We then present an example of a complete land use and land cover change analysis for a study area in the Mato Grosso, Brazil in Section \ref{producing-a-land-cover-map}. # The Time-Weighted Dynamic Time Warping method -In this section, we describe the Time-Weighted Dynamic Time Warping (TWDTW) algorithm in general terms. For a detailed technical explanation, refer to @Maus:2016. TWDTW is time-constrained version of the Dynamic Time Warping (DTW) algorithm. Although the standard DTW method is good for shape matching [@Keogh:2005], it is not suited *per se* for satellite image time series analysis, since it disregards the temporal range when finding the best matches between two time series [@Maus:2016]. When using image time series for land cover classification, one needs to balance between shape matching and temporal alignment, since each land cover class has a distinct phenological cycle associated with the vegetation [@Reed:1994,@Zhang:2003]. For example, soybeans and maize cycles range from 90 to 120 days, whereas sugar-cane has a 360 to 720 days cycle. A time series with cycle larger than 180 days is unlikely to come from soybeans or maize. For this reason, @Maus:2016 include a time constraint in DTW to account for seasonality. The resulting method is capable of distinguishing different land use and land cover classes. +In this section, we describe the Time-Weighted Dynamic Time Warping (TWDTW) algorithm in general terms. For a detailed technical explanation, refer to @Maus:2016. TWDTW is time-constrained version of the Dynamic Time Warping (DTW) algorithm. Although the standard DTW method is good for shape matching [@Keogh:2005], it is not suited *per se* for satellite image time series analysis, since it disregards the temporal range when finding the best matches between two time series [@Maus:2016]. When using image time series for land cover classification, one needs to balance between shape matching and temporal alignment, since each land cover class has a distinct phenological cycle associated with the vegetation [@Reed:1994,@Zhang:2003]. For example, soybeans and maize cycles range from 90 to 120 days, whereas sugar-cane has a 360 to 720 days cycle. A time series with cycle larger than 180 days is unlikely to come from soybeans or maize. For this reason, @Maus:2016 include a time constraint in DTW to account for seasonality. The resulting method is capable of distinguishing different land cover classes. -The inputs to TWDTW are: (a) a set of time series of known temporal patterns (*e.g.*, phenological cycles of land cover classes); (b) an unclassified long-term satellite image time series. For each temporal pattern, the algorithm finds all matching subintervals in the long-term time series, providing a dissimilarity measure (cf. Figure \ref{fig:twdtw-example}). The result of the algorithm is a set of subintervals, each associated with a pattern and with a dissimilarity measure. We then break the unclassified time series in periods according to our needs (*e.g.*, yearly, seasonality, monthly). For each period, we consider all matching subintervals that intersect with it, and classify them based on the land cover class of the best matching subinterval. In this way, the long-term satellite time series is divided in periods, and each period is assigned a land cover class. +The inputs to TWDTW are: (a) a set of time series of known temporal patterns (*e.g.*, phenological cycles of land cover classes); (b) an unclassified long-term satellite image time series. For each temporal pattern, the algorithm finds all matching subintervals in the long-term time series, providing a dissimilarity measure (cf. \autoref{fig:twdtw-example}). The result of the algorithm is a set of subintervals, each associated with a pattern and with a dissimilarity measure. We then break the unclassified time series in periods according to our needs (*e.g.*, yearly, seasonality, monthly). For each period, we consider all matching subintervals that intersect with it, and classify them based on the land cover class of the best matching subinterval. In this way, the long-term satellite time series is divided in periods, and each period is assigned a land cover class. ```{r twdtw-example, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_height/3.5, fig.align='center', fig.cap='Matches of the known temporal pattern to subintervals of the long-term time series. The solid black line is the long-term time series, the colored lines are the different matches of the same pattern ordered by TWDTW dissimilarity measure, and the gray dashed lines are the matching points.', fig.pos='h'} n=4 @@ -143,12 +139,12 @@ df_dist$y = 1.8 plotMatches(mat, attr="evi", k=n) + ylab("Time series Pattern") + geom_text(data=df_dist, mapping = aes_string(x='to', y='y', label='label'), - size = 2) + + size = 2, family="Helvetica") + theme(legend.position="none") ``` -To use TWDTW for land use and land cover classification, we need the following data sets: +To use TWDTW for land cover classification, we need the following data sets: - A set of remote sensing time series for the study area. For example, a tile of a MODIS MOD13Q1 image consists of 4800 x 4800 pixels, covering an area of 10 degrees x 10 degrees at the Equator [@Friedl:2010]. A 15-year (2000-2015) MODIS MOD13Q1 set time series has 23 images per year, with a total of 23 million time series, each with 346 samples. @@ -156,7 +152,7 @@ To use TWDTW for land use and land cover classification, we need the following d - A set of ground truth points, with spatial and temporal information and land cover classification. These *ground truth* points are used for validation and accuracy assessment. -Based on the information provided by the user about the images to be analysed, our method maps them to a three-dimensional (3-D) array in space-time (Figure \ref{fig:3-D-array}). This array can have multiple attributes, such as the satellite bands (*e.g.*, "red", "nir", and "blue"), and derived indices (*e.g.*, "NDVI", "EVI", and "EVI2"). This way, each pixel location is associated to a sequence of measurements, building a satellite image time series. Figure \ref{fig:3-D-array} shows an example of "evi" time series for a location in the Brazilian Amazon from 2000 to 2008. In the first two years, the area was covered by forest that was cut in 2002. The area was then used for cattle raising (pasture) for three years, and then for crop production from 2006 to 2008. Satellite image time series are thus useful to describe the dynamics of the landscape and the land use trajectories. +Based on the information provided by the user about the images to be analysed, our method maps them to a three-dimensional (3-D) array in space-time (\autoref{fig:3-D-array}). This array can have multiple attributes, such as the satellite bands (*e.g.*, "red", "nir", and "blue"), and derived indices (*e.g.*, "NDVI", "EVI", and "EVI2"). This way, each pixel location is associated to a sequence of measurements, building a satellite image time series. \autoref{fig:3-D-array} shows an example of "EVI" time series for a location in the Brazilian Amazon from 2000 to 2008. In the first two years, the area was covered by forest that was cut in 2002. The area was then used for cattle raising (pasture) for three years, and then for crop production from 2006 to 2008. Satellite image time series are thus useful to describe the dynamics of the landscape and the land use trajectories. \begin{figure}[!h] \begin{center} @@ -168,6 +164,7 @@ Based on the information provided by the user about the images to be analysed, o \end{figure} + # dtwSat package overview \pkg{dtwSat} provides a set of functions for land cover change analysis using satellite image time series. This includes functions to build temporal patterns for land cover types, apply the TWDTW analysis using different weighting functions, visualise the results in a graphical interface, produce land cover maps, and create spatiotemporal plots for land changes. Therefore, \pkg{dtwSat} gives an end-to-end solution for satellite time series analysis, which users can make a complete land change analysis. @@ -185,6 +182,7 @@ The class \code{twdtwRaster} is used for satellite image time series. This class + # Classifying a time series This section describes how to classify one time series, using examples that come with the \pkg{dtwSat} package. We will show how to match three temporal patterns ("soybean", "cotton", and "maize") to subintervals of a long-term satellite image time series. These time series have been extracted from a set of MODIS MOD13Q1 [@Friedl:2010] images and include the vegetation indices "ndvi", "evi", and the original bands "nir", "red", "blue", and "mir". In this example, the classification of crop types for the long-term time series is known. @@ -193,17 +191,17 @@ This section describes how to classify one time series, using examples that come The inputs for the next examples are time series in \pkg{zoo} format. The first is an object of class \code{zoo} with a long-term time series, referred to as \code{MOD13Q1.ts}, and the second is a \code{list} of time series of class \code{zoo} with the temporal patterns of "soybean", "cotton", and "maize", referred to as \code{MOD13Q1.patterns.list}. -From \code{zoo} objects we construct time series of class \code{twdtwTimeSeries}, for which we have a set of visualization and analysis methods implemented in the \pkg{dtwSat} package. The code below builds two objects of class \code{twdtwTimeSeries}. The first has the long-term time series and second has the temporal patterns. We use the plot method types \code{timeseries} and \code{patterns} to shown the objects \code{ts} in Figure \ref{fig:example-timeseries} and \code{patterns_ts} in Figure \ref{fig:temporal-patterns-soy-cot-mai}, respectively. This plot method uses \code{ggplot} syntax. -```{r, echo = TRUE, eval = TRUE, results = 'markup'} -ts = twdtwTimeSeries(MOD13Q1.ts, labels="Time series") -patterns_ts = twdtwTimeSeries(MOD13Q1.patterns.list) -MOD13Q1.ts.labels -``` +From \code{zoo} objects we construct time series of class \code{twdtwTimeSeries}, for which we have a set of visualization and analysis methods implemented in the \pkg{dtwSat} package. The code below builds two objects of class \code{twdtwTimeSeries}. The first has the long-term time series and second has the temporal patterns. We use the plot method types \code{timeseries} and \code{patterns} to shown the objects \code{ts} in \autoref{fig:example-timeseries} and \code{MOD13Q1.ts} in \autoref{fig:temporal-patterns-soy-cot-mai}, respectively. This plot method uses \code{ggplot} syntax. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} library(dtwSat) names(MOD13Q1.patterns.list) head(MOD13Q1.ts, n = 2) ``` +```{r, echo = TRUE, eval = TRUE, results = 'markup'} +ts = twdtwTimeSeries(MOD13Q1.ts, labels="Time series") +patterns_ts = twdtwTimeSeries(MOD13Q1.patterns.list) +patterns_ts +``` @@ -216,7 +214,7 @@ plot(ts, type = "timeseries") + plot(patterns_ts, type = "patterns") ``` -TWDTW uses both amplitude and phase information to classify the phenological cycles in the long-term time series. The EVI peak of the "soybean" time series has a similar amplitude as that of "cotton". However, the "soybean" series peaks in late December while the "cotton" series peaks in early April. The EVI peak of the "maize" time series is at the same period as the peak of "cotton". However, the "maize" time series has smaller amplitude than the "cotton" one. Therefore, we can improve the time series classification by combining shape and time information. +TWDTW uses both amplitude and phase information to classify the phenological cycles in the long-term time series. The differences in the amplitude and phase of the cycles are more clear when we observe the EVI signal in Figures 3 and 4. The EVI peak of the "soybean" time series has a similar amplitude as that of "cotton". However, the "soybean" series peaks in late December while the "cotton" series peaks in early April. The EVI peak of the "maize" time series is at the same period as the peak of "cotton". However, the "maize" time series has smaller amplitude than the "cotton" one. Therefore, combining shape and time information we can improve the time series classification. ## Detection of time series patterns with TWDTW @@ -229,7 +227,7 @@ matches = slotNames(matches) show(matches) ``` -To retrieve the complete information of the matches we set \code{keep=TRUE}. We need this information for the plot methods of the class \code{twdtwMatches}. The argument \code{weight.fun} defines the time-weight to the dynamic time warping analysis [@Maus:2016]. By default the time-weight is zero, meaning that the function will run a standard dynamic time warping analysis. The arguments \code{x} and \code{y} are objects of class \code{twdtwTimeSeries} with the unclassified long-term time series and the temporal patterns, respectively. For details and other arguments see \code{?twdtwApply}. +To retrieve the complete information of the matches we set \code{keep=TRUE}. We need this information for the plot methods of the class \code{twdtwMatches}. The argument \code{weight.fun} defines the time-weight to the dynamic time warping analysis [@Maus:2016]. By default the time-weight is zero, meaning that the function will run a standard dynamic time warping analysis. The arguments \code{x} and \code{y} are objects of class \code{twdtwTimeSeries} with the unclassified long-term time series and the temporal patterns, respectively. To perform the alignment between the time series the default TWDTW recursion has a symmetric step (for more details and other recursion options see \code{?stepPattern}). @Giorgino:2009 provides a detaild discussion on the recursion steps and other step patterns. For further details and other arguments of the TWDTW analysis see \code{?twdtwApply}. In our example we use a logistic weight function for the temporal constraint of the TWDTW algorithm. This function is defined by \code{logisticWeight}. The \pkg{dtwSat} package provides two in-built functions: \code{linearWeight} and \code{logisticWeight}. The \code{linearWeight} function with slope \code{a} and intercept \code{b} is given by $$ @@ -241,7 +239,7 @@ $$ \omega = \frac{1}{1 + e^{-\alpha(g(t_1,t_2)-\beta)} }. \label{eq:nonlineartw} $$ -The function $g$ is the absolute difference in days between two dates, $t_1$ and $t_2$. The linear function creates a strong time constraint even for small time differences. The logistic function has a low weight for small time warps and significant costs for bigger time warps, cf. Figure \ref{fig:logist-time-weight}. In our previous studies [@Maus:2016] the logistic-weight had better results than the linear-weight for land cover classification. Users can define different weight functions as temporal constraints in the argument \code{weight.fun} of the \code{twdtwApply} method. +The function $g$ is the absolute difference in days between two dates, $t_1$ and $t_2$. The aim of these functions is to control the time warp, e.g. a "large time warp" is needed to match a point of the temporal pattern whose original date is January 1 to a point of the long-term time series whose date is July 1, on the other hand to match January 1 to December 15 has a "small time warp". If there is a large seasonal difference between the pattern and its matching point in time series, an extra cost is added to the DTW distance measure. This constraint controls the time warping and makes the time series alignment dependent on the seasons. This is especially useful for detecting temporary crops and for distinguishing pasture from agriculture. The linear function creates a strong time constraint even for small time differences, including small time warps. The logistic function has a low weight for small time warps and significant costs for bigger time warps, cf. \autoref{fig:logist-time-weight}. In our previous studies [@Maus:2016] the logistic-weight had better results than the linear-weight for land cover classification. Users can define different weight functions as temporal constraints in the argument \code{weight.fun} of the function \code{twdtwApply}. ```{r logist-time-weight, echo = FALSE, eval = TRUE, out.width=paste0(page_width/2,'in'), fig.align='center', fig.cap='Logistic time-weight function \\code{logisticWeight} with steepness \\code{alpha=-0.1} and midpoint \\code{beta=100}. The $x$ axis shows the absolute difference between two dates in days and the $y$ axis shows the time-weight \\citep{Maus:2016}.', fig.pos='!h'} # Maximum time difference in days max_diff = 366/2 @@ -262,9 +260,9 @@ df_weight = melt(df_weight, id.vars = "Difference") names(df_weight)[-1] = c("Functions","Weight") ggplot(df_weight, aes_string(x="Difference", y="Weight", group="Functions", linetype="Functions")) + geom_line() + xlab("Time difference (days)") + - theme(text = element_text(size = 10), - plot.title = element_text(size = 10, face="bold"), - axis.title = element_text(size = 10), + theme(text = element_text(size = 10, family="Helvetica"), + plot.title = element_text(size = 10, family="Helvetica", face="bold"), + axis.title = element_text(size = 10, family="Helvetica"), legend.position = c(.3,.85), legend.background = element_rect(fill="transparent")) + scale_linetype(guide_legend(title = "")) ``` @@ -273,43 +271,44 @@ ggplot(df_weight, aes_string(x="Difference", y="Weight", group="Functions", line \pkg{dtwSat} provides five ways to visualise objects of class \code{twdtwMatches} through the plot types: \code{matches}, \code{alignments}, \code{classification}, \code{path}, and \code{cost}. The plot type \code{matches} shows the matching points of the patterns in the long-term time series; the plot type \code{alignments} shows the alignments and dissimilarity measures; the plot type \code{path} shows the low cost paths in the TWDTW cost matrix; and the plot type \code{cost} allows the visualisation of the cost matrices (local cost, accumulated cost, and time cost); and the plot type \code{classification} shows the classification of the long-term time series based on the TWDTW analysis. The plot methods for class \code{twdtwMatches} return a \code{ggplot} object, so that users can further manipulate the result using the \pkg{ggplot2} package. For more details on visualisation functions, please refer to the \pkg{dtwSat} documentation in the CRAN [@Maus:2015a]. -We now describe the plot types \code{matches} and \code{alignments}. The code bellow shows how to visualise the matching points of the four best matches of "soybean" pattern in the long-term time series, cf. Figure \ref{fig:twdtw-matches}. +We now describe the plot types \code{matches} and \code{alignments}. The code bellow shows how to visualise the matching points of the four best matches of "soybean" pattern in the long-term time series, cf. \autoref{fig:twdtw-matches}. ```{r twdtw-matches, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/3.5, fig.align='center', fig.cap=c('The four best matches of the "soybean" pattern in the time series using a logistic time-weight. The solid black line is the long-term time series; the coloured lines are the temporal patterns; and the grey dashed lines are the respective matching points.'), fig.pos='!h'} plot(matches, type="matches", patterns.labels="Soybean", k=4) ``` -The next example (Figure \ref{fig:alignments-all-patterns}) uses the plot type \code{alignments} to show the alignments of the temporal patterns. We set the threshold for the dissimilarity measure to be lower than $3.0$. This is useful to display the different subintervals of the long-term time series that have at least one alignment whose dissimilarity is less than the specified threshold. +The next example uses the plot type \code{alignments} to show the alignments of the temporal patterns (see \autoref{fig:alignments-all-patterns}). We set the threshold for the dissimilarity measure to be lower than $3.0$. This plot displays the different subintervals of the long-term time series that have alignments whose dissimilarity is less than the specified threshold. ```{r alignments-all-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.5, fig.align='center', fig.cap=c('Alignments and dissimilarity measures of the patterns "soybean", "cotton", and "maize" to the subintervals of the long-term time series using a logistic time-weight. The solid black line is the EVI time series, and the coloured lines are the alignments of the patterns that have dissimilarity measure lower than three.'), fig.pos='!h'} plot(matches, type="alignments", attr = "evi", threshold = 3.0) ``` +\autoref{fig:alignments-all-patterns} shows the alignments of each pattern over the long-term time series, note that we can rank the alignments by their TWDTW dissimilarity, i.e. alignments overlapping the same period usually have distinct dissimilarity, which can be used to rank them. In the figure we can see that maize (blue lines) and cotton (green lines) overlap approximately the same time periods, however, they have distinct dissimilarity measures, and therefore, can be ranked. Observing the time period from January 2010 to July 2010, both soybean, maize, and cotton have at least one overlapping alignment, however in this case the cotton pattern matches better to the interval because its dissimilarity is lower than the others. + ## Classifying the long-term time series -Using the matches and their associated dissimilarity measures, we can classify the subintervals of the long-term time series using \code{twdtwClassify}. To do this, we need to define a period for classification and the minimum overlap between the period and the alignments that intersect with it. We use the plot type \code{classification} to show the classification of the subintervals of the long-term time series based on the TWDTW analysis. For this example, we set classification periods of 6 months from September 2009 to September 2013, and a minimum overlap of 50% between the alignment and the classification period. This means that at least 50% of the alignment has to be contained inside of the classification period. -```{r time-series-classification, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.5, fig.align='center', fig.cap=c('Classification of each 6 months periods of the time series using results of the TWDTW analysis with logistic time-weight. The solid lines are the attributes of the time series, the background colours indicate the classification of the periods.'), fig.pos='!h'} +Using the matches and their associated dissimilarity measures, we can classify the subintervals of the long-term time series using \code{twdtwClassify}. To do this, we need to define a period for classification and the minimum overlap between the period and the alignments that intersect with it. For each interval, \code{twdtwClassify} will select the alignment that has the lowest TWDTW dissimilarity taking into account the minimum overlap condition. For example, in Figure 7 the interval from 1 September 1012 to 28 February 2013 has three overlapping alignments, maize in blue, cotton in green, and soybean in red. Without a minimum overlap the function \code{twdtwClassify} would classify this interval as maize, which has the lowest dissimilarity in the period. However, if we set a minimum overlap of 50\%, the function \code{twdtwClassify} classifies the interval as soybean, which is the only class whose alignment overlaps the interval during more than 50\% of the time. The interval of classification are usually defined according to the phenological cycles or the agricultural calendar of the region. The classification interval can also be irregular, for details see the argument \code{breaks} in \code{?twdtwClassify} + +In the example bellow we classify each period of 6 months from September 2009 to September 2013; we set a minimum overlap of 50% between the alignment and the classification period. This means that at least 50% of the alignment has to be contained inside of the classification period. We also use the plot type \code{classification} to show the classified subintervals of the long-term time series. + +```{r time-series-classification, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.8, fig.align='center', fig.cap=c('Classification of each 6 months periods of the time series using results of the TWDTW analysis with logistic time-weight. The solid lines are the attributes of the time series, the background colours indicate the classification of the periods.'), fig.pos='!ht'} ts_classification = twdtwClassify(x = matches, from = as.Date("2009-09-01"), to = as.Date("2013-09-01"), by = "6 month", overlap = 0.5) plot(ts_classification, type="classification") ``` -Comparing the results of the classified time series in Figure \ref{fig:time-series-classification} with the crop cycles in Figure \ref{fig:example-timeseries} we see that the algorithm has classified correctly all the eight subintervals from 2009 to 2013, which are, respectively: "Soybean", "Cotton", "Soybean", "Cotton", "Soybean", "Maize", "Soybean", "Maize". +By comparing the results of the classified time series in \autoref{fig:time-series-classification} with the crop cycles in \autoref{fig:example-timeseries} we see that the algorithm has classified correctly all the eight subintervals from 2009 to 2013, which are, respectively: "Soybean", "Cotton", "Soybean", "Cotton", "Soybean", "Maize", "Soybean", "Maize". -# Producing a land cover map - -In this section we present an application of TWDTW for land use and land cover change analysis using satellite image time series. Our input is a set of images, each covering the same geographical area at different times. Each pixel location is then associated to an unclassified satellite image time series. We assume to have done field work in the area; for some pixel locations and time periods, we know what is the land cover. We then will show how to obtain a set of template patterns, based on the field samples and how to apply these patterns to land cover classification of the set of images. In the end of this section we show how to perform land cover change analysis and how to do accuracy assessment. The satellite images and the field samples used in the examples come with \pkg{dtwSat} package. -Our method is not restricted to cases where the temporal patterns are obtained from the set of images. The patterns for the TWDTW analysis can be any time series with same bands or indices as the unclassified images, such as in the examples of Section \ref{classifying-a-time-series} above. -## Input data +# Producing a land cover map -The inputs are: *a)* the satellite images for a given geographical area, organised as a set of georeferenced raster files in GeoTIFF format, each containing all time steps of a spectral band or index; and *b)* a set of ground truth samples. The satellite images are extracted from the MODIS product MOD13Q1 collection 5 [@Friedl:2010] and include vegetation indexes "ndvi", "evi", and original bands "nir", "red", "blue", and "mir". This product has 250 x 250 m spatial and 16 day temporal resolution. +In this section we present an application of TWDTW for land cover change analysis using satellite image time series. Our input is a set of images, each covering the same geographical area at different times. Each pixel location is then associated to an unclassified satellite image time series. We assume to have done field work in the area; for some pixel locations and time periods, we know what is the land cover. We then will show how to obtain a set of template patterns, based on the field samples and how to apply these patterns to land cover classification of the set of images. In the end of this section we show how to perform land cover change analysis and accuracy assessment. -The region is a tropical forest area in Mato Grosso, Brazil of approximately 5300 km$^2$ with images from 2007 to 2013 (Figure \ref{fig:study-area}). This is a sequence of 160 images with 999 pixels each for 6 years. We also have a set of 603 ground truth samples of the following classes: "forest", "cotton-fallow", "soybean-cotton", "soybean-maize", and "soybean-millet". +As an example we classify approximately 5300 km$^2$ in a tropical forest region in Mato Grosso, Brazil (\autoref{fig:study-area}). This is a sequence of 160 images with 999 pixels each for 6 years, from 2007 to 2013. We also have a set of 603 ground truth samples of the following classes: "Forest", "Cotton-fallow", "Soybean-cotton", "Soybean-maize", and "Soybean-millet". The satellite images and the field samples used in the examples come with \pkg{dtwSat} package. -\begin{figure}[!h] +\begin{figure}[!ht] \begin{center} \includegraphics[width=\textwidth]{./study_area.pdf} \end{center} @@ -317,13 +316,17 @@ The region is a tropical forest area in Mato Grosso, Brazil of approximately 530 \label{fig:study-area} \end{figure} +## Input data + +The inputs are: *a)* the satellite images for a given geographical area, organised as a set of georeferenced raster files in GeoTIFF format, each containing all time steps of a spectral band or index; and *b)* a set of ground truth samples. The satellite images are extracted from the MODIS product MOD13Q1 collection 5 [@Friedl:2010] and include vegetation indices "ndvi", "evi", and original bands "nir", "red", "blue", and "mir". This product has 250 x 250 m spatial resolution and a 16 day maximum-value composite (MVC) for each pixel location [@Friedl:2010], meaning that one image can have measurements from different dates. For this reason, MOD13Q1 also includes the "day of the year" (doy) of each pixel as a layer, which we use to keep the time series consistent with the measurements. + The data files for the examples that follow are in the \pkg{dtwSat} installation folder *lucc_MT/data/*. The *tif* files include the time series of "blue", "red", "nir", "mir", "evi", "ndvi", and "doy" (day of the year); the text file *timeline* has the dates of the satellite images; the CSV file *samples.csv* has the \code{longitude}, \code{latitude}, \code{from}, \code{to}, and \code{label} for each field sample; and the text file *samples_projection* contains information about the cartographic projection of the samples, in the format of coordinate reference system used by \code{sp::CRS}. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} data_folder = system.file("lucc_MT/data", package = "dtwSat") dir(data_folder) ``` -In this example, we have stored all the time series for each band in one single file. In this way, we can use the function \code{raster::brick} to read the satellite images. The algorithm also works when the time steps for each band are split in many files. In this case, the user should call the function \code{raster::stack} with the appropriate parameters. Because of processing performance, we suggest that interested users group their images in bricks and follow the procedures given below. +We have stored all the time series for each band in one single file. In this way, we can use the function \code{raster::brick} to read the satellite images. The algorithm also works when the time steps for each band are split in many files. In this case, the user should call the function \code{raster::stack} with the appropriate parameters. Because of processing performance, we suggest that interested users group their images in bricks and follow the procedures given below. ```{r, echo = TRUE, eval = TRUE} blue = brick(paste(data_folder,"blue.tif", sep = "/")) red = brick(paste(data_folder,"red.tif", sep = "/")) @@ -335,10 +338,16 @@ day_of_year = brick(paste(data_folder,"doy.tif", sep = "/")) dates = scan(paste(data_folder,"timeline", sep = "/"), what = "dates") ``` -The set of ground truth samples in the CSV file *samples.csv* has a total of 603 samples divided in five classes: 68 "cotton-fallow", 138 "forest", 79 "soybean-cotton", 134 "soybean-maize", and 184 "soybean-millet". Reading this CSV file, we get a \code{data.frame} object, with the spatial location (\code{latitude} and \code{longitude}), starting and ending dates (\code{from} and \code{to}), and the \code{label} for each sample. +We use these data-sets to create a multiple raster time series, which is used in the next sections for the TWDTW analysis. \pkg{dtwSat} provides the constructor \code{twdtwRaster} that builds a multi-band satellite image time series. The inputs of this function are \code{RasterBrick} objects with the same temporal and spatial extents, and a vector (\code{timeline}) with the acquisition dates of the images in the format \code{"YYYY-MM-DD"}. The argument \code{doy} is combined with \code{timeline} to get the real date of each pixel, independently from each other. If \code{doy} is not provided then the dates of the pixels are given by \code{timeline}, i.e. all pixels in one image will have the same date. Products from other sensors, such as the Sentinels and Landsat, usually have all pixels with same date, therefore the argument \code{doy} is not needed. This function produces an object of class \code{twdtwRaster} with the time series of multiple satellite bands. +```{r, echo = TRUE, eval = TRUE} +raster_timeseries = twdtwRaster(blue, red, nir, mir, evi, ndvi, + timeline = dates, doy = day_of_year) +``` + +Our second input is a set of ground truth samples in the CSV file *samples.csv*, which has a total of 603 samples divided in five classes: 68 "cotton-fallow", 138 "forest", 79 "soybean-cotton", 134 "soybean-maize", and 184 "soybean-millet". Reading this CSV file, we get a \code{data.frame} object, with the spatial location (\code{latitude} and \code{longitude}), starting and ending dates (\code{from} and \code{to}), and the \code{label} for each sample. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} field_samples = read.csv(paste(data_folder,"samples.csv", sep = "/")) -head(field_samples, 2) +head(field_samples, 5) table(field_samples[["label"]]) proj_str = scan(paste(data_folder,"samples_projection", sep = "/"), what = "character") @@ -346,186 +355,151 @@ proj_str ``` -## Creating the time series and the temporal patterns +## Assessing the separability of temporal patterns -After reading our data, we need to create the time series for analysis. For this purpose, \pkg{dtwSat} provides the constructor \code{twdtwRaster} that builds a multi-band satellite image time series. The inputs of this function are \code{RasterBrick} objects with the same temporal and spatial extents, and a vector (\code{timeline}) with the acquisition dates of the images in the format \code{"YYYY-MM-DD"}. The argument \code{doy} is optional. If \code{doy} is not declared, the function builds a \code{RasterBrick} object using the dates given by \code{timeline}. This function produces an object of class \code{twdtwRaster} with the time series of multiple satellite bands. -```{r, echo = TRUE, eval = TRUE} -raster_timeseries = twdtwRaster(blue, red, nir, mir, evi, ndvi, - timeline = dates, doy = day_of_year) -``` +The classification is highly dependent on the quality of the temporal patterns. Therefore, it is useful to perform an analysis to assess the separability of the temporal pattern. Ideally, one would need patterns that, when applied to the set of unknown time series, produce consistent results (see the guidelines for single time series analysis in \autoref{classifying-a-time-series}). For this reason, before performing the land cover mapping, we perform a cross validation step. In this way, the users would assess the separability of their patterns before classifying a large data set. -We now need to identify the temporal patterns. Usually, this can be done using the collected field samples. In the next example we use the function \code{getTimeSeries} to get the time series of each field sample from our raster time series. The arguments of the function are a set of raster time series, a \code{data.frame} with spatial and temporal information about the fields samples (as in the object \code{field_samples} given above), and a \code{proj4string} with the projection information. The projection should follow the \code{sp::CRS} format. The result is an object of class \code{twdtwTimeSeries} with one time series for each field sample. +In the next example we use the function \code{getTimeSeries} to extract the time series of each field sample from our raster time series. The arguments of the function are a set of raster time series, a \code{data.frame} with spatial and temporal information about the fields samples (as in the object \code{field_samples} given above), and a \code{proj4string} with the projection information. The projection should follow the \code{sp::CRS} format. The result is an object of class \code{twdtwTimeSeries} with one time series for each field sample. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} field_samples_ts = getTimeSeries(raster_timeseries, y = field_samples, proj4string = proj_str) field_samples_ts ``` -After obtaining the time series associated to the field samples, we need to create the template patterns for each class. For this purpose, \pkg{dtwSat} provides the function \code{createPatterns}. This function fits a Generalized Additive Model (GAM) [Hastie:1986,Wood:2011] to the field samples and retrieves a smoothed temporal pattern for each band (*e.g.*, "blue", "red", "nir", "mir", "evi", and "ndvi"). We use the GAM because of its flexibility for non-parametric fits, with less rigorous assumptions on the relationship between response and predictor. This potentially provides better fit to satellite data than purely parametric models, due to the data's inter- and intra-annual variability. +To perform the cross-validation we use the function \code{twdtwCrossValidate}. This function splits the sample time series into training and validation sets using stratified sampling with a simple random sampling within each stratum, for details see \code{?caret::createDataPartition}. The function uses the training samples to create the temporal patterns and then classifies the remaining validation samples using \code{twdtwApply}. The results of the classification are used in the accuracy calculation. + +A Generalized Additive Model (GAM) [Hastie:1986,Wood:2011] generates the smoothed temporal patterns based on the training samples. We use the GAM because of its flexibility for non-parametric fits, with less rigorous assumptions on the relationship between response and predictor. This potentially provides better fit to satellite data than purely parametric models, due to the data's inter- and intra-annual variability. + +In the next example we set the arguments \code{times=100} and \code{p=0.1}, which creates 100 different data partitions, each with 10% of the samples for training and 90% for validation. The other arguments of this function are: the logistic weight function with steepness `-0.1` and midpoint `50` to \code{weight.fun}; the frequency of the temporal patterns to 8 days \code{freq=8}, and GAM smoothing formula to \code{formula = y ~ s(x)}, where function \code{s} sets up a spline model, with \code{x} the time and \code{y} a satellite band (for details see \code{?mgcv::gam} and \code{?mgcv::s}). The output is an object of class \code{twdtwCrossValidation} which includes the accuracy for each data partition. The object has two slots, the first called \code{partitions} has the index of the samples used for training, the second called \code{accuracy} has overall accuracy, user's accuracy, producer's accuracy, error matrix, and the data used in the calculation, i.e. reference classes, predicted classes, and TWDTW distance measure. + +```{r, echo = TRUE, eval = FALSE, results = 'markup'} +set.seed(1) +log_fun = logisticWeight(alpha=-0.1, beta=50) +cross_validation = twdtwCrossValidate(field_samples_ts, times=100, p=0.1, + freq = 8, formula = y ~ s(x, bs="cc"), weight.fun = log_fun) +``` +```{r, echo = FALSE, eval = TRUE} +load(system.file("lucc_MT/cross_validation.RData", package = "dtwSat")) +``` + +\autoref{fig:plot-accuracy} and \autoref{tab:cross-validation} show the 95% confidence interval of the mean for user\'s and producer\'s accuracy derived from the hundred-fold cross-validation analysis. The user\'s accuracy gives the confidence and the producer\'s accuracy gives the sensitivity of the method for each class. In our analysis all classes had high user\'s and producer\'s accuracy, meaning that TWDTW has high confidence and sensitivity for the classes included in the analysis. The cross-validation results show that if we randomly select 10% of our sampels to create temporal patterns we can get an overall accuracy of at least 97% in the classification of the remaining samples with 95% confidence level. +```{r plot-accuracy, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_width/2, fig.align='center', fig.cap='User\'s and producer\'s accuracy of the TWDTW cross-validation using 100 resampling-with-replacement. The plot shows the 95\\% confidence interval of the mean.', fig.pos='!ht'} +plot(cross_validation, conf.int=.95) +``` + +```{r, echo = FALSE, eval = TRUE, results = 'asis'} +twdtwXtable(cross_validation, conf.int=.95, digits = 2, caption="\\label{tab:cross-validation} User\'s and producer\'s accuracy of the TWDTW cross-validation using 100 resampling-with-replacement. The table shows the standard deviation ($\\sigma$) and the 95\\% confidence interval (ci) of the mean ($\\mu$).'", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` + + + +## Creating temporal patterns + +In the last section we observed that the land cover classes based on our samples are separable using the TWDTW algorithm with high confidence level. Now we randomly select 10% of our samples for training and keep the remaining 90% for validation. The first set of samples are used to create temporal patterns and classify the raster time series, and the second set of samples to assess the final maps. Ideally we would need a second independent set of samples to assess the map, but it would be very difficult to identify different crops without field work. Therefore, we use the same samples used in the cross-validation (\autoref{assessing-the-separability-of-temporal-patterns}). +```{r, echo = TRUE, eval = TRUE} +library(caret) +set.seed(1) +I = unlist(createDataPartition(field_samples[,"label"], p = 0.1)) +training_ts = subset(field_samples_ts, I) +validation_samples = field_samples[-I,] +``` + +We use the function \code{createPatterns} to produce the temporal patterns based on the training samples. For that, we need to set the desired temporal frequency of the patterns and the smoothing function for the GAM model. In the example below, we set \code{freq=8} to get temporal patterns with a frequency of 8 days, and the GAM smoothing formula \code{formula = y ~ s(x)}, such as in \autoref{assessing-the-separability-of-temporal-patterns}). -To produce the set of template patterns using the function \code{createPatterns}, we need to set the temporal frequency of the resulting patterns and the smoothing function for the GAM model. In the example below, we set \code{freq=8} to get temporal patterns with a frequency of 8 days. We also set the GAM smoothing formula to be \code{formula = y ~ s(x)}, where function \code{s} sets up a spline model, with \code{x} the time and \code{y} a satellite band (for details see \code{?mgcv::gam} and \code{?mgcv::s}). ```{r, echo = TRUE, eval = TRUE} temporal_patterns = - createPatterns(field_samples_ts, freq = 8, formula = y ~ s(x)) + createPatterns(training_ts, freq = 8, formula = y ~ s(x)) ``` -We use the plot method \code{type="patterns"} to show the results of the \code{createPatterns} in \autoref{fig:temporal-patterns}. -```{r temporal-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_width/1.5, fig.align='center', fig.pos='!h', fig.cap='Temporal patterns of forest, cotton-fallow, soybean-cotton, soybean-maize, and soybean-millet based on the ground truth samples.'} +The result of the function \code{createPatterns} is an object of the class \code{twdtwTimeSeries}. We use the plot method \code{type="patterns"} to show the results of the \code{createPatterns} in \autoref{fig:temporal-patterns}. +```{r temporal-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_width/1.5, fig.align='center', fig.pos='!h', fig.cap='Temporal patterns of Forest, Cotton-fallow, Soybean-cotton, Soybean-maize, and Soybean-millet based on the ground truth samples.'} plot(temporal_patterns, type = "patterns") + theme(legend.position = c(.8,.25)) ``` -After obtaining the template patterns for each land cover class, it is useful to perform a pre-classification analysis to assess their quality and their informational content. Ideally, one would need template patterns that, when applied to the set of unknown time series, produce consistent results. For this reason, it is advisable that the user performs a pre-classification step, along the lines of the individual analysis described in Section \ref{classifying-a-time-series}. In this way, the users would assess how good their patterns are before classifying a large data set. +Our method is not restricted to cases where the temporal patterns are obtained from the set of images, such as in the example above. Once can also use patterns from a different set of images or defined in other studies, as long as these temporal patterns stand for the study area and their bands match the bands in the unclassified images. ## Classifying the image time series -After obtaining a consistent set of temporal patterns, we use the function \code{twdtwApply} to run the TWDTW analysis for each pixel location in the raster time series. The input raster time series in the object \code{twdtwRaster} should be longer or have approximatly the same length as the temporal patterns. This function retrieves an object of class \code{twdtwRaster} with the TWDTW dissimilarity measure of the temporal patterns for each time period. The arguments \code{overwrite} and \code{format} are passed to \code{raster::writeRaster}. The arguments \code{weight.fun} and \code{overlap} are described in Section \ref{classifying-a-time-series}. Here we set the parameters of the time weight (logistic function) base on our the experience about the phenological cycle of the vegetation in the study area. In the next example, we classify the raster time series using the temporal patterns in \code{temporal_patterns} obtained as described above. The result is a \code{twdtwRaster} with five layers; each layer contains the TWDTW dissimilarity measure for one temporal pattern over time. We use the plot type \code{distance} to illustrate the TWDTW dissimilarity for each temporal pattern in 2013, cf. Figure \ref{fig:plot-dissmilarity2013}. +After obtaining a consistent set of temporal patterns, we use the function \code{twdtwApply} to run the TWDTW analysis for each pixel location in the raster time series. The input raster time series in the object \code{twdtwRaster} should be longer or have approximatly the same length as the temporal patterns. This function retrieves an object of class \code{twdtwRaster} with the TWDTW dissimilarity measure of the temporal patterns for each time period. The arguments \code{overwrite} and \code{format} are passed to \code{raster::writeRaster}. The arguments \code{weight.fun} and \code{overlap} are described in \autoref{classifying-a-time-series}. Here we set the parameters of the time weight (logistic function) base on our the experience about the phenological cycle of the vegetation in the study area. In the next example, we classify the raster time series using the temporal patterns in \code{temporal_patterns} obtained as described above. The result is a \code{twdtwRaster} with five layers; each layer contains the TWDTW dissimilarity measure for one temporal pattern over time. We use the plot type \code{distance} to illustrate the TWDTW dissimilarity for each temporal pattern in 2008, cf. \autoref{fig:plot-dissmilarity2008}. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} log_fun = logisticWeight(alpha=-0.1, beta=50) twdtw_dist = twdtwApply(x = raster_timeseries, y = temporal_patterns, overlap = 0.5, weight.fun = log_fun, overwrite=TRUE, format="GTiff") ``` - -```{r plot-dissmilarity2013, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Illustration of the TWDTW dissimilarity from each temporal pattern in 2013. The blue areas are more similar to the pattern and the red areas are less similar to the pattern.', fig.pos='!h'} -plot(x = twdtw_dist, type="distance", time.levels = 6) +```{r plot-dissmilarity2008, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Illustration of the TWDTW dissimilarity from each temporal pattern in 2008. The blue areas are more similar to the pattern and the red areas are less similar to the pattern.', fig.pos='!ht'} +plot(x = twdtw_dist, type="distance") ``` -The results of the example above can be used to create categorical land cover maps. The function \code{twdtwClassify} selects the most similar pattern for each time period and retrieves a \code{twdtwRaster} object with the time series of land use maps. The resulting object includes two layers, the first has the classified categorical maps and the second has the TWDTW dissimilarity measure. -```{r, echo = TRUE, eval = TRUE} -land_use_maps = twdtwClassify(twdtw_dist, format="GTiff", overwrite=TRUE) +The results of the example above can be used to create categorical land cover maps. The function \code{twdtwClassify} selects the most similar pattern for each time period and retrieves a \code{twdtwRaster} object with the time series of land cover maps. The resulting object includes two layers, the first has the classified categorical maps and the second has the TWDTW dissimilarity measure. +```{r, echo = TRUE, eval = TRUE, results = 'markup'} +land_cover_maps = twdtwClassify(twdtw_dist, format="GTiff", overwrite=TRUE) ``` ## Looking at the classification results -The classification results can be visualised using the \code{plot} methods of the class \code{twdtwRaster}, which supports four plot types: "maps", "area", "changes", and "distance". The \code{type="maps"} shows the land cover classification maps for each period, cf. Figure \ref{fig:plot-map}. -```{r plot-map, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Land use maps for each year from 2008 to 2013.', fig.pos='!h'} -plot(x = land_use_maps, type = "maps") +The classification results can be visualised using the \code{plot} methods of the class \code{twdtwRaster}, which supports four plot types: "maps", "area", "changes", and "distance". The \code{type="maps"} shows the land cover classification maps for each period, cf. \autoref{fig:plot-map}. +```{r plot-map, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Land cover maps for each year from 2008 to 2013.', fig.pos='!ht'} +plot(x = land_cover_maps, type = "maps") ``` -The next example shows the accumulated area for each class over time, using \code{type="area"}, cf. Figure \ref{fig:plot-area}. -```{r plot-area, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Percentage of area for each land use class from 2008 to 2013.', fig.pos='!h'} -plot(x = land_use_maps, type = "area") +The next example shows the accumulated area for each class over time, using \code{type="area"}, cf. \autoref{fig:plot-area}. +```{r plot-area, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Percentage of area for each land cover class from 2008 to 2013.', fig.pos='!ht'} +plot(x = land_cover_maps, type = "area") ``` -Users can also view the land cover transition for each time period, by setting \code{type="changes"}. For each land cover class and each period, the plot shows gains and losses in area from the other classes. This is the visual equivalent of a land transition matrix, cf. Figure \ref{fig:plot-change}. +Users can also view the land cover transition for each time period, by setting \code{type="changes"}. For each land cover class and each period, the plot shows gains and losses in area from the other classes. This is the visual equivalent of a land transition matrix, cf. \autoref{fig:plot-change}. ```{r plot-change, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Gains and losses in area from the other classes. The $y$ axis shows the actual class; the positive direction of $x$ axis shows the gains and the negative direction of $x$ axis shows the losses of the classes indicated in $y$. The colors indicate from/to which classes the gains/losses belong.', fig.pos='!h'} -plot(x = land_use_maps, type = "changes") +plot(x = land_cover_maps, type = "changes") ``` -We can also look at the dissimilarity of each classified pixel setting \code{type="distance"}. This plot can give a measure of the uncertainty of the classification of each pixel for each time period, cf. Figure \ref{fig:plot-dissmilarity}. -```{r plot-dissmilarity, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='TWDTW dissimilarity measure for each pixel over each classified period. The blue areas have high confidence and the red areas have low confidence in the classification.', fig.pos='!h'} -plot(x = land_use_maps, type="distance") +We can also look at the dissimilarity of each classified pixel setting \code{type="distance"}. This plot can give a measure of the uncertainty of the classification of each pixel for each time period, cf. \autoref{fig:plot-dissmilarity}. +```{r plot-dissmilarity, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='TWDTW dissimilarity measure for each pixel over each classified period. The blue areas have high confidence and the red areas have low confidence in the classification.', fig.pos='!ht'} +plot(x = land_cover_maps, type="distance") ``` -## Assessing the classification accuracy -In this section we show how to assess the accuracy of the TWDTW method for land cover classification. To do this, we split the ground truth samples into training and validation sets, using the function \code{splitDataset} from the package \pkg{dtwSat}. This function splits set of time series in the object \code{twdtwTimeSeries} for training and validation. The argument \code{p} defines the percentage used for training and the argument \code{times} gives the number of different partitions to create. This is a a stratified sampling with a simple random sampling within each stratum, see \code{?createDataPartition} for details. In the next example we create 100 different partitions of the data. Each partition uses 10% of the data for training and 90% for validation. The output is a list with 100 different data partitions; each partition has the temporal patterns based on the training samples and a set of time series for validation. -```{r, echo = TRUE, eval = FALSE} -set.seed(1) -partitions = splitDataset(field_samples_ts, p=0.1, times=100, - freq = 8, formula = y ~ s(x, bs="cc")) -``` -For each data partition we run the TWDTW analysis to classify the set of validation time series using the trained temporal patterns. The result is a list of \code{twdtwMatches} objects with the classified set of time series for each data partition. To compute the *User's Accuracy* (UA) and *Producer's Accuracy* (PA) of the classified time series we use the function \code{dtwSat::twdtwAssess} that retrieves a \code{data.frame} with the accuracy assessment for all data partitions. -```{r, echo = TRUE, eval = FALSE, results = 'markup'} -log_fun = logisticWeight(alpha=-0.1, beta=50) -twdtw_res = lapply(partitions, function(x){ - res = twdtwApply(x = x$ts, y = x$patterns, weight.fun = log_fun, n=1) - twdtwClassify(x = res) -}) -assessment = twdtwAssess(twdtw_res) -head(assessment, 5) -``` -```{r, echo = FALSE, eval = TRUE} -load(system.file("lucc_MT/cross_validation.RData", package = "dtwSat")) -``` - - -Figure \ref{fig:plot-accuracy} shows the average $\mu$ and standard deviation $\sigma$ of *user\'s* and *producer\'s accuracy* based on a bootstrap simulation of 100 different data partitions using resampling-with-replacement. The *user\'s accuracy* gives the confidence and the *producer\'s accuracy* gives the sensitivity of the method for each class. In our analysis all classes had high *user\'s* and *producer\'s accuracy*, meaning that TWDTW has high confidence and sensitivity for the classes included in the analysis. The average, standard deviation, and the 99\% confidence interval is also shown in Table \ref{tab:assessment}. -```{r plot-accuracy, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_width/2, fig.align='center', fig.cap='User\'s Accuracy (UA) and Producer\'s Accuracy (PA) of the TWDTW method for land cover classification. The plot shows the averages and their confidence interval for 99\\%.', fig.pos='!h'} -df = melt(assessment[,-1], id="label") -df$variable = factor(df$variable, levels = c("UA", "PA"), labels = c("User's Accuracy", "Producer's Accuracy")) -ggplot(df, aes(x=label, y=value)) + - stat_summary(fun.data="mean_cl_boot", fun.args=list(conf.int = .99), - width=0.5, geom="crossbar", size=0.1, fill = "gray") + - geom_point(size=0.2) + facet_grid(. ~ variable) + - scale_y_continuous(limits = c(0,1), labels = percent, breaks = seq(0,1,.2)) + - xlab("") + ylab("Accuracy") + coord_flip() +In this section we show how to assess the classification. \pkg{dtwSat} provides a function called \code{twdtwAssess}, which computes a set of accuracy metrics, and adjusted area such as proposed by @Olofsson:2013 and @Olofsson:2014. The inputs of this function are the classified map (an object of class \code{twdtwRaster}), and a set of samples for validation (an object of class \code{data.frame} or \code{sp::SpatialPointsDataFrame}). Besides coordinates, the samples should also have starting dates, ending dates, and lables compatible with the labels in the map (for details see \autoref{input-data}). The output of \code{twdtwAssess} is an object of class \code{twdtwAssessment}, which includes four slots: 1) \code{accuracyByPeriod} is a list of metrics for each time period, including overall accuracy, user's accuracy, produce's accuracy, error matrix (confusion matrix), and adjusted area; 2) \code{accuracySummary} has the accuracy and adjusted area accumulated over all time periods; 3) \code{data} is a \code{SpatialPointsDataFrame} with sample ID, period ID, starting date, ending date, reference label, predicted label, and TWDTW distance; and 4) \code{map} is a twdtwRaster with the raster maps. The next example uses \code{twdtwAssess} to compute the accuracy of the maps (\code{land_cover_maps}) using the validation samples (\code{validation_samples}) with a 95% confidence level. +```{r, echo = TRUE, eval = TRUE} +maps_assessment = twdtwAssess(land_cover_maps, y = validation_samples, + proj4string = proj_str, conf.int=.95) ``` + + +The results of the assessment in \autoref{tab:map-error-matrix}, \ref{tab:map-accuracy}, and \ref{tab:map-adjusted-area} are accumulated over the whole time period, i.e. the total mapped area is equal to the surface area times the number of maps. It is possible to assess and visualise each period independently from each other. However, our samples are irregularly distributed over time and some classes do not have samples in all time period, which limits the analysis of each time period independently from each other. + ```{r, echo = FALSE, eval = TRUE, results = 'asis'} -assess_mean = aggregate(assessment[, c("UA","PA")], list(assessment$label), mean) -assess_sd = aggregate(assessment[, c("UA","PA")], list(assessment$label), sd) -l_names = levels(assessment$label) -names(l_names) = l_names -ic_ua = t(sapply(l_names, function(i) 100*mean_cl_boot(x = assessment$UA[assessment$label==i], conf.int = .99)))[,-1] -ic_pa = t(sapply(l_names, function(i) 100*mean_cl_boot(x = assessment$PA[assessment$label==i], conf.int = .99)))[,-1] +twdtwXtable(maps_assessment, table.type="errormatrix", digits = 0, rotate.col = TRUE, caption="\\label{tab:map-error-matrix}Error matrix of the map classification based on TWDTW analysis. The area is in the map unit, in this case $m^2$. $w$ is the proportion of area mapped for each class.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` -assess_table = data.frame( - Class = assess_mean$Group.1, - - MUA = sprintf("%.2f", round(100*assess_mean$UA,2)), - SDUA = sprintf("(%.2f)", round(100*assess_sd$UA,2)), - CIUA = sprintf("[%.2f-%.2f]", round(as.numeric(ic_ua[,1]),2), round(as.numeric(ic_ua[,2]),2)), +As we can see in \autoref{tab:map-error-matrix} only nine samples were misclassified, all of them from the reference class "Soybean-cotton". From these samples six were classified as "Soybean-maize", and three as "Cotton-fallow". As we see in \autoref{tab:map-accuracy} the only class with producer\'s accuracy lower than $100\%$ was "Soybean-cotton", reaching $72\%$ with high uncertainty ($\pm13\%$). The user\'s accuracy for all classes was higer than $95\%$, with maximun uncertainty of $\pm5\%$. To visualise the misclassified samples on top of the maps we can use the plot \code{type="map"} for objects of class \code{twdtwAssessment}, such that `plot(x = maps_assessment, type="map", samples="incorrect")`. The user can also set the argument \code{samples} to see correctly classified samples \code{samples="correct"}, or to see all samples \code{samples="all"}. - MPA = sprintf("%.2f", round(100*assess_mean$PA,2)), - SDPA = sprintf("(%.2f)", round(100*assess_sd$PA,2)), - CIPA = sprintf("[%.2f-%.2f]", round(as.numeric(ic_pa[,1]),2), round(as.numeric(ic_pa[,2]),2)) - ) +The \autoref{fig:plot-map-incorrect-samples} shows that the misclassified samples are all in the map from 2012. The six samples of "Soybean-cotton" classified as "Soybean-maize" are within a big area of "Soybean-maize" and the three samples of "Soybean-cotton" classified as "Cotton-fallow" are near the border between this two classes. This errors might be related to the mixture of different classes in the same pixel. -x_assess = xtable::xtable(assess_table, - format = tab_format, digits = 2, label = "tab:assessment", alig=c("l","c","c","c","c","c","c","c"), - caption="User\'s and Producer\'s Accuracy of the land use classification based on TWDTW analysis. $\\mu$ is the average accuracy, $\\sigma$ the standard deviation, and CI is the confidence interval of 99\\% using 100 resampling-with-replacement.") +```{r plot-map-incorrect-samples, echo = FALSE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Incorrect classified samples.', fig.pos='!ht'} +plot(x = maps_assessment, type="map", samples="incorrect") +``` + +```{r, echo = FALSE, eval = TRUE, results = 'asis'} +twdtwXtable(maps_assessment, table.type="accuracy", show.prop = TRUE, digits = 2, rotate.col = TRUE, caption="\\label{tab:map-accuracy}Accuracy and error matrix in proportion of area of the classified map.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` + +In \autoref{tab:map-adjusted-area} we can see the mapped and the adjusted area. This is the accumulated area over the whole period, i.e. the sum of all maps from 2008 to 2013. As the "Forest" and "Soybean-millet" did not have omission ($100\%$ producer's accuracy) or comission ($100\%$ user's accuracy) erros, we immediately see that their mapped area is equal to their adjusted area (\autoref{tab:map-adjusted-area}). To help the analysis of the other classes we use the plot \code{type="area"} for class \code{twdtwAssessment}, such that `plot(x = maps_assessment, type="area", perc=FALSE)`. \autoref{fig:plot-area-and-uncertainty} shows the accumulated area mapped and adjusted for all classes. In this figure we see that our method overestimated the area of "Soybean-maize", i.e. the mapped area ($110173564\;m^2$) is bigger than the adjusted area ($104927204\;m^2$) plus the confidence interval $4113071\;m^2$. Meanwhile we underestimated the area of "Soybean-cotton", i.e. its mapped area ($18782634\;m^2$) is smaller than the adjusted area ($26260270\;m^2$) plus the confidence interval ($4805205\;m^2$). The mapped area of "Cotton-fallow" ($47600561\;m^2$) is within the confidence interval of the adjusted area ($45369285\pm2484480\;m^2$). To improve the accuracy assessment and area estimations the field samples should be better distributed over time, which would also allow for better land cover changes assessment. -addtorow = list() -addtorow$pos = list(0) -addtorow$command = paste("Class & \\multicolumn{3}{c}{User's Accuracy (UA) \\%} & \\multicolumn{3}{c}{Producer's Accuracy (PA)\\%}\\\\", paste(c("","$\\mu$","$\\sigma$","CI","$\\mu$","$\\sigma$","CI"), collapse="&"), "\\\\", collapse = "") +```{r, echo = FALSE, eval = TRUE, results = 'asis'} +twdtwXtable(maps_assessment, table.type="area", digits = 0, rotate.col = TRUE, caption="\\label{tab:map-adjusted-area}Mapped and adjusted, accumulated over the whole period, i.e. the sum from the sum of the maps from 2008 to 2013. The area is in the map unit, in this case $m^2$.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` -xtable::print.xtable(x_assess, add.to.row=addtorow, include.colnames = FALSE, include.rownames = FALSE, - comment = FALSE, caption.placement = "bottom") +```{r plot-area-and-uncertainty, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.7, fig.align='center', fig.cap='Mapped and adjusted, accumulated over the whole period, i.e. the sum from the sum of the maps from 2008 to 2013. The area is in the map unit, in this case $m^2$.', fig.pos='!ht'} +plot(x = maps_assessment, type="area", perc=FALSE) ``` @@ -533,20 +507,24 @@ xtable::print.xtable(x_assess, add.to.row=addtorow, include.colnames = FALSE, in # Conclusions and Discussion -Nowadays, there are large open archives of Earth Observation data, but few open source methods for analysing them. With this motivation, this paper provides guidance on how to use the Time-Weighed Dynamic Time Warping (TWDTW) method for remote sensing applications. As we have discussed in a companion paper [@Maus:2016], the TWDTW method is well suited for land cover change analysis of satellite image time series. +The overall accuracy of the classification with a 95% confidence level is within 97% and 99%. With same confidence level, user's and producer's accuracy are between 90% and 100% for all classes, except for "Soybean-cotton", which has wide confidence interval for user's accuracy, between 59% and 85%. A small sample size will likely have large confidence intervals [@Foody:2009], therefore, this analysis could be improved by increasing the number of "Soybean-cotton" samples. In addition, our access to field information is limited to a set of samples irregularly distributed over time, which are not enough to assess each mapped period independently from each other. Nevertheless, the results of the accuracy assessment show that the TWDTW has great potential to classify different crop types. -The main goal of \pkg{dtwSat} package is to make TWDTW accessible for researchers. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and evaluating the results. The current version of the \pkg{dtwSat} package provides a pixel-based time series classification method. We envisage that future versions of the package could include local neighborhoods to reduce border effects and improve classification homogeneity. +DTW based approaches have achieved good results for land cover classification [@Petitjean:2012; @Maus:2016], however, a reduced number of points in the time series will negatively impact the accuracy. Remotely sensed images often present noise and poor coverage due to clouds, aerosol load, surface directional effects, and sensor problems. This leads to large amount of gaps in satellite image time series. Therefore, methods that deal with irregular temporal sampling, i.e. irregular sampling intervals, have great potential to fully exploit the available satellite images archive. DTW is known to be one of the most robust methods for irregular time series [@Keogh:2005; @Tormene:2009]. It was successfully applied for satellite time series clustering using FORMOSAT-2 [@Petitjean:2012] and using MODIS [@Maus:2016]. @Petitjean:2012, for example, showed that clustering based on DTW is consistent even when there are several images missing per year because of cloud cover. However, the effect of a reduced number of samples in the time needs to be better evaluated in order to point out the limiting gap size for satellite image time series analysis using DTW based methods. -The \pkg{dtwSat} package provides two in-built functions for linear and logistic time weight. In the current version of the package the parameters of the weight functions are set manually to the same value for all land use/cover classes. Future versions of the package could include methods to search for the best parameters to be set class-by-class using field data. +The DTW approaches will search for the matches of a temporal pattern, therefore the number of points in the time series should represent the phenological cycles of different land cover types. The number of available observations might be a limitation for sensors with lower temporal resolution, such as Landsat. We believe that this limitation could be addressed, for example, by combining TWDTW analysis with pixel-based compositing techniques [@Griffiths:2013; @White:2014]. These approaches have become more popular with the opening of the USGS Landsat archive and could be used to increase the availability of gap-free time series [@Gomez:2016]. -To aim for maximum usage by the scientific community, the \pkg{dtwSat} package described in this paper works with well-known R data classes such as provided by packages \pkg{zoo} and \pkg{raster}. We are planning improvements, so that \pkg{dtwSat} can be combined with array databases, such as SciDB [@Stonebraker:2013]. We believe that combining array databases with image time series analysis software such as presented here is one way forward to scaling the process of information extracting to very large Earth Observation data. +\pkg{dtwSat} provides a dissimilarity measure in the n-dimensional space, allowing multispectral satellite image time series analysis. Our experience using MODIS data sets has shown that n-dimensional analysis, i.e. using RED, NIR, NDVI, EVI, and NDVI, increases the separability among classes when compared to single band analysis, for example using only EVI or NDVI. Further studies on multispectral data using TWDTW analysis will help to optimize the selection of bands. +The main goal of \pkg{dtwSat} package is to make TWDTW accessible for researchers. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and assessing the results. The current version of the \pkg{dtwSat} package provides a pixel-based time series classification method. We envisage that future versions of the package could include local neighborhoods to reduce border effects and improve classification homogeneity. +The \pkg{dtwSat} package provides two in-built functions for linear and logistic time weight. In the current version of the package the parameters of the weight functions are set manually to the same value for all land cover classes. Future versions could include methods to search for the best parameters for each land cover class using field data. +Nowadays, there are large open archives of Earth Observation data, but few open source methods for analysing them. With this motivation, this paper provides guidance on how to use the Time-Weighed Dynamic Time Warping (TWDTW) method for remote sensing applications. As we have discussed in a companion paper [@Maus:2016], the TWDTW method is well suited for land cover change analysis of satellite image time series. -\section*{Acknowledgments} -Victor Maus has been supported by the Institute for Geoinformatics, University of Münster (Germany), and by the Earth System Science Center, National Institute for Space Research (Brazil). Part of the research was developed in the Young Scientists Summer Program at the International Institute for Applied Systems Analysis, Laxenburg (Austria). Gilberto Câmara's term as Brazil Chair at IFGI has been supported by CAPES (grant 23038.007569\/2012-16). Gilberto's work is also supported by FAPESP e-science program (grant 2014-08398-6) and CNPq (grant 312151\/2014-4). +The TWDTW algorithm is computationally intensive and for large areas one should consider parallel processing. The algorithm is pixel time series based, i.e. each time series is processed independently from each other, and therefore, it can be easily parallelized. To aim for maximum usage by the scientific community, the \pkg{dtwSat} package described in this paper works with well-known \proglang{R} data classes such as provided by \pkg{raster}, which offers the option to work with raster data sets stored on disk that are too large to be loaded into memory (RAM) at once [@Hijmans:2015]. We are planning improvements, so that \pkg{dtwSat} can also be combined with array databases, such as SciDB [@Stonebraker:2013]. We believe that combining array databases with image time series analysis software such as presented here is one way forward to scaling the process of information extracting from very large Earth Observation data. - + +\section*{Acknowledgments} +Victor Maus has been supported by the Institute for Geoinformatics, University of Münster (Germany), and by the Earth System Science Center, National Institute for Space Research (Brazil). Gilberto Camara's term as Brazil Chair at IFGI has been supported by CAPES (grant 23038.007569\/2012-16). Gilberto's work is also supported by FAPESP e-science program (grant 2014-08398-6) and CNPq (grant 312151\/2014-4). diff --git a/inst/doc/applying_twdtw.pdf b/inst/doc/applying_twdtw.pdf index 166c6f7115cff0038046275dd41967d4ec1ce502..a7992c1f1a14c43e86d3244cd40ec9621979e595 100644 GIT binary patch delta 297581 zcmYhhV{j&H6Sf)Kw(W_HiEZ1qdB>S}VkdWO+qRvFZQI%B-EVhmtGense|6RMr>l-W z`eBH6q=OEXQb|mL@fQ;Z9Oc~7@EY7NmbCgKNDOc;cJ?$-DhL|DFrK#S#>ks~FIfCx zOqdZf!Bx^Zca7?XyeZ`wyR;<!Tsuoey5f&kYW7EjUTIi^yMnx1=MVu zY&9Ug_MSmmydtG>{K#G9ys0dzXpT<3MTY5QJU;17O&&>AUrw+qr1YKT z>+Rk)u<=*+qICu!Eii@Tn_a+Do3lOiYLesAdIpfLVX?651qyH|fl}gpIBTcfJd{mu z-&(rGeqplY-vE6`XW1lx<@<=*YD#3bW) zgE1TrdMe1mCwi~xUrdGBxu1gMU(nZVWrwX~+%b!!AIJib9moDsuYfzrc(5N8YAbWa zdjSb&U~Hu0Y2}@^pV94=%x(lvl8Wf+FN|j_RAN%r)-g{Z!*DWlka^`Oca7QYL^QZv zEoX{+^X;g`K4Ap1`!sYP?{Ezb4HVO@j(exBCpv%jPfe@n|0~K8Bx`5f1M`T4$d-S| zcHC_~S>6CR8VyEU`i)hd@nnCVgK?ah#GtBY=;Tm$NHxl^bPeIx)_??=(A^F(MjyO2 zWq$geXXki*kz$sQ;+WkpVF^5sZM|NX$jKJIrvb=*reMF@Ki;?w!`OgaEu-*>KLO)1 z!g$X08;~|W^5gTKf|=)1GCQ@95VZ#p1^x-|OVoy+HuPKE0QyjHT*H<Krz6R zMNq4pav-l2b+9Ae9syFQgkKT)AAwf;)=mMBihqQyhGQ;8ZseC%_{KOg;++b_ySR7k z9B@h-eKzUXv~!PLu)Lz|H%i3J2vqU4Lp)e-K&YFdf zpgz4=ihVGE>Ipf6qw&d+SDHne8hkkw9A6-+JK}?~`xsvP_eZ9E3Y9 zUoo=m_Yb&JFD^7PsgkZNz@upH_EIegTL}Q)3cS1%xHBe|%K!XBmK?-eFs5dRK6(+q ztxOeygUf1XCZXK$fD(ywF;W8nc}U%QOcfsjc_;~~;Oera^&WFLYA$~`G1Hi=zNFY_ z;ao8hbHi1GpMHf#Fo~68bVxrk`2qPTw!*;0{P!J$%|w31kKN$A2{x3PV$zBMfkR#4 zH>6{&S^QvNoi&ygw%DarzUk;j+(-(<9VdC1_QRPXrOmT@0^D44Fbq_Ir;>{Vz3D(}% zTza}CuCKXc@~1U{tp@o3YrQ7{UT@1J9@x3I9QLB4J1)*8Y2mh) z;eog3(Nk!;mA+IBh>3phja`;M(nCHrw&IX+)(I)n!Nd4th=y-!cxOaE7WIJ<{atLy zO+wIU_ml2k=^tC%4#E->*IGIJyi`#N+W9jXhltyQpi2d)NdB#PfZ3P_F=ev&GCU$n z*K@AZ@Il63Ze4eP@0javwTrGMI|483wckeWvIuZ3$NyxN;W!BomY|*8Ve@+P)1cyX z^PU=;^L=Uht`geJE!z`wV>owUDO3Ar>V*x66aLdf^n3^*GC;&x zSzK!j$%)kBBaC0&kT?5BA%P{ryUzG4Orf#D%c?aLf2Gem z3olERSy~sc>vxS-h%gmdK+CCrDbh>Egx51mltP&`72g<=KJbTqw|r>}(_Om+E3s;l zqoiHU;)8V{-K8j|Puo$J5!v4+aJK$z(5d?<91T&Bfs)*6ZM@&CKU~I}^zu&@>0hfc|hB?{CN5*Z1`~E=z zH9!d>fNO@npgrrXl?#nTK8NwCjG=Q*lOt*b>vVX%P!aW(L_cS?r>K|UElSnyX3jW` z?%M{)fc{ZQyTpzJd_Es(-;>wXlw7_=xnIZG`gUKsTmQ}ZI^N|r2wv7seeoEa{_(qU zhqf`{VTXDsl$r`y+>)6oG`B=EpyDw`m8s)?qrxThiFlpKeSUXsPz z@YJwHd~($Esv)sGRE*^_%i}W3+rFugh#+>8BsCxJHE?7OFt}U~FkW_+?&oeY_vOZJ zBs)s>^-OAQ?ci_)UBlPoNTsRvrhL)%xgy(dN?Qcy>3)YRL98eptH=nYES6>yI*tLt zq5@k+jo`ai2ao_B&qyQN~i7Qks^ zJZa4PbvRwa-3{)C{$dE$e27&)n5e0b%8FFcM1rUNcg_*jQNca$OBsmX4HAYDtLUr$ z?@1Xl@uRE;J&B(00?kZ*Q}W3kI}``7Ck53G@(^-~yzn{yO}C7CFI*pzRe9hWJwH=1 z2T|G6*MH)xgH?w|xu%sU_QciyY@^Lq=I+P097L(VEmb))O0ZfltWGsesm-IDf_w0r zRXAlJ;VSY|y_fmGcSbg{5J66SR0l4L^Zs2Q>W~O+)ZkqV!zOvtuvq|;hd%*;lC+5C z(wn}3!cC{~uXH5NDV%H3Y^OaFkT6 zeXy)6zV|ytyMmQ+kF1LnWrz1+rpmT{=l&ktY|l1jdZf3ygAiTJk5nDg$Y1?O!vL+t z#{?!`aZX9zy844Ukf33-gk~Ke|Ls1Q`Gs2GmQc#{lP#7pRU#gcM)L(xz@o;rqbN&= z=|)^QC%j^y!={iwdHNBlNVQ-=H7(~HA>|nl2`bR2BxB>CQGKqIAJ4%GMG7oi80gD! z`J104d<(MmLv2#oN{w@pKo99nfoB|}|7ua~-(e97UR8js|^cU=uV4;Ri*=$N@VG5m){rHMDoulT$j$;Ql28yXX#w>wXY|hzAnqJnd);Qs@`j#U|2YEq$6^HbK%T zXaq=jti$!LRfbZ0-fRcO)HTW&;hmZ#$4y;LTr}^C$Uv*CrpAVsto zQ;vyWuHg2`qBfo#uwoSC&X%~jXfPgJ?)SS#SaOVwk?Rsw-Yild2s@r@UR} z32=BHZ7dZ2?1odDg+2QybU7jt9#|0PEg8#vaqy8qx7f`f(a*Z=627NDn{gg=@P za8jGzdKH;&?WykWt~ptcT+mivUdsUOxe`^^g)9#e67&a@YA%rW-TdeMSY7h~2gJgT zCER~GBPHeSNV&}6m}`0ZgUgY4`}5k@+)+*4L%r02h{!R!!`fcgV`Q6O&M|=Y@#=^F z6BW+gUPne<{tOQ?Rw@6i8bH_5DuJ1|cig*MyW`T_mFjGMrDJ%OweiuxIXnn9AvC+& z`;q6UuRnF`{q0t%Yo&tuYV70d7(=_=e}(<3YYaWBWj8?F21_S5yQI+T9mEAsD>Qo7 zRc^L@AU#MJTiKy322wp)appjjYn+a6hn=-v<#=PP#J##DmigWG2)*biLgv$JHC-R5_2 zbdN~3_ouE9?l|wbO-|UK6_`yXvgk!6A15;%AI1|ZuX&XUr3%-ZZhtH*Zr$0TQ^(#R zn-$unKl%qkTQX5$LpGjS!_=wqYr3o?G4rK~sxk~Aa0Tx8Hvl5LmTSOn>7~(dg^r0H zBZq`uKzZi@w>NX9XvVG~1scVXM^nCts_@Pmu1w`{+*bq0g&Ql~8N z3$2o&Tzh-loeCrIEOR9p?v*@?4f%9i3tW}HkEFf z+2Q?QYnaR=yPoE%$o*M2INk7ft0S`xSk#^L7x_`y9DtB?6M;GDCG<1gH%sAQ2v+Iv z$|wy!57$H-E?vaOOy(yG332DHYObegJE!YjbE)*FG?U3vZ8248aurxFUpG;z3m;`7 z*qO9ftm`}Jk}i1b@<6m`J=k~I0!Q~anBTM4VlA4$*76Qr((gIM{ksZ!{kzC&cCD1F zR|&rE1CSZw8<`ms?95LK`M%N|)VD)kEG19oZB_V~K=gq|YM!z_r-E8pw~(v)Ww|L4 zT^GVmkMFbnG?XxyWBF>m>$sp0YgL-v%VE#HFgp85fZ0sfWBR3Nb4t5dTkudm0^Q>$ zTY=!~i{>E(|K*MKinvq5m1pLqOT|$IqCku-25^vevZ-d}^)CO%QS8u}t)vJGy@o@c`hoTREjN&ucOzo8KA;$Y2c%#NeC9u!!BaaQifV_i9kDKF-@1#) zWV{ZlmlF}%*1$nzS;ys@F=J?VBP~oR5$%RRh4$F@2UHNYV&P!R<+%0lh8cU1c{fop zavixCc=HW1L~AT0KOPH43~=o$S+Y*S*J_vg$T`PvxDQLjE@k;}Ugi}K!ve-Y01=L7 zAk-_l=sTp(?-{8lAh|8_`W*_V35xBEVt*cjuY(BvVJ=S#jU@Bp)6sDQ?T=GHf)kjiNP8vA(~?`;|EovjT*Pu#N`bXXt_PU*dzxX(A3fC}qb? znc{Kf)#}CN!xBHj9T{FUfxRQP02kd)RNAeZPEduI+ORwKV-vcKeMZDUtB#{>@}R7J zzF8tbZ89lRdM;l9l%$#=e?iGBI_oQ5Q&@~d_HAAc)FXQzCu&Jm^71ff2jr0HWhHd-T<6wQ04?Y1gg#jHTBYSmD#zWp<;EBsSYz!&WU*%edO`E05L#hw z0^O#7_$7GKD**w^=jLRxC%s5jBwXlfMoZ8MsX9m3ek@}E2^`Np?0%sLp5>9uOG7{q zVCaAbzfiU>9CD#4#S7Pm+3L8`vPB3Y0*L`?s2pA2qA{awXzU0E5GDk@yd0BUDp164 zfBfUe6O_i-36AH}eCt4IlY3QVg)s=zya zDoza=l)DuRp(2$)U+hVgm)xDoeZNQP5^^6iznaJHpJ@`ed^ky5sR&d$MEZ_&Z0V`X zco2!-<_r4##ZWXGKy1;i)6(BUfncCk;D!*5QX+CAVT31dSfUU|`cCHxVOQ(!K(#V5b!9FZA3&EviBu#HqpDzz>pMNLYvbbufIa9HllJ$t`H23XgO%!-Jlj zbQV-|hA!=atG?EEg)Z1Uk3}IZPb%$1+bI`r203U0=*uZUD@f~=>;MJ^LEt|jD zf%v%3P`y0+h765&3LoMK#P@0b*vra}BGr1%Wwu8 zU9nRy`o5)+2Xztg?{Bw|Q%=}GoK8}`wr4+5Wd;blU~I$P7WC3E`n9W%`{YRB)d`6( zFi>+x08bQ2g#}t*!@~sfXam9Pvht|k?}&AGO24VG+da&6(wMT2`3iqmn|`~nd=P)T z;ke8E8L6iha98-13b6Kqp+_E3nQrRJFdQFLxj?p9@#q$HPqB&39J3iHDfN?(l)wB_ zPh9t_U95)7)+j*iS?uEbOL)>)d8`o5zfWXW4tgmw0kbx*-q0xVU>xI@SLcM3eE^M<+mq;D5lp{=q)8%@g z_|;-*jzJ4lR>6mwq9xKw93BMOf9FYoek(tXQZ zbmGkr$mhhmEHqX!9?lX(08~Io8P;a7XHfOKJ4+!72wdB&_0t&pVw?PD zS>?af3mn%o7n+zZ_Cr&GJk;#j!i5UHSnjG+wL2ep#Py{gPBo?A?ex}l^kvrbO!Z1h zBX$-k$5Dr*BI=RxZtXQcG3;)pZEw-c^h+OLyWp6OXWMEf&%Mmz7Dc#1xShJ{0hWbZ zhzy{z*@&&JO*0j-q1F}Sh|V!8Qi4JX!^)jnrD)?BW` zgAWuP#Qss{D%Kc|N6hefflZtLZ6>8jH9UpUbLK5@v!4f*eQ7D}uh!9)#8yT!aQp;N zFtYTquUT-+=YLsV)5%w>>C;`@14hPJ)1RB#YPF+*DjoA$GwEL&r-*u2&RVLUCzp*} za@WcPHF->t=K9nZ?yJdTaA$H*{|pk-I=$i$+*R0@^<95SfB!aUjVIVefpHx#dcss5 z_QO@M&FZM{+o8(SbTW2dr4BiVQ0A_Dd3~qz3I?0)8?kHlWGvrT?(&i31h}~sG!+j< zmORbM64q%qFm`gpTdZ67bCy)0yqEX~AnND3FQ}WxPViqHR#I)`ZW{$U>3jO==Z=>u z^z=L(H;c34SVQ68{kSH?gu%w(Uf41xIs+x_x?80*rC)S7iq+p~#^=dj2YMO{7X%6Q zwm3A zOc**QBc2Wvx$}u%y8_v^UF%{3e*DqiY0o=M#Bw;jpkRrej`z#an5Re$5Vx8t)U0xg zw;&8Se(-9;6?ipUrmd!Hikp4`;gv2+Vx3v6ePM%O3Bgf1ll#g;0K!^K=DioBk-nIT zbFWEyh~TQ=Ba_d4(zU%{VJWLR@8AkarjKwaaSe=_UGEtR%TG+a-|;MFQ%KezR9RhF zeShC&UjIIUyyjQyZx6yQsLrWl{^aDPyZEJAIn&pFSXH(0J1s((Mth6b5Dzoxcuf=` zDg`EHXlLHhUKkNq1Bm_QxFyPO2i>p+-X^_VVy5zq3G~EKHmou}?pPw2HY$9dK1I-@ z5d2NToy(HHPOYjfBj>ymBgfIusRf5K4<3`Wa|P;S6EyM+2(#y+|C=jmG1d(W77Oow z)9oOmzpa6_GFOXU#JPUyjw4L)j-mxGCTB_Cz6!xlg0YKOfNro0y?@F=wHc2v<4SUt zDM)LN?4eUOwYHp+27U{h4-AJ~2%iEJDND~I~2m1xLlQ-OxJ_>xudE!=0LE(s9IIYI8)9$wbazcaAG&Wjf7I}Nwqb=H`+(ndI7I)#G%hs6krn4Y% zX_~jOs5Qmi-zh?HHeRSPOOTv4a4E4P7$pS&#+^bgEY)S}xJEu3ICvUj*L>Kgc7bB= zV*w?R`6U`~X*jfP-uatXsKbF*s4#t+)>~3E)7zWo1s*jan0^fjmTaAmWUB`Gs{HsHJNq0*0lqk*)}+G5hxg9;1i5gyYxq;_Ve%Ej zJE|(a#2qRqIoso2($M~W5Mg}&T0GYtPhm)wh+2WsObt!JH&dM9C&5({ncFE@T(jgK zwmsUOh}tyLkA8Yk_7e|Zzl+!Omq;6+CzcQNwJ%N z$w$u@WAFqSDlc)J!sn@p8Zb7rtkZ!r3Q-`yqvUc)_G1oXP&A(z=C*zGd*+#nF9Wb| zrpOuX*Uvzi2PrUNbJH_madiCe%r zrEw#}UD2rh{4EM2o!RiK5Q}Nrq`6{g<$>J14kz7pAJAM;ZhS%MyvH$Q~z@%ll z$DBeIF3+#qt+#o(ewR}Ng>C|%T~>yc*0`qfPA{WHb3(!U$S%;NCzCcJqw#%P|90I*Rn4c2bi=faDJ$S;` zV4=S6@B+Z!^(OS-At;n|L?Q{s!}fn)t(Xq$>bj!yUcAE@&1s|u=Au*WMblIv|E4xZ&rRwkr39dm6`s`nVYLXR*7=tkzOrOOm@{XheE?cB zXQIAuAv$DYR9tn7YCi&ncBerG!d%7;+Xp`r1=3ZP2-6p>g{SI{0EDht&(90X?-sXP z3!5(<6`q3ev~Qq1TR?<~?*CiQ!OatuLLZ-hkY0VNQ~h-XJ;Ye)DoF5z|4 zJcc4N=8;I6X?7MVJYw(>dFm_0EDDvNRz=bTA#EiuGsvrxfK!!g zwO$)plmQ`b!wN4oVXO@kUQ}a@HV+Wp&=G;?!KM?$(6b`c9fZCldsO6aIoXGZ8vJIp zM!{G5+Y$cWSaQZvYR3qY-+77Y53&lfNLGH5F152}5ndw=020H}B;Z;jUvA=~Ey+_A z>Xj%qdDWSbHD$mYVUW*iDJ0SE*}jX7SYr5-Jv8nD$NNCy9mj^)&3L?Tg#mJkLyy)U z6_|_U&L>5=?(&^~WdwEICZFB(JKk$v&q~L0wk$mzvn!7(AKrT>$~C9O{cq&AvU5D2 zXa%@Cbkwk??Ox;7xt8x6w6o_&qdWK<6(gDhNd>`C_&Wh~Ub!?$4pgRx)-Pv8=TGQ zq0MUe@p6G@DlV?^4fgVpk#yF49XTHLy9gM0Plxyqm;Cr(GSPp(Ift*VKA*JF?shzu z_iQfgeBW=-*Lcw10N`P50v{LFb9O4@EFBPRk?F~anJ+5vrTopej5oQMos;a@u|KKW zbMzkv9k1Sr!tc85g^(=fuNe60$EK_BK0DICEv$2CZg~q+gotMY8XUcU+B<4%wWlpE z%sDzcYF3$7@S(OZ!3_i*up8fe!W4dxOg@BcG#a05nq5CH0Ws)ipBt#Fb30_*D*IY; zdF)$!nV-A;X7A_c4Ook5b3xpnZkpRo;Ge zK_BbjJ_-Y6srkeWm9P#NSbsL%ATbjbBg&47m_#%=Ng-9m!s$Z#!rDUH!s|ll0>*Vc za40wtPKJ&KP!25mKk5Gq?ZCs}M0gn{n*S5vAocxQf#%nk>~L)p!2H^F#=aY>Jn5da zSu;4 zA)D_%{{Hb-p}+GU{H!Z`MIxCEr09uqf25IrL$$|E!JNBKu;Nd=WEZ)CiPXo8?8qIT zHN4$|8V}mooxPHP^P7v(&{|LLy^{3)6;xr8yI!VdV7Xj+7;Maf{Q~GQV?CW=&MT9j zPm=Jn8_SMTN^4IBt$EQWaM-VDSxIl`!kn}|a47#GUyp5I4X-w;iexzqVX21-=*Emv zy|uM}R^_QMRGkxb`}A@na4YawOD91z2!l3Bs!~g~J@Arpi#0-juiH7LWWCDz_qndd zNC+1oH><&_-Ft*~YXBHQbw^=JK)PVW7LDN961hAGM)+~t4j|)o7V906Ra~6@CrrKH zwRCOdW+EZiUToSuYU~7G>8&c^6?HoR z9=5w|r8_zv$Bw7g4A98w6sf>#jx-=Hs@(6@!5R)ZzwDTt=>TZymmhxRV47~AVs)5m z#1A6c0w9}pB`^{HiD`s$o(hki6gjVYarM@j76QrU=Z_tulKeB`aOzdKX9QgslCGNI ztJ`IHDuW|gg6MX}&@IiC#8WP)fo;D^jN~mx9_~cbWXvl)DQw|N!iyDP_m_o@@MQP6 z6=mk(so=W9<`6J7!_%dA9}cfVD+@gsG$t03?n;guuSR;J`{&rb+t098h$T zZ}gcqc7m8Uj{YSBG#=vb>9aa_ z^9>8JYY$QYu#FR|jyCiT=wH^@vRF)H#_PJAY};$x$U z))Ut6uuRW#F06@!GKyn?eb6sV0B_AUSF{f-r}tDQlNbo`*21{Rx*rYapR>Q)ABgx~ zX+>U6%sG-zF3#(yS-+S8<)qO~dnU1TBd#b+E8MuP8V6&EXVkkNW1o7hmM7$zr@Op~ z0Vmz0Nx^Dvm@X)+xc^BIJQf9c>HK#CR_4FI2LneLXZIlfU?tYUC?!~mh<`DbdDa!3 z2C(45<%Jad2%*5iBQ5c(#N+hNxBno+qbKzK|E)I6|D(0J()xJ8Fu+*3QvWBlt95ka zH`*`*@(lMRadGGi$}XO1?Kz|E>({P5=0LXWL5EkoV%Ab^|4CEz_MA*aU$$c~2FI93 z@lzo8KIFjXOs_QBJr7P3-#Bb{IbLLpFL2sUUtR}TOz=-P+GwP?udlm!v}S&#e)_X< zjo(86!Vmn6?Q=gNX1BaUdt284MoGlGsJ`}lKcjDCwMj&|aqQu~pRhzf9L$XjXC7Oq zqmxRB?#Xu~NmO3G+=0n-XRW-Fxqis;m5#TiY3iTuOb@TPqxxJwCtGwkv@xiZT3ravmuGsv37Q%q09jWd_nuk4H`o66~kl1d-M=Uc`BTcXH~tKu#Z z7h6x!$PX(En^_#6-s!qY*|!&Pohjp;l3v@MYiUbk)t*^|4xzn&AufY-xb2>8bB!WY z6p+h$5sLKxVexa+h1pqH#Ac}5XX$Wvb`mnYj0zx38uQn{QWL#>tTCT#M#T;ixo2UI zV{1^lYbTJ^Z#bjVtVSmREY(3izgQe)=KBZv9m{ol_Z@C_-<;zkE-`kY57G!CG-HCm zIZaPEdfx=-*~r#eqBe#=327+> z;ZB`~Ah#APyy!T|Bp5zrC;vGMERGri+A_Y5uTJHt<&{9&MwOiDXn|WT>Q&fa_;Ok+ zK=pNBJ96x?hkb^0S@qHV)7@O%M(O)0wRsBuDX-y((YOoCiBc6v|%z&kYffJq}-8PPx~O9L_itqu0Cfg|dG zHq@A%R#!>RHJ@d@S9;q4a}}J3VJ&rI2Yfyvcnd;?HmiwFR|D=U)2W3lwU|+}1Hs24 zHJ@i}t~grN5;3|Ed;Y?gGlEVeN5Jj($|gWE9SYvcDg__FVd?b*r}hcS#d)hiug@p& zvwr8$2jDL|szVf~@fne_>C@8?3~JXk*q6BMI}>&z*sq~cHPssA{t5hgV8xvU#eAxy zo>GFuwN=5vl;)e7Qj50N7WstaCF00pf9djJtE;a&^Yv=-1bNkN^0v8 zE&pnmdD~RQmUspA%RXKfnL4-IyJ!rZ@&xAo#(CE#?r!u0R zWbZ3oEgyf_{#4P=rHD+VtjJ?BPvbfTFp~LpgLX>Wi6BvD4mH{T;@I0r3!dhEgC?XA z<6^e2VqEO{a(sT;L?R)V!pE2ORR?DLqqd?mt3&|*QUUwQ1b=eyoVzvH4@m$uOnLx3 zL|V3{S;5984U)id_O^2~mNYlir~)mv!4Gh$Rk+}zVFiddNSX;F2wjDGYCcyY zo#q-;j@<}*xAZpUcI!CGlje`0XY&ATVj5Kz0>xqf79ABW*YduIsjXb!bgBbv1(u-p)qg6&$Pohhh7OTokwSQ!LqTIc4Z0x&;hTB)D|3 zwm)R({T)uhA_y)IRu#8Cu~+KcTUtQ#@fnu-TcjOtF}eLP=Z|nTBomhvO|S89h{qBt zJ2C{B!B48pjIi+a>qKNsJe3qcs(B#eA!0NaKk}!U%LWm9fzFD?X>+28QLpr;HEnOr zwiGoTg5`BzxPC}_A^%_-UX?E$&lFg|+E{MCp(QCmi>@f0%-h_j1cCryaKVRfb0;;I zV!}zYY61u_9HdwE><_Y_cbudVaMTo?P@s)c5m49l3GJu;^r;+<2|^Y3W@Zo=Kz z#T9Jf> z^cO)PtqocUM10bMOeh2X0-~v|FI2cWt4Nin)4BfC>fKL940Qizt&gBIaz#5t?H=Zb zkG&68Z@p_MeTCW7-kfu~TP5oM^rUa>P}A{$0rhYQ;_!m~Q#G9}3|Wt` zn=DN@!G;y!o40-)o1Dv<*Tn@F6r!2skLhkh8=s?pH__N|8%qPKXMzwO^kN%b4#V}! zA7yRVY%nWO5?e$pRP+eA(|XQQ$iIRwOhOlx_S91{s}KX}mYH%(d$`}%A29L)C`$p8 z2R_iIaDeeY{}9E;ah`H4bO`|jO@<(DuqFO)H_^gb7r~za#Rli=u&N@*^Uw$WPF+Qx z8aIpsQ^+S8R0x13jpLg0ppN{cbHQKouV7en3DF|FtxQcKjNG!G*pYo`)HEj&h(a`k zC1c9IWznAP?h%)*6f9!lBrh!BLP(ah^ua^jX2l9aYzCtcX;5CcST_3{3Zt>J7fn#5 z=y7R@gM}Kqo#YW~vF^(-oRcKs3m_9d@hM@qE!E7}S5N>+ePo?lgsRz=v2l&4V);L2 zbCFnPMSX|MOtKG}7O_xCS?!y-gdsZ7OAH<&0@PJe@E|o>+3C7xV=KnheD#v+LsN1S z3uH28Hv5yotnn*<|Hg%UR*p?Jm5KSIt{k!CljyRw9D| zk)c54aqt1DyjwKCuWRAve^;MWOII^Q#Z}tIT_#TaLB#7}TSnw1s%|Tr^d^J4Zbj$d z#z_q#ij~$SZ*D8yN1j3^6zxFk9wELHPyKH+lS%Sa@^;PQ-@Q(Cj!?7dX!U~G9zOs( zEakO68joqoGHV9NEv|0#HIJ8z9t*a^j z(a($Iq~ENef4K$i2Oxa83@hivyi>5>G$1x7pI(d6m)u@E4R>n@W>oW>zrOx%iXbHK zRGS1)z3VdT0qY+8{buwP@3IU=sI$K^IAvX%iHJ#@oih{TV=?}*=bL8-?2p(hsip?? zyIW}O`7f|$MP4oMsfVzoYON(M78Evi8d25+E=zIx@MOolL=NdxM3#=nV&sgN*O{Et zRWCV_eIGLPZjAdsi{njj?MRv<_og7Q%YOnSk?1Gts;3QN1}UaynD*;*+aY#7K3;DJ z{QVz?#WTIdF>X=xb{w~)L61??_Maj$Y(J2bKp*rg6iUb$xnzj1N*xv!aGJMlMZoq&>hn`*~{jfe0Gp?#D zlW=SQ7b|xFEmi)}#+xhqvp?=Dz-I?$&~&n8bz(|mc5pqexx?1$@`Ef@LC|(9QN_3Cz zlxFR%etkVpRPN-w^K|3Z+HJ-tGmTUm{ru%L#YTM6s8WAOV)t|ZVm9Op+z!Tvp7Vr0 z_2ug|uU(%`X*v`AeA0N?GvXv3al#&TnldYNZHqSL{$Nb~wK@FDuIXPk?m;KNJPT|| zI+dGuCdi@|m;5QKB8@IDKyKAz+{_aJAH7v?=1xDG8bgsU7rld9zHZ#k@zZ)RD;7D24Js4}sCdx{DWY1+TI5lvc?J zW>hD{+LTQEJu-+en8k_}JkklW@|JwiZucq>f~Ec+;&(`MUC+U|Q(s@fjS0o*_+>Xr z^cKTDibB5(`F=JP7RRSSRj}wTMXE&gO4jVe4U+ku3{S#LS0iM&lhU6|j}{~EHmKM9 z+cIZA11}s$fH|vuFn9yDZlDEu_n0@3K8~jbo3JQ^wj5 zEh3a#0QoZzV>5uIhYL3Xia3b1Ah?Lsu=I-zHg`I$6@2uiSY6dBA3Qb)Sl*w{(B+IU zMD$QnLznDW!eR39oP<7}WaLd$U>UfP`N4g^bXd;cxDD+wrvWm6kQVHwF%U6pu#fka z4_pDBz$=wJ)(WVaJFPcAymV#$7o=a8>ezu4K*qEyb)YXq17|zuvvD*h+q@t@Hih;2 zl@n`Z%#C=qIx8v$;s1wP+t3S}Kd8m{5>XuHpX7}nWN(R`$?TmR$=ol`=S2wUjN-N& zLT9Crx>o1WesX7HjTTolwtNWMLzemI|jej6fg=Q@}` z0Q;ViJ4Hwyx0YCFo)d{iBRUdR`M4Z6%5>0;0s+G(l_Iy+E27rkQ2#RyuXEikKG&WR zBaYQTP`a`rHwwowEVOMqvD)Xf<-t7}Q^gfpG9*7bpK-dKYItv$CF2lBe=XwD=VnbP(z(ZKjW%>Pn(DZ2m zHzE}lIy!nz1Qoqlz}Z=aW;G~8Jo_()vvVrcWN*+YD!MX3wl$uRi*w`x0TdFFOLPzb z9lBUq*x5OV{t+GePsJQMbVL{*B7wnqr_c7b+0ha>G&X+?5l`!~_d{=ApK@slud{R6 z-DyyWe0oJCI5d#2X7zhEd>Q&bEg~MCXV?RdAaIO~yzkW20xM}mU6xsATOL5k)~gKq z77uBsuB(jlG!Jagb*)L=<_%s5ML-7pSu;gE9ei&yi(Z)Gh{Yv5A;qsszV(;CVMuJW#>XTLL*-wY9)`yxdllna z<&0syUwrW|vaDfzw~CR~zaWG`sN9}^H5s#lp=nL*kET|v!%h;N?`SgsL*Q#x1p`a} zt#k}k%gj}*%kp{`2l*sT%bun5b7M>Ox%5U2ONVu=?#@|JEDudCZ-+7-Cbz`0+VS8{ zMgl|2k zW{7GLYI2D?$VO-e02poMH;gP(w$g+DndC%cC8&uKQu`&Un6XPN%@}J{ZBT1PvE^5L zdHuyV`PH7;oo-=Ww^rtCN%$=$x^`c}bhW)rO3N7Uo#gtbvfDK&{peeGVCvFtdGNU% zvqs{DR@#>O>O8c@i*7z6QXvV&A1bfgRLN-{0(C+J*zePD&65^0~_=5 z$2eUNe$%ft-GxU(-0pG`VQhm^lGkN;Op6eARnf8Wp6IJ4hZix`1JB{+#%XNG)bkT2 zjl`*8y@~MzGR=TBd-;}tf?#{{uXn#Ic6Ir-Kda1mx&t7CN=uA21g)Go)e))ty${_NUxwy${_a?PqX9Bn; z%7x*T`di%$xc0A_k0hc!vU%ITR2;d+oWf*S28aoe%SA;2%bVGzC+uz~@Z7Hs^5SIk zm-6o|jFPqbEzJJwqF%Oh65#%}0n60OG8#WZY;6HmZPHhrQ(?GXCtlTm=AK&G((l&2 zeR(!0AMIs-9v$qsrSVA3bPSwOxy0sbM=VJ0;s`n+mG};_@-ZPD`AXmc7DYc3Z67CO z4-2WP>5e+a{B;a8r?pdFTU*Prov5B%TvU0D6^Qok1P|ga_gP%8q$2t-ziIK%A z`ZWOiW_Pv5`b-r=yAeM#m?W~(V>Sk2ROeN@IG8`^DCO9DvTW=9sAIA{Q0Xn~eGV2z z9{oPYoxFyYpF3{2%u9cZER|+r0``HWz@x3L@+*~%u9xSx!gRo=A3Tw!Q^3b<*b(Vx z_siQO^~3$gkK2KZPZhb>`TsQ@5&kbPjEB9&6dxi24wsFW>%VrS035S|rK7c*4G9}B zFW0|uC4e+N*@QJItnQ~8ltT$ntSLR+Y+lMpbgw{#L>F_{$RiPX`zET@Ldxd5vm=M5 zL^Xcc*h{Jy&6-WEE&cH+d%6BW;JG2oU(D3|Aj%m+6I*LT-V>wq1)lFwjN*Ig*6rXtoOcc zfb?JqPlZ%L`56n(xWiLuh7oJsg|Qa@g(4Cjm?7d`-f5%+n4ocFZbm5q`1(hs4r`L~ zBBb_n;i^udEN!jeq{OTA*}pYa<8Mt7d6I3_3hg-UAH}Y5&``Wfa^>w^Dn1-)dv({N zX#m`LZjN7vIs3uiDmKm=bB9I%fG3t$UCpiOkTRwI`#@hu$(Tn-D^8tprA)$69QZvE zXnn70>B6n84|uq_zNP8-mf1KvDGgHu2#rR39ZgN+y2Wuup&9qmaH6Grezg$CEGtm; zpy;2P91p%YFQ5Ul<4#)}@w9h8`vsF)h5-sykboSk!)aAesTAA6p5vVE9GM<#BGG{d z>qmSCFZ+=0#hSg;J--5f&g^H#v7hDI0=F*W_=V^wnL=0oy~ZYxcsv4eEu`2|@D4XS zwV&eYa}O3X1;%8=_2fhyh~Gu<##b|6rhgk&=x`4E7&%|64(@TORZ)+C#~QT{fB|y- z_pIO@?1=e9*?7ppbY#C&ai90>qI+34D;yhPv*HtR~7Mr98m{ME>h=DM94R}gR{u-iO%qjz}&{{Q|vboQ$A9d z;U@kF@hrpAEM&;xKp_b66Q4qlM^6CN`R;%$k>Dhg@Aet`W{gnxPLMLs{H5Acqr zbq0t*^@7GX7Zyaxa~J}fGPkW4wc687<;FK)z>mVm#c3fllbsph;_yVR74d9$mH97$3Bd_6aj4E2DJ% zROZu_viTv2;t3$E=}5?$cmXErG;DrMo@9Av@=PnXo|K}{@4KmAqD%&O>$Zv98>rK? z_V`mE99iyFtJe)_2dY3`)NE~EA$dI8>>twTBfkL6ow^$P&^F8&EQI2*B$19}=V*^RqQ z%1}{_911jMm733lhyDh;!NQUnuLF*lE^xw2sOp8iU0rV}2QTTBzFeJZG3VyMbrDGe zgY3SiTh#M;X04KUNph--Hau?XCFHjvt7L*k*d}qT7a%>=)#TbOK{DJlI5!5tLyAOs zX_Wd=5DySp7D-$}c3+i0Ri{GbXy#1P+RRsgq@?h?WW~QSElOD&ef#<682r^z;6iKO z5x!J(jTJ>@jb@}?Ti_rBVhpk5RnjL?h6EC&jyIL_jnXMSQNYv>ih+Iz6DfZD>UA4Z zXa8yk*wt4v_=o){9d*K55o#&5kT9hhart!t!EqFZTHV_uf<}9`@ z@b6z))hL_VPFGjquq#n9L}od5V+qe^iEqq166!7EGnL3q?rQSbf`#AXCsK;?OI#(g zqmTiAUWUN3*s=NL*wl*IvL`fxm6eO|*p|LkF7@#b5sF~8BY23T#zg49>4b2RHlsbG=CjCa%dK1PQdqTzdP(1t~xBe$WeCtBX3 zAv8p(ia~*G%UDLodkFTBYJgR+-Of<>M9KtUnoyc8pYbJ_C-EkvxiBRD`4@t>&?&8C zu6&gT_)`S}Q&1>id9mYy(?)i}lFNOm1{Je$|5xeAS+k8ETN^e%9HgMPk>cv@QT+~Y zZu`_2beh0Oa&e>E{c|afZYurOH@axGmALYG2SPmHnr?P2V|K+y#JG{sZqUlQt&bAm zv)kKml(MV5$%TC*n33d~!4tVhXwku@tsp(`SloaTyk_h2{`|+lK*BZi!{W{&r|1x- ziZSPW03DGri4|Ghs7*E)Cz{3vh-6Ig3_rI(HQBSa=)evvyWu zUihWMFb*~Ska#|$rxRk7u`W<{e_rS%qKJp7Vs5zk?w{2sjGl$qohu=eq$ODZr3wOH zgG%u4PLpl2w;RT{`|fmFhD=yL%OlgOP+)b-h+-JqB1wAzu%A|w%G_V&e+SK~nb z8Y^s~Om|=5x?L^8e${tVjPxu#xvcK;^yKD5+0Bs$)&dG)e#hN1r*}T}UO%)DwKj2D z{LSJEZFU4Tv4$&HfFx63}|N53?ug|SQ+*E&FYm{(2@&HI`Vn#S~r~Pn{R_m z=_Jdk+7M+`Ttk}Ce%-#3Ao@gi?;!QsD2lKqwdVfN_}`D!4*F_)Fo0sL{1Im1iTkZ- z9_B2l;>COeKW4EyWse%q0XNhqR(%8gTpp|0=<>V3Bua-pH5XXQh#dvtmNr|+Ut1}k z7^n*tZyy|5?L6dBmIcdEIdZ4?E_a&MR?asEyu#lpQLkE=l8!+DCdBl$9lrlKdep)WH$1LBq(S@UbO5*P ziaW%ao{eaa5yuRJ^br%zpr7{t@5XJ!iJ#JYJIMs;0(XczpM@0Bzca_$P*`Ttk zJ*F|y@q^f7L%{ieLBRiq0e~zQ5Yj*-IWXuHH(h2>)|4V$66i16{|gYP0pnO9h=A>K zV3exdB0q)SlPY^?*9_(Uj$Ez zu^tWz%fGK-wqj&u{E}s5X^7CHfd*v*_VIzCgR!!vh**iFK$;POvZWxJW2f+&86}}& zfU&aw%fB)s0AuC&>Tm~R;rNdP=f6EG=J;T&T>pw7T4;f?{2M2fGGHo}(r$s5LS}}S za%rKH@~Q1q{5L zb_sE!b`){uSnmZe-w5}?&0h>Cq3|N0bdscykVA-07k-Nj_A5$4imKBcRS8BzIW7&G zLx4|2)OEx$RgieH33CaDcI$uc?doJv<>v@y83W}p>Mh5R#v^37(UFDyHmga^Sh-h+ zhOy(#&>0wq2bO}##oOuV+HX5lvkRH4_@>k^%3%a)C;kmkq#pVasJ)j)5JLScR6ZA9 z&+AetI+Qxbi020CSVnh;aHlH`9VZ*#$crdP14W2em<-btl}Mfvb+nc&4F>8oLQ2xP zRrwf2U{OrP1*VbLG7Dhs~Q_#%eMR&Uo%nEe!g`ub>F$)P`|(4KtGY3t)u9OoH$&a zGA+0uDt}sic#UYsxJyFs4CFJFFy^!OBP*sFm}@uIi?;_8=8Abf*;H^?P=AwE2wo|I z5%fX7@2uGjmr7Q6R-2>)eE}rx;9b9n)4D790FeZsQs}2(OQb)GGkmLdFjxF7%GM10 zsRnRH>H77(v2V|LaiC9D$!xxIN2PI|e0n_^aHM#HO@LGE`W{pe=^e7&1Qr4Urhj7E zi{n*G-2!DUC~JAS&65}Hs+W2`TrSP?8~@~e*vTFtIb=)3nM~Pf`d$=jq9BhV1+ya6 z?YBX~p}3|ZPW0(qSW^4uMszpc^dTO8$9{oBM47V z07LT;%N`e(n3;4&L5KMXD`}v-Obs^|7){-l`wofb^L|!6P{;&eDwEHodkPmB?XKq`Y%bNW40c zXTfXRoqE~RK2~`&97Q|z-eHtpwb2|fSm1&ZJ$+_s5@#gMxnRcFigDXmL{c|WydLktJg6lkAD`k-;EocfYVUg69z3@%qjKc@NuuD=M zIJ6slbFbqqMX5S=>uJ_2BZi-f0V2&-biHlzd?#XdP&w-bKfMDzAdV%q3&>dBDcJvY zIRDHg5D|0pr@M~C9kQn{xLutsiWmfL=1dEb5JQQulf|7rJx`3)%bj!;5l_h=oOYPR zrggCak10lG!Czy(v%J422L|4DR!t*wKaoO!81>o< zai+oe4EnQPnThOn2r(DN50G&tzB5OIx7y?ce-`Q?sd4H7Yfz3tY3E8d&{X9llM?>`(**z5&*%JMQ=;*+ zMna5StxknU5}E2{D$~Pjp9i!k&J2}K(j6!o`h|=Sl|;}G+dt#H>wuK}ZIvDtI_C%x z?*3Xqx`A&ey?d?Q2ZDIxvJ>yE zr*-dkzZ=1xyvad$Jl0I}S_1MJj;EhP7f&Hfw1S_!^Wu!Qf^JDHz5DTDpGT~5=zX=d z_d>v!Y3e9}Ymh}|zb99~)^j-7Yp@<#Gg|OTSLL0eWiy2< zm?SFXM<$3er0489NmK8cYDj~h1R02(s#r(1J9+vOy?&0|w))9#9@Yt*Udn9Dxk;}P z8hYMu)TJEpdk4@uy<2s-CpEY~-WH=IJbIl%SaOWx0!J^m6#?aAO;1P@R?+-dLw9__ z!i*!|Z#?F*nX_`V@=}IcOJA7m(91mA>*%^0zw6f7A21s4$M~ear{~=W2J0W*6+VQX zOR~OeaUAdvTlP#3FBN6yG_G0b#Gt(J`nGjUC@VWRElmQeOtl2l!T!8AARl~vxQ!eg zuhK?FAQ=X!-T@4i6Vu}>W=zR5A1$R>W_knz9>1+p;I4%8fGe^&0Q!2rPMc_ZgWr2Y zD>sd;qgk$ewDk>s2HFkhrA*fd?fZMIv3ZJe2&~LUnV}cdLoQn8=k^FAnM!vLLkmOa z6gAk^3x0DJOR}8w-Bx?q2U9b5E;qlJi(|Hf7mv(aS^;q?5w~;Xx02$>o#qnqA+62f zyM4&9#1Ju}lZpad->C=Sl~tu6uxHlC)@y&?m6s|hy&Anpdxf%bndGBDmb1X%m^0R+ z^r}2{(UI!kh4VSnIIR!(#LmkjpVVH11&enRfyE%HLY;+yf8X_wx8S&?AO+92iQ#sD zq>qNZ0tcKmBtvEHS{2Qo*tTol%ZTN#LXe{N69(`ag|$dE#iuM5JY(sQ6Ub%bLv=rBR^ z8^y^F`UbQ1RF8p;9E=q*?ehpbofLP=affuJIRZW|SQ){95*WU|)m2ozkTnHuz}9oh z>9fq`)2ljzXqqE&gYPQ|P6;0tjP<=h>+LqtckGJ(B|T z>oLQ1*G*{%cdr0CJg;m19H^?cAiU<6PV51C>~C#(!-pn(HqUbpImLEpd~-ahv+n2d zAERGdnPy%#5Zv8|5n~_3d2N+|kP`eA-LKZMD}3Ki|59rMf{uq@83uR(d+!H%O;-NX z_#@gQ7pTvl6x&;;AN4>7?j-`Po*95#rvs?_Gv>x8X-KEf&BxQ$V>RBskG#z%p?qkE z{w~)>bGy1{!@$fl{5QsmG|uW{-s*dTGR=n??5Vqu@EB2{2~FQ+f4c4ZJ%UfBfeWo? zTubaX!t{ALlnJ({bjkC6J9#bVC*V-t%|mmg7eM!7e5$$94WL{0qgZfsR})Rmyu~u? zbz=8GsfF_F^A?-B6L{kbiVnuc^*<&%JI|NNjtA`Z1xIL@_%06mPrS;%);|Z!|8>Z_ zLw|Mf{-@&~Sila=Llnao8;Ym#RH=S4|EMTg*EVf~L6s0m-!X-(>{V2{``HE!fG@0jW9BEDh% zvV%DX@(;tnW*dEdT{Z^Zvb(dV4l5qS}*a0>Y0jZJ|GVhC%s1 z0jNYSzwth*nXrm8pRVs^zIhF|TtDZ**DQ8+ToJi+G5qQxvivgqq5a;kK{^)i0tjbV z?yh!RUQj>IAfo|Bqr3xA9j}+WTR_0)F7(CSx0Uv)tlxcACt^fLood11%mJmELf$*s zxr~)HVup8bl;@oz145mmCQs_pz3Tb{fPmU1g)fO>hw9XSy8tGdD!> z(d*Xj$r2K+?wlUwOm4x(39aw+a*6>)$R*xBTy*^7VjydM{XTLo>Ga8mL0DsR0r}Q( zaB1AJnP-gC03CRa*W52p8Y1xR0Iv=(B7OYgkjD~%L?&0P(g% z);iev%Fh>NqocMNK8}aqwog#L1_hJzd(_`6r~_5a!Sd6A{anBO_BZ+!6`nM@o1m@-b|?p{8?u>` zJ+VvaJV$Pdlp+i_s+JKxaD3zU2i5q434+Es$!R-esL`P%F)9&|t71oVPl6)NMKaN`#$KUB z%V8L%(o;EGj@~W~-YKW6LUgDw1e#C)dox(`qNm#K>Dz78PPKda<%I*EM`OsZWRazL zoMUoG#sir0lHFaj?ig?tzfu!cok(ZSUkoht~X^rxoE>-N`4i=j^Cm^{6``wPw#giR0O5xTPE z!A>x-J&XHmCbZ z`X`h9m(Frm~ttwNP$qFzMsnIAZpi`M~ z;JgJElAuF*(xz@|MOHTR_la%`gt5*2*R0Cy>dTV(f6n9d7)(fLgVXNHp0TQ@kUd(+~XY~s!o(6Bt+zPEG6uqj0G;$h|Qchz@9%O zs+jkqUh0(Q&IUkjCJtJxf59uyO9`jZ|@Jgkm^FUVjH==emM1#qu?A@R{^mXo(edO6T z{l;+*%`tCsUyp4ZJ-5)v?QfLxCdIZ((w}Z$C2Xfb25DHRw?%+~k(!)F3{w6ktj9|L zM1U1pobe1NiI#7z#RaLW@DI?j%ahbFniXKgf$H3#Bn{XEg)3#Za07y(?sqO+gA>RS z#RdwlTg9Ihs}|Zn6Q4dTlz1syf#|UE)H}F`e!N&ryU)O2FM3xh_aK=G%#uY(j9k{4 znE5?{h-xxFdQs1GwS;K?gX)>4oM~-9bu*=7I|HdVy@|A|6WX7gQamq6tg$}_(}nG! zJfuJ4kkg{!s`;xySLrg^4U+6xL_Nh;`!wuybnI0^_lLg+WPw#L1Zh`9 z`;Q#_xIm}GhH{U(@@l4V3T-g)qIs48(3;SU)OFkpVWDS|uKbG8OrMJubzY7EFhNQi zV5Kee2jPZa)0ld6fCkNzC#n7#&G0ziMc*qgdUTO;-9NdM_J{k!lH{r*)FS+Y$-%+U z`)GV&ng!%HcjbtZ=u26FY4D`3CauW^T0M?6C@O_Lnq_{tMIPSTZh%3PYpTU}a@bX6 zYx9I(@Wx7=JQuww?3TyitDl{T044q9I`;#nJo@^jZPm6_;E$TYFpWBj)_vwWuBe#~ z>68Z>d9fqv+ktr9_y{ywO{uHd~(#o%6)tzi~ zDUAJmg|Y=1*$&L$&r3C+66K6_?9hI*^vid00hfK6MoMLnD* zfmB2FMHDCR{Ix_wU!v%~2&+*j75*mqp`sv~JX02CSx=zB9lbA0_JXP*Ld1r~9xK8SZ&rKHDp*K$VB4}zfZ zxvXf^q2US~NZj2cfmW!)ABrn&K!+WKUmY;B!*&tbc?+32I-@X!u!v^=HcQoyTXktGsqhw0s2lGsqyzLY^`?(G4}-^q zw3(AKV3-XJ*PC6#z=(oj7+x-p%K)+UyfCJKo%rY6!z5oN?L_cXU6h^8bu?u6`lkTX zd-41x@HGI)>--K6$dv?B;4Ua9Ev56IvE3>zhGsRR18tijI5baJ$25d7R%9TuS%Y)< zQ^w{eP;e?btN;s8I2N?-@8A!EJrdP>u=%Q#r&mwzw^567PQ_NX9jWD@(8OLH?}K&y z7;X5Rbw0AawNUox%p$>9?l5~;O?8>Sg-~rbPGYwvfsqe@k z6&1RCL~g;t!vO_>5B|NV5)H zh`M}pMAS7e73Jl-7{!?&entW|uQj!;TfV=hjmSn1yc=Bw&LdpaG{Eg^kW=0!_M53V z%X#E(@6?)n5|VbJ7SI72PPxl!^Gb@Q9MdI2^_ZUXmbXmEsn+8|TvMoyo)b%>os8bZ zTp2SJz?dP}r=~q`*wDVV`xCM%{AIY{RjE3Z(r&deZxzdf%b{wYgxHR(^Mj~qs$l&YH73Jl)e()!fQ@D(f zUy^iYub#>2HB)Vu4rKz$j^UJg5y~=IGBf4QrzAn>kT3+Vyb-@( zu2dCB?jgkXBr9!DUU?fl#MMMq$8;hbI;W`g|(_hvDZ$;i(bt8i~|cmyyYRMeF^wmGPd9`IYS=UR#j&J+?h4q8_AaELt(J zWD<5x^GuMJ6kf9p${$5UzA6P7jBSDs7Qic6HUnLr64ZnGif?=^x;9@oyF7LI-da*F zEAaO^m0QD-$ndRcur{7yBm(%|@G$TuqXEC%5#;<1w~Vc}RTUFI2+m5T z>r4on*HgbC>s3nLAgjQC6D(G`_aXU4F6_xSN^1@KczEu|wj9o-A-ZJk#(FT9H?L{# zU)RjQQ0ogvoRpdJ={&PXY54~4n7hdP_s8=itER}C$xhc|x$(4{7|EVu7T9{fP+wkH zk1^U7aZOIibuE{Oq5kEeSY16mxKXD3z*W|NLb@S18nv5>)eY< z{ozx4L(wEb?vh{l7PRXe=i)6B9rC-PDQ7j^)Vo3U^DmNhINC|VgHaZJbT2Ap@<8U6 zpEHjEYgWSG@oPnoa?Qu1Tvif>Hzgf!)@DLBL0a)7y`|Jb$K2XvJRtJ}CZwE=h%#&{ zCrtzbdLz!V89+>*DJZtb`F1L(C~r**{H7f)yA?~VS7F~CdUFQ#@uL%$KZo3Rc%GyCLnlgn7Q_Gz-m#H!kML;JjbUcj9H#Y`0336HvLt_#zxZ!hcy{S zM0LMk>*opH^0@#szw$UYUr+GZs_*CGiv5B2314zE4Zs2nO+B@JvP!P(1L{ui4YkvH zB?(IeU6SwBB`R>jUi4fg<$znz@$AA2jHNmE)kl)R=}d6O!6R0YLY>s$QHf44xA*Ax zSuc1%jD^6E&v(J}urbH^f(9-W2f@C8SZaMInUH~JFn3L6K`2EIE%M1J|80jqXSot$ zl~Q%Np@0cR(Zoe-w|v(F1(vqF!112#BSzVN<-V(}_351>#$qd3ut;z@4yz%prJW4) z)s(Q&FfG$5g0^r+ZG`MKxx?SAXEnpq+iVwqZTZnOndE;cbz8{1m)Doxj+eT5z?XAU z`CdWhT!b+4T$^j~G+UQyVfd*FzOpZ-k=E z1p`J_&XIcb`|Ug6PzzTGp!B?cKS!z(q`uW*F&>zEItaSh%wc5~9Z!Py8xAr5|6yE|qVx;u7h zGOskDQb(-+%wk08Lq4YC^S%0$!`du>#R-6!vDWn>bvMER)?Z693ZQ9%R*N0Gu?Eh< zv5LTC_n@o!w&!@-Q}-78;)Lfw8foKf9i<5IR7aTllLUIMazHcfn4=;92TV1kPFZ>z z=I6rszi|%?_lW3}<=DI5uWk!j@A)ew-4y8?jvW4sn?X?^xtaa7kG0Ii=9cWC%nUfA zVp!j`^1>btN%QN^0gt;BbFOf!Zd4rFATDitB?i+?DByy6lPXDt?woNglK+97s7Y~` zy9t@c#Z#l8DVAjobppbAeN{<*@27Iz_p4mYx$n{T$;Xx7EfH7OzjKq42)@rQl3H!@ zCk5)qfmP#=1hKoLs~Rb%q$%x;93cP;8KP`)@;;i>--H>F9_PUBt#dOerUROpBo)!& z6TjFOHMCQ3*mtHfOIcdIkzrz280GDO4EEurk~(!9SCUZk(H+I{lnvKeho^|L97c?6 zQ3jXXOUrXVNd8M(@;XVQ_76X{5%fTU-~7N1TEOi|iM!h>4b5Dca3Pd)x+{Pymn*3i zfcaB)$F;=PW4X@>cRVXE_n6;w@T-I+p*NwMd~}KkUbD6#>KI}|lsRO|-`lc_-eJ^s zGSquU<~w*V=K{aSg#Jc`GsYUNJG(yvwSfEf?S%mr2{^;uj65$oDshF=zeQMT)mU@& zo--r1Gq+nPKzyIheCZnu8cKjf8aVp$^=j&BA8IMvsQOzszALleR&8aXubeuXAN_O( zK2gKvh!=TgXqvhDO(ooi=7;l?19lUlETxXy5Q|QBvRl|pfQR=wW;=2_+y|}!^U1C9 z)CRSc?4e%HZLoYH=CV+s?JKSH8-1?Bbj)FVM|tiT`5Z!wXE^l*jr>ZL~Xi*R$E~_b@++ zJM&IEaLq1EG>(!nO|l+JAtg{=iW~%mF=F9J_<&yyVs_}4?Di{KI3R^6l zv%O}C$|>zCZh@0I<)RKW>91KUKKNnh@8Ba{;lm?KO^ldCS*3KtmZB^LA??}l(uS2X zVPYt;t_K$ftYSVrg(8BDzJ5NPM^+FRK?owewSrgCD&Vg3CjQha`9 zO~q%kF7l<%RmlRdwH;sO#@zZgJl9{8WWvaQAUbg(0Ni7jj-4 zi|B{Vyp}#c6LNgsfzxt4@A`?$HLfI2l+G2j}WjoB7^w5bQ8U%}_H$C`m&>P@8nSi-?V&R&vQU z4kBzagOds#IAV|6on;p2a%FWfo7W(Dj9CHR&Q(ziInJ7C1tAp^|>jCw{1G z0-7@f%h+enG;^;Y?ICMa3r9L`wdMG;(E~Ww49cdctBU9W8JM@B;U^4$lF;MZUF>mX&5cRr!bn`1UIa_Lyd{vI+ybk;JNj zgW2d_aDL3Afx2VVQtml^yQGyg>Zxh9j=ovV+NOK$HjEUf$lXfIn4=#rJ3d71w75Ep zeU*`bjj4^^N#{Q_e}BBc?p^S#5YyTwj_(h%vJ6~MJ)^?v&uX=9gyB6n=Xu^vn{F4$ za4V7FU3@lLp6bFN%R|JRMQXi5bg+G3y#{cgw$W_ZfWgKGZ@_=&AUuWI>?W!18Pu~8 z5Br_&xt{zK=<~oY zn_X-6ul0t*%kvdb_iys||5^%Qd3>J{1z2>^ch2sB5+xb>vN`r+7f;neQe04e3 zjmmbf{2}CrhfP%aVq_8BxO;ZZ)umZq41=!YusAGOZOlf4FOm%|0~oGBd_YG0G*dl= zikbzRdfz>hiWWTM9!Xu)Xe?s!-8R*mKd07?Jv8xyhXAWo#Fc6G7WefFp;E>*+Sk>L z-w@3}(vI!_*gtqOG$hBZPbk<;7zlSmcBKs!G(w(n*P$ zEH}}L|G7*td0c)w>%t^zj*hLh5ha{efytJQ;} z|7oHp_Kc12dzZ+R9wnB7V=I9=NfG83GmLH9j?Zj3#%X>(e+%|RJstmA@%1xy+>bU% zWJEGPPjvT#?m9i;fxAaSzqivvE`4pQ(m!vTO>mPVJHD$h32ekS3qv;%jmbBF2w~HZ z3$p8hiV0QcpCo;Yq59F*O0p`l1pyH^QD*e{anLxX$2d|5@8KkoUCG3Je>2e}bScP6 zpSKYh(+FC_ct@39_P~tqG@^CH0@=cY^{oXAecli?X?rg4x?L6?a?ZXp{Hb@|L(J5K z;km5D9#oV`(y~r?h9}*-_`0wZFqLcv!QoxHK3RV1ePvcqUFHc*!as)f(|nZu9jbYG z!Ty9-xEBlO7omQvLTL07YO{?DA6Rn+1ANKeVz^`*DQKB&`fo?*-NdWX7#SB09ZpBc ziQp~7-pufXoacQ^nzUxB?4`}x>g<3Wd!wlw(@b*#V_uA6yc*IO4to1MfOVGqURiqh zXzXj|x4YkkGtB|eRQUjsBR+OAtyR!FHZmy%iPYc_2@b7eio5w!Z)myusm_HcnJQ1O zXbWXLbX)p{YTQQ1%qcS`arkN#v#;=jOlqcNypuoRBFZ?E*SXTB>I zHey9_h%?)~ieR_cjX0qg1^m_+1vgxbQTQc=gR({)4^+Y|*zQdVRwgm6dfuD|L6goo zA;0A8G%U88d0&dk0JvqtS^g|gMLdOd?`ZoR4+=xeLp|OO0av(m z?f4TId1)qIb(S*TbVY;+6&%h@oLME16gsh0bORD-60F@k2T&T3RZA4Oej5tJ`mj6P z%lFaxaozNKs$U5Mxg-^QZA^Thzut97*L0X$(oD!By^)T=Zs`7SEe5jUOtg+Lsy`i;LGC+u&el43%_&gRivXQrmhGwusgFtAc(gHzb0* z749#zSW>g5v=hC1XY&PUo_E{!xzW$w{_l`8PS@E_q`CJuy8s#d@FJ@oVxF=3p^mUq zFWGkXT$&Ik!-M#a9io^0e!$Y%z(*9(s;1MWrnPH1TCL`a#GKJq2T^air^Zsnas`m3 zW~cF6_2j3>;jI2A!0$BQ?%z)`DEB|n)V9D4BXA*5)?^6`Ad4|L-{&oBZ1hs}F2V_me6Skm+Kiy$;rD9!P2dwDHfZ;d?N`Td~bS7S~l8BR&C-0GX z|6M#_u8-$Pj;{>D;^Qh=*|5;OsmNQ39pRQ12;lv2zk7OmdtLYeczZb88{y`@K_Dgg zc$!M6aef(F@MRMAzvce;Cj9aJZsSRsC!jrQg!}#q_yebO0B~~=_a+@6Cu6NBEw>YU z;pOk^=K#**Dm1|(#Usz}I$`Y~9lMvE*71Jl)g+BGojJb)@D3&J{5(ofB+2Hyf(>n< z%Ad)-Q@CsG7;DHY$bY##q-7deoaTLWZ#0b1n?B^XS#!rZncQ%nfb z$Jzn=ZOxr#hoC#8*)W2OC*K3wp#!NQXUy|!YMd7WTSXcD0HLZn-*m;M6Gbg5=_$5M zuC6gxzO?}GtZHycrx|GD^sPk)R+m{#*lrp5Us~-Q04*ZV*n>&?e!gkZL*$3ax61VC zLC`G@sxsaf1H|>VsRs2UzM0{j#kXxl0E=DK48Y4 zE>#oyFc9A}HmR8DYR?9qjZn8F>9=97*#W~0J;y7}A8LHIYGhq@ z3>3p^^x~q88HPV&^HkEcT{fMXe-}PSpUFFN)>6PD*3@(J z0Pr<&I4P^-2YvMPRwnFeML2QU4vm5Mi8cxgL~vuSl-0REOm+NZ)C%X;oIL(2%3vMi2@ib$N);5j8T> zSp!@V#M>hEjp2&ECq5&tGMPCw0|oMa1Bh#w?W6+&iq9uLil1Rtu}0Y4=Pq}hJkYoD z#%kw>ea82hQY}>Xov2EKE+c66)~R35l*cK?a<3hqG!C z7k9ZNA5*)83~|^6JNPHIc1bH{H_kEzC{WUprFyy0m9g4May0tvpyXehE`$BP=VdEG zwOJ+AK4<8*y>`q_kOPmnyKS{nt5hb|GM3)FLl$nn2Z**ARMDhMx{D>knLj-z-jc;D zk$a`uJzFJPA##8FjI&Wz2pNZq0gt&NOJ#zJmy@~~oJ;6KfVZuBiNWGt?H0`iDDZ3{ znOR@`s+Yc6os3Ex*zeg}&sIr>Oo?+A6n9i7yGeWvceW6Xf7ZK&Sn0HOif{&cyHY## zGXvqQ5rdfq^bBpM&b^YXsl>pCA0BIoBT6*pq10lgR(5`g8KiRlnWPi`46WKwuirD` zk4!)3U)2k!%H_hjg<9Ehu?vVwRGHHmfxgRGy z7X*|`z3qd%nW6}ki>6OJG5l#!sSEsRBd4Z={B$xn+Ub08g>*@=IQOqCV&EI~FVEOX z>ynv0_iN$fj8`zu7ToDjUICA(>G9#Eep1 zv_sSi0^(zqEWFWzd|3%ue))Rxk$-h4&pXd>ryBa+a7t%_Hw+8Ye(Wx)7v*rrUT>n$ z16Bgp8Un-+>>{1S9M@ib1whzs4U>gFHur{WyjWVeZAL7z@qLKyNvNKupOoqRSN(Y1tu|*@rrfvhpP?YxQE5?_+tTm(ff0zbY*g8sBgum?}l{Gd6EY8!ePL?F{Td&Jsu@}d8#Swpp1pD(`IaTeTQyj zGY>t|KhlFL*9$w=HP87#?m8$D7IYF=&?!Z-i0QwDHyD%fzft?=jpg z46o8-X7sdik}6v^6wnd=M4ABuHxr5#SD$HBPbn}zP1VB5xVSh#uUDEi{uw+4iW#R? z`YOsqDJ>M4H4%#>dK5KaB$j&IlZb<23vN_|;4oJE=V%sl8tSU?<3{GlWqM$h_qVh{ zl{b7$Nk0v>K4BFaBXpuT=yyTzL@}m#rhsXJ!60?)YA5-SA<*etO|mCr3Wb-HB=-k< z(%4nOENxA?9xpGAjY~5;7y3|Cm`Fbmnk&?R{Kwb`3{zomnkAb7ac; zJBBl}T9;ZKGDVwCmpnyV{+2QoGRNSUkWsL_s-2x0y>TAXH9EPhvMgj~cCICDhFUn7 zUEJWIw=+&gQ=!P5C3BM(4)K0;akKJ{0W^cM*EA!t2KwEdBB7*DO|6inO-+G5`cIL5 zg~c;x*&M>nuNoiijFF)VE|ilU6#Zl!=BRr(;FrA|(={)t>}QB23))6mlh>6pSCiqW z_Wb>Cf?rnmkFL2=E219Hs8UJ-v<*zj66?4J$3UII)b5tg195J9Q;@6B1B75k@LvFD zjbj;YI&el-vAI7yxMd6Of>vH=;E$T!Rx|K-CFo4%n;{ClogqA)U)bE5Tl5* zBo}2Y+Nn`BHapctVQ);I#KKrCiVqwFVUX*bb2YcwZaV7}1x~0A3zkeUnmNA0F@2ei zv!tZclwpTy-ydqbz8l2#ik5c{ddYU#cX09iu``m_RZIOesgm`&VfoaV23E*MdyRrn`C<8^StdVsF+ zZ!Y~ikhrYH(_GfF#2w2mZA0gYqZ%w9q72w;C)5lv4Ypd1w+=@(s?27G$+w&~i^-5j zAU4;x;CY2U+2Y(?;pC{oBhb^3kTW=|vu}uX7R_~A->J%d%>XGi( zeWPG7So&qsH0sXX1bGL)p4$RRZjFFgkB74~^lLj2I(hnwt{Uj>R5b`AHQlNCD!i;- zLv;<#_`GG0q{%uO$XI1aR%@@#Z=#c@m{*Cj z18)*V15mf__^q@AnkFK_T@2YuaptOU{S!~bjr^?o4O?lc_yqlXb)XB2n{y%$&a`8B zQhzH-qs~Z?4}0-|?!rhnxd!6kv%gRpcWkW-I<xAWFdIfq znLB+b$Z_RSeP}no!ZU???5Gwk^A?Q34z)mSQTSp}QDMSr*1|Hd9i1n4r1C`J zng>w?v7%p4PN@e&uNgsnrNTPu^mA_BE~%;@3H9+`F^S))}osLj^2s+S8It_?4hQa2FR{|u;~ z>THi1db|n??)kcOnt~#@;D^R#ib{*qRSC0OWXU<0GXw*;UEDebd-T(jpt}Jzjh*yE z$+cF}2FpPahsiu{*6KDaAZvA+qszXgC6B>fdew#wMUWZRz4FBJv?_=3*BEGZHmRy)5*V@7{uGFg0D~exA(>Us~EHGjrSKUzK_7^-qFyH zTroM{kBb&PJs#fd568t1kuDY99)8nK?-?=#VhfR@JE%UZ%EJgKpwHX;;1~Wp>xH|s z*MpH`V*LAY4B2vCNj@)k_PCG-d!2E~=|Lkj!lBOMXx0W!(!;>#CfU=L)7LciT+mhj zmoIbSi$kKHVyxX0nOTn^&w*3%6qf8hoPA#eUR+!sUZ7b zzH)~$yt_#;Qvaf#xqs8t;N;Dz$fUYm3R{Vl-4Qs`HnwYX2H(;Bbffx$PFX3NFNSlO0%k`Q+bX%6Ggfy+p?UGQ-WaY)&}3hq>|*ex0ACI&^NYeKxMEVv9x`^ydEQ< z*qbZoG^V2c?99EKuj;hAX1?sRw!EVYq2#rp+_y>xQ-MJGd4H~R&UjQ>*WwnFm9F8& z=FYc|>oK8wttLW&HN@M}jNBgx^p4H+xPL$F-j-ki-s^L!ElZfI*6CrGXc?c8w(a{* z4CbJjccdKvxn~BYRT(`#0UyD0H{)(*M~KzqWaS@kt72y_^2jy;>KC^uh^z1;HGEoE zm}6cDxIq0p;cENEK(xEaVLTbdjq=Hq;dwd^nNMD|n)D-|;mhiDQakzA>|)mL1!kp0 z&dG(pm$F&!dq%FL{c&l4SN+izQ7wSjT^w# zj-bJUvSY;?yXX9R_`Z`~5)~%jqqFYz=lx8>tXkzNx97v&gHU%q`lo55>CGn5c;i!8 zSSYf`a`x0_$NbAlcg7c>yZJ|8ivJe?6^x$cf3Zw}|IFnwT!4F~YAG+`rig?X6r7#^H<-px-4RxZkza?x zzTl(FCD}SI6kAe%j1z;a<=-N5FS-c(Z){uRE?51>oyc|CE(}lK% zBC&qGjUy5izAR>NrEhn>!hSZ{eQ|ofZDKSNAUb%#fBA9_fZ#O)?GWw>J`!Ue)0^3c zw)P)yEWOH5R~)y-S43B&)(y|SXx9C<{r-M-`P4#(_!Kw3=wn&m*=J zPuFC7?c3&aMjA-jo}BfyAi2c`{YEW z=K%>S(rA|9*azx?>8p;D3`Kt@7io*H?9sGAvG>?*s;xQo1UD0`FxOqguRXW@ZI<&O zGecbTakv`8nHFE_4@+>yc=KQSi?8H=Z9Rf9=s@YKOyjW5+ntU%81}C~(X%i8-Jd79 zEK_Wx;?=(0;!HJ~MxPYTlXRFyp7a-vDEs|=q&p~)Kb=;9USdpFSbozo6Ow?Wfu02c zKraEj{>|%L&9buD+(bVBm30)uq#n8 z7Ep;gV?$PvHL%QU6*buVA!p|Tuqw?PA<3XsAhD`i>4&l*8PuDogABk0gFzuHzDqPF zDia~-CO9X8Y1OM}?R&q;OxC?)Mgp+ijkS375OEiN$t8^TY+RI>%sfd>Zpx4Z?FEb= zno+I3?0-N@4eW;rdde6gN&sv0m=3=hW$0y@bel-& z#uHlP*e}z31*})If?NF|`mS#b_)f5gqN|Cs!;*)8gsRTU-t7pirm#dAcduP0%4rot zKMkc=P7>a8GVilxrUp5g_pj=L1S^Fvl7ZDb1ErUtu(M1E;JHoHR4D@-yu_OvESu(7!Wn$BXu znBGSFxOe9#&)YjHL~W-Qy8WpM;wpu{g&vAJCp&Z{k21gtalU6wNTl|W9vRnac%#c0 zCqs9+m#E*kKDAm1EYz=RO+b1(PZ?;ud)jj(K~b3+fsss+fuWkC3|ApX8qyS6Zg62> zsqR&ZyM#{~N;KD*v2N{Ivtv`+OI@>akEeg5%*7w}A}@gJ|D$ZywEqu&SPWEHBKu?s z61jZto%x+yER$v4X4tJv0KA>+nj*@PJe(ACVZZq^nni~&p{ODiXq!VPqXuvzz&h{d zmm1Q#;_nAkLOx?p-dV%2R6?Y>L_ieOW8v{Gckn*wytqLT^)tBYIJA>i0R;96Ex8#3 z*apEIR>uG!hV1=ZE9mIRs;Ri6Ygc-q}Gi`I2(BWH#)019=-nIuaAH|D6Uv#kNm)_dW z&L46$IVdoVhYb~GB}R4z90(FZrSKx0@N9rl`(KGbJiPnN?jgX1xK?A-JgMyTuE`wUr zHU9hYUTYU>euB6f1pc0vX>=ZHtp~=Fx=6@(@Qb*o#th4i-iu+yp=LCMQ?lAHw_xl~ z$P1h|CDKT5ZzjBhJs;;D6?Td6!=6v?7w#34RotK#e+}oi;Wzuv{~CYC8o@;qv;PJk zgl6*wAK51Q=~s!DYbS8n=$k}8?H}`Z%!}RCQ9bynbgE+CXKeU|?2n?#R{#4d-!9K} zdUV3r+b>U94@BF?T@SrFqyu-l;n~4A>Vw#KfGbByQ_{0Sxw@dYh(Z{_vwrmeVn7)Z zpZK3cB?Bt!9YwgneSm= z7)tk8HPJv#eC1DPdJhw(BcXB(FpktglXSx1TqP83P=c^pP~K>BL{rCA{2()z!kjVa zQDzmHDP|GUX}^B^%v7vvialMB;f4=;KPd=N(46)cjBx7_glh_lRM4DM8F+hrQ0uxOAipzfIgWHN#XBdP4V!c+FFz}q2%F`t5SB1A8SF^8 zG+w5g1WQAU=9gH)NIfzlhKf>+m8cs|{9KwCOGQv}Eu1hyN`Lc9%z7fX)uEHJC>jo} zg1Xa!2sjZn*80pVd4XjgdsCS881ZAC7&?(_k;f=J^W2GSFb2YPjos1iTrd9SHTNbS z?aCmb;0A<`66zohOD>KItDYkovq;db-jBw?PX!jIYmiV>*9aS`(pv>Hbj1U(%-++Q zxXOYcp;|Wtx~J-mTmJ(*3Xp8=jr-gYI+!=I1*~-NvxqC4X9uw*R%b@wpt{6`>WMn0 za}fVs$b#@EVF;8hZEp`GAowGW9h~Er<@aw3w8-Iyf;+W?--fShniQ_Qt}HB|+^Pjz zK7tjtFL9BYp(oy@8EL0ll?KeHwlWDsf0qvD^Lf|i=X+}h*S(mL3&~F4=)C^pUz**!6k8FPawQ1XvRTV#_23{UfmjTska=dH2T+krb7fx2(G4 z`$*rjF?!~{A7zEc&yf5n1%4MEnjh*)uA@*IsJJ&G)>Jo#fE<+W(R!pkH5-$?Wup5! zXHC2tST(Z5XAvhD)NWkV?rRFAaJ?Wl!S~-UGGThgNk48P&MgJKXWu2g6~MsSvVoZ3 z4v9AoTvZb+^Y;Z&m{L6m*Jb0%j4}zxSl1dXh{!wYe5sJni8Xa#t-c8;Jq|jg09;j? zD<-g3Z$X8g5@Qok?|Kl`o+5ENKO;_QNOnpmvHLXUUDO?V3H&h79F;i^058if!%iu5 z#5i+;QpEU0Ffv~k-EU2CXBKkmpFnZvmq1RA8(vDtIo}J_(eUGKJY8LUc`yBwiU*B4u>vs2iQSPpVL5fJDTdyfVmubZP|Of7`Nob ztV@s0kH41!7w@q6usHY{4o~W>I-RdC%X;6`TEmWaiCs@xC_B8*OONf3zrah-^Z)cy zANr%j`FoAcZu{#PqSK?Jl{0_U>BZyJb@fHdtcE^@G`-_nEl_p-_%Fro)A_a2)5)1? zL?ZGVZL`u@xK z3=dQ9Yri1=bIa_8b+*$8_`Y4u1N!ME5I?N|Yx-3;=94-1SYW4exxQfelvAF~AdNie z>8_cHqH4Xqzn4h4IQ|KAh(8V5^H%P#Z~C)%DTP_BT({Fl7m*V{*^7M!&8N)=bTjUP=|g-vZN4?r zXT}$+f;_3NLeVQ*r1bk-E4R{n-~(DMHr1TwbT^*9e6pi}_eug|#Ob4VaVeu6x3NDU zc00X+PA)i8(?9#S-jm>=j@J@To;RryzTAHCg@v~)ckB3=rgM7C<92`b`4`@1Sy0|2 zM(QO>K=Yt#yK8^|5o(_gf8%+9rR`8Hh+S>_PQ$*vDouM2=D^@Tz8!SjI?SMNEwnO^ zLCC#7q#SKxGcC+$MRd5E!a+oIV3Iv=+AzM8 zY75t!^8AW4-|{Hfsmy@p9-~cnx4V-PQ`+!$XTExm{z75{sxhxl(&5S=A*8b4<4$t% zCFHm@I4@vQrAJvp=&`4tAN3b_91s5sZEedRwe`jRCXn8_rNF_c$~wivpY8R>baw;$ zs@+ZDDh0&0B&%Cf0YsZy?Ae^k)#Q2;SutvSye(|p1Gh(e{ckbTJm9$DLQHgBe`m5D z5s+`VjncT_*>c*QuDbhypdr-AI{$rvY+11bgu{R_GyaET^_}2mWck06pe7uh1fr%J zA7F5QLW?6tK}=d)Yd*O+J4|`Jq&_raPk-gb183z$)UYXsuaD+tRn5&xS4V97?S(tB zON(aJrz?wwO!=r9o8HEz_q*s4ss^+g2S=M%syc_onJT-E+$BZ=LCNFeafh9Lyw#mX zy=9OMF-WgOgcb+Sh9!p&&_DEQXCBX7qfJuFz#0d`$B~OkRPJht9@r4{Wy-gq$)xjo zRWYo!2qw<)0Ek#mTk?EWbN7bLsS>NbnLlFY|7FkUE5^S0iM3Vf@i{i{q;%88=Ub3C zL6IS$1Z#6{Tx_+XE`;VA9bwiqH;>1?)qlJ#J%{Ut8;XB2+wF4;T<9$sM(q0hL5Avv zv3j@de(Dd(%V^(yPb5c=#Fe1AFoY}s4D~0vddMcbk}GAu?n)#2R3RXnm+!fSh95^1 zd*ragl8Z|FySu8jr4=I4Ys3|DL4ZYvfYr1>$m&hm7EBqpb% z$N0{p5UPc%4H(5)80w|5KSTf;WWUCcJt;kJ^d4(Hdn4um>1ymJl@-lb6Zs+HJ_3*N z?F=HsF-RJgh#})>bV1}+0~{fKuG`By&Y6=@wyuOlr9n&PM`_ud&cNZ%&wu1k< zW?NBpu8da!F(at*^`lBJK#pLQ&}cu1l&IwCV%Ju53zpoItX!a1?+jj)NTNb(;jHao zUBy0{Gk3$K6jU7TSyX5uBNqCQf46N_KEnf2bHVqBZbD_js-#T~jYAT;h@_24G_e>N z=<|#5(o4uf=4$%!5AI9Kt5@K9`4r?HDzGgk-#zxtqdb?V|GP4$a;-VR!6x7aS4N}qj40)j3oS>*qxai8 zHJf$Gwqb~NU%h^+?PCAAuWTM7zorT1WD2bON9Lnx-R%ZLI)wbrXBQ} z`^Ft8aIt~>ZvrnVRf)^KI*i4;KMr|)M@(l)T9Hjql{55qedS})UUL4&d^3dM!GMWy zRQEZp-oqny{t|S-{V0Nelql!pglQAuwl%&F5CPlg8fbGn@k&M6-ju&26KyRmvW-rQ z@zPoHxp?UFZzv$c#z~p(Oy4c+^jGO8RuSDcq`slZzo?>|Tj4p_P(C3xu#0c4m!E7) zlk~cjp~Waq*xN~mQ_ZXj^p*IcphN zAc}=-g?y3uF&>20#fdEB=vz7qb>D5L>%4 zO^ICH$Aef)Gfgl)Zl6VzoE|1%wVB;mG%glX7x+JOoc>tizD<+yT_9*9g0kXIfK8H_O>&Z*X+A8IayhCUU1z|#> z1w3md&Mc{R@VY7}gi>> zQ5&e%48{Zz|NQtT2kehfM|>lhtW~99Pv*Nd&;s}qe!beAg5?(&IK=M_+boU(f^LiQ zn-c|?AJ2aViN|XDQ@EAflR>Wcf+}600gHnhTGx8EKAjDkC}KHJBVaYe%idPW6V`(N zzSg5~>p*e#Yjh)Nnt7?)fP9QCf=EU*1ZUhv7q(Iwr4g)^5U57Z;l~tG@*&oXpMYtJ zG>oY-3gpf}q3RyOmfUgKA8-8-qsNKPI={gLEE6fg|5@V;!orV#+QBiq5G7)X_hB7^ z8^*Ss2AhC@RAC5IJ&s7YYKc9S?5t=jzij@DlLfiji?oppGaHNg0G_|xa@e@|m01KS zF}(Q#(Up`<(%-Qir(28%VbF_UnXgL^jf>&P$z^gsna8|ZrxhWrbE10#{JMwW_@{6e z(H>yp;Ev=}NQQeB`W9SL8?g(DLWJm{;o2;P=y#xc;g2NzhaKl&mQ|Iv{jxenrT>UJ zK?p?mTlh1qRxY&h`SYfH(&3_jRa2DM&jR=jp)tFAtx(T zJQCa9(QWs`J1xkx@urUul_+wd(nHwluejEjS-aaPc_ei79B#2R2*!acYkxThX`^KYhkW zZ;hvcL#R;zGcn$IAAH$RJ4FcngH+8k!XGD7+}OU_Mr|YjTJuw$e>uOGcb=VCG;-_Wnv1i=_Q@qK*m`&bST6A_$|Q3W}}R` zo1I~;8TfKp3WP1hZ%fWqXvCStES`ZJ`{%zz;rEj|R2#}Lgi{REcexo}rquEbJ*aX` z@i-MMsZ!v_eBrWDt-e}4NRen6zMr6q)7XXK8>afEX(ekxWpdd}<}P-rvYr|U%}R4# ztM}@f6L?84x#qa+q0WyV(lCq(1ug5*k~-N)Ho;&;n!p}y9Vy$sqDK24gZVC!F$3*T z+-^_07MGpP?D%t5aMyn`Lq>R^f1xj9H6fZ#0Cc}B{Cd&lp2N_73jzu9>f5||s=IEf` zvD~QJYj2kt-xjLrPjjhC=kG7K(ZjB!Tr8x|3Qaj+{MrWPn1Ye*A!kp_yJVWV%}Ca; zJ3IIwoc|M2#0E{e&j4>9i#@+4=H6hQ5XgpVA`qTlM1t3w24$VF7sQCD%?8`O*U=#- zSN@Ma`d$LaFCC&+e7cToU!g_f+^e6@PsyUjzpce$bbcNCu!=%}UFpKZf_Kqp4AN^V z&^xE!u&SsM{=l+xf7zQo_smwbt(VNGv`-i&+rl0=yhZ<8qHU}auZ`Yj=It*#=Nd

QLFy7bn3UuIxXf z2VB|D2TSHR2ciVvp(XIOCzXFw2P^(9BU+3aDPL!bO12e8oAMNKL2bW&BlXXl@%DA- zx6G-N1)S)Gw^&Kcsm;#w-TsYLrg1r8f&h`d23^}_7umb@Q88v1&?hwl*DG+r2}HBH zyV`f3*^cf|q;OWeRQvMh63pgV$E}BO3E3QS-%Vq^rmSc&v<+e1T{Qp_l7z(zb_tG` z$^_XZTM*^4Ps8@{pQ#z?+A%CfnZucLgwqDiWt7=>qpvGT^LKaB4Nz*mfF_NLEMN{% z!_HfpWaDHx|B_y~*=#B+EvaV~LjCIEGx&V+%@jjDu6>^Y(OnnKVc>w_*+0@5V z^7RkuV{Af+CJChxBN-j;cb#lZfBr%G6?Fyt@V`qaMg3>e;?gt9B!c8Nnjy%bc5I5| zOVk~X>?J3;ZS@Wg+SOhmLb(NNM17H@8yTstpn?i$R#rE8XI zUHr(PWOWm`#zWifOy`lZWDku*f1ebGvv{q}q(Wj0HW_=|b2AY?gVEI_Y8RF>!>d|K z{2Trf|3w4`9~1!8UtsiI@S`_9UOab@zmTB00@>Vg=$~V zCBx&HhTao`xQ#DmV(ijfn?I#0ugs{PjvE3JqWu=cLaK{j;HskWV$#zY;9E>f)HOY( z7s22^kJ7O~E`S~JFyYA*qnh#7c+~M3dpcJa=@Ubx*2~16XR%VS*qsmW^^CY&-fn-q zRoTxHL8>NmX5fP{={$JI1l(DUEAv7C3mRP8WU%HRw8Dij9tB*@;bXM_5ak-U)g5m} z`IQV&VFXl{Dt{Kz9B@RdMz6A!#5Lm@5FOxe?n`$?T6k84!!-Uof;=^J8oF_`Mq{pX z**&1h7(Gy&P`c!a{UddjMcy_tUc1--*gFIPMOs9MeS^+O!2JVF5h%%>?Gk~(m_tj~ zV(m*XY0g*4P&)PleeL!LPl`ZO2GK5q*tOS3kh4eg$nC80>atTb9A)F-UzbS2(vgdL z(i^>k*03T(P0VEmr4L0>n3C9nKQh?N4BrTqX<={f{aP0_b)T^qC2U%Woc=J_`JWh_Gn&ZC;8l-wY_H+wcfm7*Z#Z$Z!1CgHtja=v$*aDmTp;+v81>a%Dp zVcQ3h$KitW7HY2sU-?sJeW_5N&k~CmUZCE4VQfES2T;2a`ud8()kBwJ!;NdJEdsVS zntsAFD0zR8TCW~!+_9-YapJ52{GfY)L!O7}`R6ji*pA#!m{0HPtm!n{Oi+T}q}cwl z)qhfadpn#|){B|1Eyd!&2i^S=|2ox;rwxJ3GZ`~c3l|hEki3-N&gVo}j*yW8AWwFq z)VYhn2X{xE8V(hjd5X7@8jHvf>NT2Qc)PmZb*%W9IZdF-u~6@|RXp|E^Gb=gd3LEb z0vPV=+u`OU5Y%Qf?8kp+HhwZQ{Qr?BMh3S3V{KPyd?(ghk-rn`f+GoV3vhhG3hlXw z%N_Yr4~R<@1fXGC)N1>0}T>yI=4b8COXeJ=~sB z7B|mYilh?Jil|lU*NDd~^GveY&x9?k>ov))=7GG_%9>BtF5EQ9LnrrSKh=&Ay1s%? zNr664LWOJn$G&w7v^M(A`8?TsTt7d(42q;HnmKoUw+{Gj9f}t1!MaKQ11$OpNNJbz#&@f3ZL-|DMb;W(n%%tGxWqJ>{d& z6p?@b6VMnN#JB%_yUMDSZOiDN!wtN7t+C1W26nzdy~9obM&mW;h_;MFDfCGS zkMM#et*b?BZ;cnho?9x*FR+r9sFSfqY`Zn6os8xQVzji+JW0zPIOM!=Tg1D(e-|MK z;vsm+)d?K-AItY+#t|UuCTLbb*XQO7hI7v=S$oy9)gJB73W<_aencTL`97bUe*pjV zeoqo-M@Ua6Uf=g*y5aLeZ~QtN21IjmzB@fPNW&hQK2y2wiVcNyf5GLDcoTbrdoRGc zS55TPx(@wRsB$m^%NRiov}r&Y$s)5N2!jL76(hJgHpq*}EySrOzW^__7(2^$*%Wgm zPZrWmAcQQ=L`G5U4=MoBF*9jjLk5~5j%@7Z79o4mZkMD?Esp&tdT>^Iu&^f!^P+!@ z-he~97iL=W65JnNn6(kTTNI zyr3v99(`UTbn@@rChSYiOQy`Z_TvDls=hvYVDFHOle%^)3!>#vS4%DK3PptDhaf4c zHd=lou&MomRtX;S1;DQAn*)Di)4)i@-$EY_@lrR$*e7`r&I?i>F-5YNtzAknmIj@c z+$Vzm+d8)5-_h9zd%EgPnrUqe^i?4&Awcw9cR;nVQ@0q7riqfg!-g%&h3*QVPcq6x z&=;JJLziuhG>CtK$W2AHiZQn6(FoM>jsS^JiBdYoSoAuw;ZFuPE)or$E$k3l}w?l15~W5 zOjxU_bk&oyGiu)yLMU+XK_6ps2`4V|g(LT}v=RLu<(e4e184?lWG?bm@EI+6aoByl z5aTF&iCwGCQV{~@sV@1Rq`Hc9+fJF>25QiaT-@eaudLARU-D|uJj%S^W=yFILXR8o zSqS)c4=>)vk>pN1hYg3(wm$;3JSe~lvrr)aya&x9OV&-!fBgbhrxy*_C;i%)?I*$n zgDM~}i;g@x_z1Yysr$9Cm@Qawq7#Xs=?KRBvCk}b4Sf#9&KvA|+lk0HxpqhGG((+) z|HbjVU;taHgW(=W>FBX|D=K)8sl>Z(W6z&-iiy%0a3%xE(-#D4TVpAGNq~vb>ANm3 zpslIahnGL|$#@KuUc-O+?Qg!UoLym`a`sL=i2nXp5wSf&AF3T*horA+wkBxZFwTq2 zg5BK$K)4drf#-J`6Zbf@bkYts3%QfHh7tIztj7}()TG)8%fqg)2+C!+iqe3xexxIS-KmzTWCtLPwoitK@OOQIo!e zrrLc{dEs=Mm0@5)#>PBAXv-i>LHAN%POX8r^|+kR(k8mse}Rex;cq*Mg@0pw8PWU! zS9>L>xS|25CYfyVN3O)M(7!WrNPe_(nYuE=DK6{=pPpp-U#B?})Z6xhl+U}ru80k# z!Qe3wEG{(e@9$D)-xR;;qZ#crHXq8&8?NwA%6*6GUU=z5}YzsHHb$ux@;eUJ|`B*UHD&WRM@%;L^@>Rs*PZ>wSvy1b|1_@F?-H~_2 zKM=APNq=_P-A-DRdX$mMJ{t-*G_Z z6z_6?>^EchI|~T*&DfdE{a%~-cBe4@Zx%33D+YfMIppd|ZK8-Kc=Wz4LrNY6nj{WI zYY(i|-W#MqM~GfA5t#Ed`Rd?IvT&3^`MA5Nj-tAd)|~3Hddkmr&M_|x+p{o-w$Hgf z#I3K>7TlGkcv+}sH{9z6D=jC@OIt--t3OIxTij_WLO-o0jvvbC0N9^Ow1CQKP=?=} z54iL!P?3$c2A*zgD%3{s3MoQ(XF%5IGbLTf0qrIE7|`!V;m@Vi=^h`Qnh92QiQLT^@uZ;T2Q`;OpuPC(}%dYULEKrG^(Y~!lE)S#>k77O4HK6O* zUr#?$4_)}>tk{Kk4^4WC586McBn?DYe?dlqlDp8XE4Q!})o5`q?f~6@DlG7!f2^<+ zBVqnf1}P;q6)W|=JC=7Xxt2E1I#PxpNUZLap*TIur|t5f{h8opE*~1>U06ZQ|K`?Q zyeK=!_`c;zS`m~`S|uZzBg&`Ry~qPp^qe>r)kCutu`ZF9Dod0}bI$_*C}4Yt*3MZT z#4@$aP0gC4qSN4{a=**~pC4;&yxmp#$%*C@7^O&gW9uc8u?Ck&&en_t5#zHCh5vaC zuN_F1AQR@U9E{KGAT1Y*QLGA$Sj`m-#ZtqVYP0cB{<@{>%2g~UL|J-m`v z$d%>^eCEQ~tc9rg3)&R*+u)|{Tz<37;NPZc$iApDyp;K~Zzrj_g=)}m{Gb!YyR%r; zW~MfkI|re)hv1j>D%fK_VwRkB9}RZZ>M%|*TZW|gp?)N^-wh;I+?059Fsbo|Sw9cu zBjzdlW+1eA<>*(|6%OZ&^@v)>^lv0M?jaJv!BEKQzw}LHb(_!^uvh(l{2;&LKvq_6L{sP0(A)1JA^R+Cq$FxK z!Yl<1^Kgf-prE4!vwrFC)9L2vhsSy3L^;dFe5GSU$1cY{!g#;8m>aY(hcZmdi-eVb zlTzkOkgP^1o?pjoYTon@6f7%R^@f{6sDGS=l844Pc8^faI>+i20tfR~CKE#=aa5&#Ql{)i|KeDK(T~dJJ1N>iW z_cc~JMXH#qWnu0|XKDT=s!d{**O6@N@=P?(4~`WGXFB#E$6dlz zxQTmp=}t=ho}hII6eLWt%_?k)mOArAQ*8pCVp5dOiX;9TYl?4w_1L_*j5aeYXy5@W zz&mqt>mt1mHT?n?gAV?Ak)!$&^f@!6!bY1mcMBEFv0wF>HrG6?^%5_i`d6=M$=d?B zwB1&v-l%*tj1!3(I}bfFt1A1QchW$OZJ=@nyVag2`^7f>2!3$6#SQA=uE6>$WzvD= zSBiu$q_M3|@c1j$lG$!5D3%TXf*n6?Lpt1uHFRVWC{}ZY5?1UuF*txRU5E>`O%|8= z7b*&~qMx+f_XbXG4tm5IIa3SZXaj_V0yt(lu>C5~xJQj`YlaO+?_l-+w%XCT@rR^9 ziUJtDLOeA5q@!ontod!7{-yeDxC%4=aw5Fb2W4CfZCmu4;7_%L7~z>z+4h?dkqd+$ z36uT{!YDY;qwW_uWC%w8>uDFl1dMTQ3_W%@+7!~58A~V@)Popv8QfS{&I-5*ZtR?S z*$ZY|-Ds){%J>?E(F{FeBt7wM5z7^2L1pA9ac@ORL$1dE_OOQ4#Nx8jU_y~Fjbser zBmmYF?%f?23v3OU$eGUf5+%kCN>6VDO{N0b_)JY?gWTvtOr{1*K`>}S>D$cr9Dhx@ zceyM{PHvhW`?(&=r0u3PU~}u<<0`K8YClg{U|>r79x=<2V(>%R zhq=L#n#*>yE&Z6ZAixYcYP7?r+u3h~a;hcgw6`jGVG?D6Z=}7z%>q8z3?7_tx0xTC zSKp;~kIuHG;kZVNYJxi~9GDl1FhgY&v<)+mWfa7b9xaT{lE~Pm<3?5UwRt-_JG8K~ zx64eW;YKw=t3`{h6;g>go$cQoopAE$oa074bOlY zh(X<}IaPC{TO{-7`T=v9QCm_AULb09*wZb{su{(g_R@wj3X)upEFGBVX+1mHIdfM{ zV=%>-;K#q0S3{0uotU4Hx0X{2(vpvKot*iHH@S6jCmSP=OuzdxrZo)C=3Q-=N1LXX z1Es&WtWJ)PT_&ynZF}k&Mk5eS`6PXRI1b(SAeAp*DlXoPcN1Y2?j-Z~Jsuiie(dVU z#T%M>NGO;El9ha)_|oi#)%O+gB4Fjf{!sfcIzFM9_SrZvzm($H!j1Ah_gOtSscHj$ zpYo@b?Du|rDu?8v5#G8C9i3CXPO}Px(%%9`XQ=@(Y+?lja<;5u1#qlRH5}>Br7sz{ zK)j)iVcF;-%YWY&j-Bt{1+Gly!n;hgX>+1GLL zLf?j`W-Tr|zHby3i&H&Ex=GcX+x8dH|-jDsA6XSADcFqnUF0Kx6gAXCP`jP-jtY!2DGy3uAa@zs?z89Fo^B2EX z-#7WJCfe^a5=<|p-xj3s{q??k zKfHY?z?^;${p$L7z1g=*F45DT^z_3@aRWLsP@KR9@_xw*b!9?!em-{OYLNJn-Q*cXGb_1-%t1R=kc$lY? z-0}4_@8kSGA!Ev)FTi)n7W;paKVdLnOlg=q{d{3K&(!H!@(g$=0n=g8^2{B zQrEvPb1tNBXl}i4$};Qjiz<5guK6YuK~;ohdK&I38Jqv#2SYA%Z&YFpL|P*7NXI8{ zCCAOZh3nm?&02~GX8w4c&R{$4mPK*<&zN=0RPvt@HZY-VLacIPS&nsAW9RS~0%I64 z11Ex(1Mrtg2FLz_eC<+PJ*UMC+lOB_!`jFxkmKP86_6ix zg+AUDA%^d{YNdcw(U(YI11f=p)4Io5=dV53F7s_&N^cZc9BypQzGJc-N84q@Q6y1C zhTnVyt%jss3+v6US{t#t9alW3+Ra|q))%JS0$8CFJ{#+cWkN8-M(KkJH-ls$Z|?^~ z{Mh}(JoONyZ6j(0xj-$YSjNqE4R9MKBD;V6f>59C9gSvJ&KQPujmD!xica^@;2+p#2#KUquIba>Z?15j1 z12Cf$GBxA)O)^KURDj;f^1x-mB-GG*@s5Kw9P;B!Hb8vag|uoK${y2@Q_^yfFcC`i zqX#cT>_b%^Yl^OBm3cK2PZ!PvkV4hNql!k5hjB8Yd8U!i^ydS;)}7-ujN<-69U5Ko zz!yPB`snk5rA<(WbU5$;T_gtgFGhq30id4;XbMyF1Z&o((D@toxdH8HZ`ToeH;asC z!WnE*AXBhvDO^)TB<7(ESY%bAGtI5_`nM*FA<Jvs3nQzk(F@=LCNT_?^Otq4N+s zDPzr1Ya?Sh=1&IDeSnZ9Qii+r+hbEc2WZ+OFirdxgsTw_3fI>vHQ&|$g<*Bn3*bO0 zo7EAeWR&j9!}P!*kR WVhqwxi}45T#LYbJ+2kY1O4mYCqRs@ZJ^ARL3=oxc zc*CAO&nw@uWjCe;icds5&76<`*dlv|kB^yzw#nQ*zqusgzyd)|xXa<5%@&F^OhlA@ zV2a`XzrlV59S+B>n;bl{2S5oO3J=B7P(7gx@eJwk?~XS(E2^w*xp-1^a|G-4OLWsw zpHWyD?S|9Xb5H})isW}PEDcUQ+C9W{*URgsLn0u9SAADmGLxa8Oci_uD~$yN{L?_E z0Z`96L~MLSfc^3~XlD?=_pP5)*lP`MujLvEE^nSU(J&1Hmz226Ex@lb1XkC>FLh-0 z)(<)LI0TaO0~Z^AcfrioU9j3Ae-D`G=evIwI2?Og*TcYm8V(^W?`IxSV$klyKVJ01 z6`waC_4TKzzaD8HtHGl{%0Abr;Llun8qWj%&Z8~JC( z+ry#qRD03!NA!*HN|rm1!p?6R%7s^tYM(`&UhcG}3VEb%0&0%*IhC=(yR+01mG110Vi+R(2g3MB0n zT1G5O2Lm*m-}D&D{wOx|dd5_VanxQ6fpTM?W-QJLQcOfbaA&0|&vfc#RTl=(2jZbQ z`*xS+DUTLxJU;CoIU%&Bc+R zCmkycKCGHl?phs|hU-n9bj_Be)X7S(Y&o;#u4U#(_$ONT!m$)pJb95;EN)_ zvkU9&mb*CU7P%hlEZtN=k6>l0)yACQ%Qptxbo7~KpfbkNoZ!k!48%$JK|x;|8o#Bz z?It~T7EeSZ#d*5J+`_9TmKl9c>BeOY#d%I`p?{7N#wV~rSCo9@zm(P-p7)=-gZX(* zC;MUmCk+PBje>6d>J!6)LjP+o9&Es=WM`p2J`w}=>6_`?&gZgp$=00U%oiTqqk`w0 zV64o~Q?>)xD9P#L{QNCUuLb@&^FM;+&*fj;bC!vZyFQkSj|T2byR1F5Zg`b^&ky_g zd`tHsxm4$>(_7uZY8ffBJEa-}>fRrV%S~ z4q*o`6yAZ=^s`pL>-oU)zbvIaan0e#-agwRO z<9lhIeXkC6@6n(q@P0$?N@k7KO>uU5wElG@iu2y(6gpzO135O#n;b+?E#W!eWpn>- z_?D0|U4+p&FuNW4`3@`QDNLgVNB2`)pH!0jMla%}KYhcT6lz$!t~tnHO`ebm=a#@M-d`f09gO8-e#f2qG{JOVx|Na21&G$YdZwL5bTs^4?4ORa( zzV@~j#}_)@j=7d!kJm>Mu@m{Gut_sfxw6Kj$ML;CdDh1uXj%K8s5QtBkM7404TVn3 zK7dc1l!He658)aXfrI1!0j}8?xl)12!6*S&S{sfy9LPSoHNAzp+Wouptwl;@2}Kf) zLlL_5j%%Ern297-?5N^Axb2rcI|~Dm5?hpt5&K)RlB9k@2==pE7LV!qY;u%_vWUYF zhBl~&iV4W5tWweQ`K$rNEp$@X6hdPK&Xgol3|R$}N`mRLG5wa;@|1;)P(?x;RC@q- zT=O9Lpn2%3WTD`o;!)%n8q9f*eBG#Y7sb;sp-PT^%13Q^oj%s>{C$Vi7o9yCjWFm_ zNKX}H>edNNGMOM54hVDu%qmF>cR>z~2q>gLGH0^D4V1v=C^?OySvWs6!?zGan%@!r zbg{M`di!EvLooi>K)wG)VGNj}CQ1RIvsPihhv2-49l%irN2^FBH1Stp<~vIf!8Fn3 zYC#l7^Mr_D|M;h7-FZ_`=hO(Y1A$VF{w~c0R_tfRL}hpI4pxBD@o%cdR575h=OzL7 z;+j%7pfm?2P;=}v7MlB&>h6-ZO0!>0*l9LeTW-EmHyhr92TQ?=V22VE?)@7ec*gge z%wj&Y7p=MLI0DUoN+OLDvzPRgrWV2QleiWm4xZRO5SpLJT>~7W02&L}B#aSXsK=j8 z!hXHffb&#J6cmGml2a=O#WC)P1~_>%nbb(3l7AD8&elz9@UJn%R6##apG7=tR&*3n znKCAATo?ziHz}7puqK0&U}P;oOeqNdmvPZwOoSkm8^lb7aQgS>4O)RON5u&58vLe? zqt+D_44W+z-S?4wT)rOs>#i-I?^uqdk$S52a{Mla%yZ`p7j4*zZ58_(&YBSY!ANPW zIjAntms-rIy%gm>^_ugU{lL|E`!YHh@IT-iMWu4}-rT#-7L2sP=@kFSQb75H{7$a3rydt@*qvY7gt?v52B04L|avyrWe zNS_btJ%6fmS@p7&8s;f=%j6l@kIf)Nkzy2fqV6X3bO+3p0(n#ZS1_pTHc_a#KDv&6 zUuAiE#@Nx=Oe5BA$_z)q;ycJ`fekrXT)uhBhsx7<;ir09!p)}TI7m4vgQwWiqOi2= zeAteDR@2oGhsU-q9_!GM<`T|CJJ!b+vX$EQd-b8# zL%Dc@Yy2^Q+^XSWg!~4>p|Wmo(ctesslK zFf7Kut2InDb+-?q(Swb$r`*gh-e%q zBTgC(jeuCtlwP#fJBQDO(*AX6%99kh=3)9Cyq-5hz|C{5W28q^cOUQF-En%xG1a%4 zZ2mIlzS5f;7F;fXW?cW$>+<%D{Oil9Rll(*Pt zqKK1B#1kPcwTd5oeuVc~jQ#1qrNi-A|2F$}BIz5~GwlYL^@9ig7nXey$Ry{;TvT|N zv?)qL!aQKWwcNm9H!uU1%l7`{5bXqRF0_>5SW`Csb&FHj=t))mZ)@0qSkhzwhIKTf z29-y?E@0tUQhHwts^)f2va5Iho&uxXtx{l#pT?jTvwFIzVROR={$G>Qr{Uv*ecu_0kZbR@k2*3rJS1JC)#aR+hx4(U+S z9_sem4ZM|~#ifB6&a1cjhV#ZI-a%$&s0`Ery6-WAh^gcF{?HP43L89oQT0$yBn9HTMUW$lf=?|f!i*3InJ9p{W%kpyfp~_^&XHL92E@!8Y zUekI|kFs%%l-qV)hQnjtd7L#SFvumT>om<7JQxdEax|%VfM$M$ui7iqArr56>G&6V z{5(&nI7-CGSL4O*ili5Gqo0=*~2@9GnxHjaJyHlp~IUQ1tG@5MSz~jUIY15 z`T;taJc6T-+I0_@h&dzYk$c(tj>IEKEs1NI*YGVnFx}AovmBL1ePsTYjb~j_pAmQwxPcIZ|KNeZXgd+e1QEgu6 z=prrTy%VO6S0@*ul9lix$N{NK&W+c#CjikCRIG#}GJ&R{5c%Sn6F40vK&i*-oNqVs z?~fK~qW3b)78hkFv(ju;J~lA(QW+|(p3g==q9A82m2HRNOEYr zNl44~3Woi5!D-Umolt5iPv1#juc72&C(q;KX8I6abALm7VPptQI8fH*)7NpTO=3@N zzEwu@Mq0E30YuH&CqA3X!Eqcf#{EJM*BWj5(AZ6sUS;UIa2@K)E^v4 zkDKyvzjLIvyX_0Ak*|EB!rbKm+*AG@5R4=3MKRB0DiF0@xNpHmvdOtJjN}kR+~uW( zfb!3yec9xsNIf7l3+m{>R)y|IPz)w8mUVy)w4r&t>lmn^P+e83)V&ksICHMJ5U0m6 zXdqG?Ech^&{lNk)^D>2{F=|n6uW(G^)7c1BZeOA`B*SRsGi}Y>xr?R7_=sr;v5$4$ z6d>u0 zIkMEMORJSo69nAw97Rp!rh?(=%bFcuWWT}m*V-fOL_%!1y!VS59ECb}4b;fo>#z21 zua~~pV5<9Oiz%Z|%egeMtBX#$Popl~kCb$yx0@#h!lmW9G$wWiS#T{0s}9yRQ_T#( zB~9lC=vO-og0y(D>~)3~q)&?um%7;n-Y9|RtKE~sT5ILV^T5IUa^AejYMpNNc>Sbr z(2Dsh&psF=>V`dJ1)-r;V$L-zfgDdfp!aDLt^V#pcR(8yZ>T>GxN<78T(=UTb{ypf z;yIK?D0{@At==a2Q|}cd!+0ms_B9poKJ;doUjz6)v%SL^^92%GUs*^`wcOyYNM!5N z$(s{&gAbv?*%TWpK$CIY!_XNsR>+r~=ic9_H4n_cE!pGcv7)KK5@*|l7mpuuJ(Sw; zE3%~YD8|WkPGG$0t$DdEb;F+nNE?Gm_)p~SXXHz`AdrFa;%jin1O8G8Ulv3O5gfp z6Uj!*rdY_C7_6x;gN!8S1yl>6r(rZnWA3$BqBACI?N~RhPqNE9C5d5S_Hcvcq@ZZwJt2h$C+>rk05UtPxN^dOc)B?14eU7B8dK10fpMmCgsZu=2 zcvG6FLW@E|bBp0t(Y~>DGN?ym3vqkt^?rCt59C>p;IK7aQC0>q)oDVSOlmRBwUo%@ zs{7*d^mV!VOMi5mBx-mQ?ytslbW;m4yHpxGk8^V^LN^}-fSt*ziSZtwF>n-D?1-|y zK+3#*R_=7+<-?r>v?r{egc?BWQNJ=4SC`_+O_VkRxK- zbjQ%K{Y!2~XW7Mgg6QK~B1YR6>TiZ02l-M)?*?I+E9xvPBfh-i#v+tf zzQ_4uDC*@1Ba8dgKWnh|s$a|!gJ@h;JdOmfi)dLjEo zfppqg^bI5yXVTFCvQrAkgY2>)hLa;~2^{!&IKe_uWb~uMfudOm>=MB=ep=|A4QZid z>!)R#BP+a3k-?248x#P+YKlr@xW0MNXRlr|-xL{bz4XRI9LcqP6w!~6PbS7BnB)l zb1;o_2FD!OLqVgIK$hsRplTsm?0Dwi!|{~mgFBDRAm#rO>r^UJ`2L)4hD{0^b&UvN z*B-bK=MAqEsCN{`S0zwKlaA;N8AWHiB(;3zqB*r?k*O~L-mvmk^yRip;-9pqM>HI% zq!cwFUR-MijQx@Ip_s;MU-lm-CIw*xK?Zf;ptJv;y!Kw4tyojF`NjA2=peEIpUt2t zF00H8rd18@ALfAn)eqlvm9*t<2}Z1WlH>t>$Nllb<;+Knk07yHJ?hK<8!w_c<-8~K zG%T=PX|M@t-}+!LZI(2Qrp>Xq?mD6#7sEG+NlXwL(C@`GlIW-)+Nq=%|0=pi*btuJ zUT|r+Px_X_+-60SeiX!OH~~v}ZDf$04E-$#2WcAYD?emOD7H2!apJHhoRnbKkXHqV z3$zaA<++Vn+J+`<51z%E6hl*{Hb-~48`9{7%ddD$1i3wer20H(s z*WX^W0y<-p7f4(E%sc9Tivdpr1R4DGmOir4kGjXm3ARRY1cBr^q1!pLcR$K^&^w$} zt|HtxiqFbg%m93&8u5Z9!IkPoVm_t~wktLZ2+n<{x|)Ebrj{I)XHpy?&D0}|$4z{a zF&av4u}MmkR3A-e+>sM*Knj>8Mp(zuO06X00^UTu=G}G9EZMurKPV(Cauz;UH8hnL zR5xveB~urTXzf;n98t7=+9%o^&wkqXZx>rDE^I`KZi~7b`+AaLbgTb<&XB>hkT7xp zEMi}6g^trjjcLcWu;-#z5WBHu5%du+Yx1}7OisjHcSki{&{-;7yZ{VViR@084$xnZgWUB(Nq5+nmV>m;gCDdnT;t#zYC#*akp7d<0jU}3VtP&ZNe0tS z`?l^)V;FWLP^yUP$xNhej!(Z3BdvVW)V0pOB*3OpGwRiOeub!Mgi)P;yKzVja2K%B ztb@B3XEAha zO#Jm`w3I}r>whIH)i?p?`N*E8hx`oW!xv!JX{#YXlP}sX&zd6?^?X* z`c)!6Cc56|P=A6&Nwr#>MhOQl4&Jl;*tE&JGwFFu9dIK<_f=_G-z(z|?6C*M;4{Mh`F}U8ZJ<-5zCnMbicf*UrEal;W2X8M zfs3U!o`Z@2v;K#C;pYU-j>tsF%FgyvvZKWCoB2m%|1SuGog>x%0u=2hCH8+X7+R8c zIvmK|FEzW0+(CQYI#)6fjB!F{i(v@A;iZT{ncHom{*sP<{41-8s2NMqYtf^1A^5mt zuF!ZFuCMYesr>i%tK;9x2Z3oL(>>Ji*=drr&_?;S?V{I{$J5m(;Op)8r)N)nkkR66 z*SCw7uNPZ`_tUff>rUmDj?b$m7T{@Pyxi^aWMQ1SgFsDOqi1dGqzhhb4Y4VDe5u@6 zt;ym=MYyoS$?8M6zwnVDSGo7}v;*Wg9>nGbvysC}I&Xzs=GM%GjR>_+8WR!h`q#9{ zmD3M)%uf)Y%0u`IHb~IxO|r)ZRF{x>&p>u***ORz7Bjug1=b%DR$SV{05HD`Jg|`` zdxZ&y-*y=+%px;xFU}Fw8+c^pSBodK+s8P4%9_^PCpap!vPgz(j~e!M22q5~TW(Vo z6L^NrYjoH?^0yB(WKRLRhmHYJfT&^4Xbz2Aw1Q8N{n{*;F+j3H4&LCmEl~QG5u-Rr zl?JWHm1LZfbBK=;$SLg`HbA?moKLKW;YBewlyU)98N3< z>+g_kSJc$T@E2I|t32blsVYM(Yu9%4UQ&m@PpB?su#i`XDkj*P3;=Erb|aDt+Js+> zCkZa3IV4Nv)8Ttj1%uayUol`J?F+t9H$T zQGq|;vOv;!|6m;(1p=ZKu(A?y8-9VIQjlyna1k-a1q#de$wh$3-bRYY82`8ks5rk5 z-Rkwg;gNTmxhWx-@GeAfFnXJ=hHKOOXsi%c zM|}@9ddVe-xd>bmtZ(b}kTjY(@zMx}#F4*<8B~)aUJGr~a`r**qzn-;i(gAaF2tN+ zAw#Hq|6@9dx^~y;A&}pQfZ4nJk|0Jve5nPR_G2tw@nhiav{bJ4tN9K^d^Leg08cf7 zP>ka4EBF8dV20wY)F_3v@9wBVOUV@91J@&rcu9~(kPH)pP)>JgSB}wE@57}(H;Ltn zEU@iPuY*edd-s?4S_auRuZyZOrvURQ-w@<|iTxeKq$@@Ye?Y4D46J$#LfV>x z!%nUuC2pLYK2O19V$XPxa4$mxjddGkocLVt;5oTS9lj^jcYqp3h6)9b)IcveMvs~MNx zIe*4(nE_qgNupf?dTL0+)vi7tqjVFShY3O}90l?}jc=y4zcb~3gga9~i-WPC+2Yg~ z#wh?*dO&P@&_<@(-v?zE(<^N^GzAMEU>jlhoBjqQt~$v-b0E-)WTA@quG-CCP~*xnH<_IV?5^5(yN|`O`Hc@ z25o;~+%@`tqh-SmR56KiO!Yq?H~*89OiTwzjaT`JlO|EiT7(yBMCAO)sML*)_$x(8 z(C!+->Gm))JR9s%DeRw1j))5JZN+3l4@^|>5=Fg4n}ht<}7;#%foX%Pf{#iXv%ZOn|%!~7Y;Dv$;~J{&8{;|W_Mp| zPF_9Jx-i@;UeUV57Q>&Pos@c^dI4C)vh%OGu#pBl&h;cXe=_y1<3kenmN!Rk*Urwj>r^&8FqqK8)@k%j>;7YEMCByc3cwr|O072v-(Ad*w zpA+)2sp$ZY z7nT&SvTDffv7C=J-3f(^y4}zB$D@n8CK;YHFOMgzna=~E9=eH_8vkt=t$>9;^q>W=62Vi z%?rW5S9F=>_kr!(>6MjGx30Z(JJ@ZMwqYUp$&@MnznAy5V3x3))Bu<>qg-ez*)x7g znJ?XHcbbRdxt#v|Pr1gOI9InPnz^4;KK1u#iyi_z;e7ot_?JRy9lo3Uk((+ndWLs& zlUb$DB4MnO1F-c;W0)^Ebe{t|ZjMR5#%m)SZ+@K*hcaBYx{IkBuTOx`4A-kMECGl> zDJTEKSOXIML*nC}I)FJf)>|Z$Vh2|&m4E?liU?x`0bw&Mf8~TiDms=Bc){|!`!$s` zQ6@aH;0TiTEvAPI>%rhz>ct&jH1dRpS6&Hfi#qq!93O7Gv_ezXTn=Lw91Nc`h4Y}dZ~(+yp=n6$-!v8of&K8NKi~|DN^c0OwKSI}C1(N4 z{c$ETi|JjoeiSPQkTIW4bPH@gnN5`&zpt9Lx9|Zsp{_F0=6Xr>;NCp(stBPzQo)eQ zs!$r)v8uyF))xi!(d@BvKsd!gV^NT+Womx)Gu^t530Rf>)6kM zsUv{|I)8xrDLt;CqtS3Gpj#g0+r%TO_bt4Y_GSK{lFUMC9>3E^OP?g~EB)+kn-Y<0 zd#R+{%fzBj!7pW-s!Bjf$$9$Ss>{1|`BnLaFs|O%_TTMWcCP=X5n^)u$M^dG2$9Sz zod0|0whC**_F(vFOJ9gq!4HJ?+TIeo0Xuf)mm(-@Oh*GnZvb>EWk9`=X^OQYmq1$^{lJ*M%km7Cynfnr1Qi_N!Fpf*V-Wi4*}wduYHz+&!ha^Z;Fv6>L+a z7>{%xA1US|YQw8D7Uyp}LS+BY6nslXYrL^fAf$k63*{}ZFCik`cE<_mt1tph3_P1< z*a~nga`u+H(QHN4eQcy&bSamuq*}8Xz);lE$U2WinJn0}JN?;I%_OS+pa~Q4zTT82 zXhgMM1Kkqq+e2`}gAnvzo|&_#2#R4P)~_r<%tm~kg0N*c-y;VG?hTZyUniRn{=f{r zgmSUt^FqyFu7L^~$*!WL%P zxb}aT#cEv!35Nv!QNzy&ro3TE7bz5bJMssrU>&QX{DtPU+}g53$cgbXpm+&{|4Y8% z1hN)gFYfKM8sF^0Bv^u3!$U9a6+piV=o=9ES;9e=B0^TdQY20?3L&G9x%UMU&PV~` zw5}jVI}SPp6E@P`)x0m%Re~Bt($p9lEMy6+alY6DDiEcwh8)(E%1mU=Kr69T3QN%+ z2@PrKLRReM5zn+E^5uDmm~|4TYSDg%b45 zQFl?YfdmT%k06T?M*WCO>;F0Z-q$Dv)xUQVOR1VhUf1WCF_X`n1xr(?) zp7YCo09G`E+z+jXBfe>uXWQb9O}_xAZ*O=BKC==+;j)t_qqCJvY>+2oEPLoV0%7(Y z_;+AUd1ekQ;G&6~7X|aj3KL@-p11Cf!c~eDR*zic? zku$V(gCTv2z0V;u<=o^y=2Q)Fi%iysF0#YdL-`g7O@@Ik^~#P(t&lw}0^U=j5csz! z7P|@Yjg#zsdWY{%(2yCTJ}z5%wp$|Al*Nk%Jq4)_g~loLU)`Q}MIB`H30RQ_y^iz| zCQ3qR3qVUt;>4^Z^B*gd;RO?AXGbF#OASXTF|9WBxxLmQt)182qIj*Qsw5c@K*djm zJ7ZI-*sLwd5=DM=P$=-O_r20!Ca*5lF-tOx5 zE@bvfC8v1PW>gON^r<#IcVqTtomaD&_d0qAIG|9iz9dhuOK*|zpULsw!t6tTn-XKP zRQnYJQTcIkNIl$;i$y<`SkCmZ>g%X0cJma8U0gOlAD}B}AV3(wOLnfJSNC_Zwbi?A z9O$f`^QLI~2af#{`K~D&!PouR>}gHB4AnECQ6J;xwXmb!-y`99h%AxZ})1uNRx(^8$?o8gLuGbGavEUg@Fu&^?>1+3fs z$p`*&#t`xP_#`OgJAZ*s_K&6_L4cruvT`u~x1NI;3QQD)gYzdo2Fm;ca8c9(Cr{)P zMff>^>nAjZM$I6u{1YXj|gcvZD`Tv-*i?o}^MIIYc60rY)$_oDvRMvr5BmpV<|3GD@PovC!V`^%i zj;gplD5-wDHb{(IUC`MI*nM}tKF|TWJHIB> z&S#RD6UufBB}y;)2XNF+7?Nv`N{q%+z0sM0j7^^1!^yaGfWm^2(9FH%&1+mvE3)F*5zV$ z%GmIHTDHjrUV9TYIVJPWfS!j+O4NQ-N7nY*mg?i>Y>H-`+O|iqDOgC}#pn`fwHw%V8}9tROi?tdb>= zMq5s4KM7S@vp4 zo&$&DNa}2Z70E?HnY9J~YcM3sa@iCzwteLY67@glOraMn3t-3YmflMr^nwf#3Q$K| zM?UHvQPmI;4D_KFVA)?X%`f~`wTRa*E?Hb!!2pBTYA`CS7N=Lj3yxSs`D#X1pn3*j z>t$#e?866atn#?U9fbVDVHQb_2pE>w0sBe`m1IM)y-!0SQ@WI_anAUCE_4pTU zY)(qdf`++-r3aHr%1hIUw>aX3MRgLQI9~FzaW)k|O)jfB9$iAIUNI9p759=ZRVM&EznwE~WTzVZw48#s~& zL~}aOB0-CBchA&!ss+O+yQ9yuCyR(vLUpODfglvDhf0o4YrBOoeBS-W>)U3YX`tz^QGSu|y@s(Ch%CC?ZT}(R_5yBhXdKZQC>JPJ$zu|Ufn!1nhk|LUC zvUrL9z}u+P`{sc!2HbS+1X94gnZp3rLnu45zpG%k=mRRKM6= zqa*5El`Y0d!R|PKb5-W}t#f&7Wm@gGg*(B<;QFA?$Lab~Sk=Qa&4Bwc!edOw z=K7$OjNTExN(tSriFGJui=9R53~}w(0=KEdkK1}ygNfTZ#PDI$e5+$ASC zck=7##e(>$RBqz5H)llL#gA**Wk)ma_P4pL5#3Uj@Uo69%~-9bzC+zOESYU7YHOI5 z9C95wFIp^aBBQl%qhvB=@1k;2jhCG^$zr=f`>UHqm#y$eD@>K5!Ej^ZhCw933stU+}C?;0nlC_~(L0u3fC~8;aE~gfC8$x8=V-nC&gR=5{ya`I+##V%rGi?h8_ zXANvlz+`E1wVDB&4@Dae>{ zbw-%+9jLIfE#ZZ=9d&ccK$Ob~D|47a1oU?q4df#hk3RiN2j!xLc<`o+>6^oZxr60n z!P=~=(1!(7)5rEcPsa4&jt`J;+J&=qevwo2kDh2Pkn;)_2xR&_Rs3-YjonbI70iR}YjQXyjVg^i$Jpkq;)S6*X zY)Caom^NlD;ah5uGxqD3Sp@h|8PsvXXLi6EX%5tJai%QFq{?;9OqPsA%nTAFhcY8Y zQw!8FNj~2%Gv)j#)V}WnV=+{_g6q?f2X|5*GAe`DnlfqpeGzrlG{>!4*fO7?EIT&G zO^n8Z=CrUqqarRK>vF58s@)p%hyhd;kxm|NZ7##4%=*+CHYG&hT^?u_YaN?q1W37x zRGwTa;jVuoh)B=il5|nvn9$fwxl+|X%$fTW!KFnF9D!x*%?o)be$cUB57daucmbggr zgdnnC#mOGa{GZwiqA8<`^bkXmV*qrh((v$Pl4G-pv$#m#UW*86hSGF#uqu`L5_EAA z-A(oF)re$sK#7|6DVqnvJ)pkUg*iPWkH@R@5a{x!)u!>$11eGqTRg5yai)aaY&GJT z%9FqR?C?s8il`^R|zi;?q0{I_J#zz-oiE5Nk5+46VdO+}1ZyT&ZWRaBN zpAF(bn>i~rPoqj@r!mkbN$uZt5)!8eK*a$qB4l0nDIyzq;MmQ5+T}O)>+abw&~|jD zM&qk7z%p3M$zuedArX`HR0j}b@TOV&(Z+&lZz3ap4i2fg70jx)@FY0_sZn!VM7zCe z5W9F1Kr*UPLsLcZi2wnRL)J$1Arl5!Wf3$;<-0rv(UQ3q;~AiT!X}tLHa(S(rWBk) zym2dxd@^e4y@Cjkm;EvbkUUx1o}5i$26H2ZL}}SBS-4-SxW_^WGho%i?i%PqL^9Gz z<;sHYyMGZ8ZD*S$G#VFz#40JEU|egD%jt~c<7gm5JB~<`1^{r9IQH&Yq%l165W9?! z;u^U4SC4}XKn||^BRLk4gVu}l7!h2~4f43+8&{dIA-{4de1RQCOvM zDP-%@VUJ0}q;VhowKPTX)jE*VhY6{iq_t6}uG7JegCYg48l<5Om@A^L(>CSOH{<=Q zAyOA{6Cv-H#sP=^D*4$DbES`)2q)eqNBJ)uay6fO3D(O{{huRWM$yTV}EH*WGO$w$C4j=b0VnW;61(-EDqcF7Xep=nPdmJU*(XBC{rv z#PWQSL?IJbLo=6Hwugbf^aXtK{4p<|X+lzex!*XqUJhh-3AJ*)v#t(0`giB7#aaKd z)rZ5hp#WYube4uxW3~<_KDR|iS~RcGPg_Pnd&v659^GpYKAI=mnteBJt}aS3AW#^f zKWsYBD#X(>B&Q8duDINq)QWR>b=@on#@epD`D^C?Wj0;EbxK|^EVrLL5Lfl6cjVM~ zHRhQ4S_!aPgUZ*A$ zh0nR2-k&uzJ!LJf@a>g|X`sV6|6M)Jd7#@H{Z;8l~7TcqB<@ec@vQ zj;?qF`Clc7SbDW=1@3%G9sza{Rgjnb4EIYzCcSC$8?PZ00ZXL}9ooG0@$uT?MCz~i z6zN&JwS+Hk;NkYdE8+77kHIlm;NK+gJVg)qIJn5~zG$DN&scT@|2RA-|I)RsaIiC5 z>(EFx#}~mr$TRyL#}OE^zrv3=rj!i8XL~Qe)_ydN(a-$znr(dEdc9)b?wvh}KfTGJ zY|_>hH9G~YCt^0OM|{OrY?lY`8?V@Qtth1=DA9LrW&yh3*Bvdl#NZiiik^(%B9esi z;N~|hc>2RSQ6b{pEo~ZZIDRqEXOVsc~O- zz1&fB6_XwiiGYHSdamWR}6Q zbG~O$A6i>KrcZUfhs^v(2YTis0%CDxw|(5W7qu|2?y>uf zH-}f1hd{ECdkGr{QiGZYP#DOiPQ?_+NodwiO=qw%A$=+p;kO@mK$S5Ax)X9Tc zEFFH8L+=#ngn|7%x;XyIqLA4=I~apUb&C|TPM$r#D(sC zf+mp%_B}86JC9)j)&g%G9c^CE5FVHx0-aOkPU-Z%%_G4ky}S3dK^l|)HGH$6N={JK zbQw@az;bTTud@8mHf~euoK(DazlrBIsyDp&a-4Zxw!!4)J&vKwz#F^CAyjEFuG+cP z)KippPCr9hL!bvSBi?G@WTG>1?k#Qh{G`gO8_lk4DTH-hJi6k+au*}e)sgyPSVZeLM6iJ#k-Uja|1W^Z$iP`S-Gj0+Ek+QecaW*64{CIA~b2JDq2slRW>>8 zNlies!m)L{r{V^K_RC$>5s1Xy0X4UHp>3x5(=#b%CU^Aym!yg?Kw!2)k!o0iP69Ys zT@x-ZPI=~-a9p`)$BYX_#=q>#U#Zza%~e}oH||Pp|D!dHw5(gvPVr`HYn`e?ZKM*Y zGtXiVIqGgNC4%Mg_^nC?HAsJGce1U>pIqL>VB7t@9k~-9Z>9_m4+gY&EHp~{3_-) zp7qrhQB<^BGq>o1{KA|fMS)ng7RyoQ%xCrjqM-@fSO}90v#Z%4z8))cQIwBl!b`DU zNd!^aVp{zsMJ(koRZ@k;@K?eqXAu4|TQ{CK_EOjIg73{fW6`9`BPt?IXqLl9x=Mk0 z1yG^uEIFtFg!e9nzN{`pHN=M=LpSim%Y#RG5L2$$VW4J@nIez=ynp~oW+Ud0SwmBV ztWXo)1P(59=h6x$YgckOt?yo?E{{m$E}q@=%A$=FM+^K8r|f)?Yw9Y``<}RY@&lG` zwyFyx04bL!YL*2p;weIlve5eMIKH43gEn0xEUGN?i7 zQQYw9lcCD9$ye7QGfl5WlhB7&h2z%Ekf*5>$)0_9=1GY2ItJo~4tSi+?3&dIb{v|U zb73NS{$L7gnQznK!6Gd%+$vN~O&BBdFVB*WXJ+>EQ~BZg`HtBssGD14@+eW!&uI za{snt&9|>Mt;W1kEfsdkn?F2L8Mb~WN5fHQ89o#Je!?XKKX~b3xnq{0p^GurwPbzUu?Dhri`@;EeL{ZT-tJj4rr)K-6Fb>bbC~392xo1@F*wZZn+B}lo|Rz zzP%}4WLAtBUjagJTVedGn)*=(QYd@gz|=PUVkR=w#4kuIq7RyCAU5;p0HHgQxiHbV zbFy8UK`TL3Ce5uKVLzy>BF#Bde=9V`@8)ZuA4m{|E0<9$h~Tx@Q=)eDE*D8_eeh>X z`N_Gb0Fx0Ctwq{2iAc*Z#kDvNgE-gjBQEHlkkX<3{hy4?15ZV8DTEH1)=nkTcNfz=?t(?P}YAnbe6-@&H*^&p*Z?Jbp1p{#nw5hb~fDaJ9A7g%^ z!oX$MBiai=ehhO@>@}WBN~e%~ZU<;Z%i}g;4$MD`mvZfGNE%jCLvJ!j=0-Pc+LTAP zR!k=xB?cuk=`U=l6qiI?7YE7hvY;=@i(;dK@KqOsOz8!cu|=5h3oHMasO~yj!F#<{k651|wAlAmrILug@i* zlw>IrH+dg^Zq?bUKh$*_2W&W4<;_A?9|Hk7W94JEctQN5$|_Vjo1S^Xx%2bFE|2G+ zx5ig&WD#CZQ;)iVeYs1_G~W!w^MOKNt(N|fkk<-q8<*U)` ztlf&(9tPM$7Pi~?>CCwEAwp`f#^J25OG^ni+}nMOfxMkkIKRQ2{<3d?7Xz&$qM>p|(Bo??}3%H)ZjQlO0@JNePZ&uOX-L^ao%kA(3jd5*tD+5Y{4jrmb8_W(x8 z?%jWws2_=UKs3jKh-(@=f9z}qNo^0VX#S4+@Kz%(8y`;gIUTN(7U1P~ug(@Ta@SE@ zC+17Cya73|@M9!NGJIwY46X*xi@`kNmwgjE}xL9sBgtF5*?W5-pTmrThO*HPud?P*zX=hfY zfE-Eicb7U)8d)Z-XC|2&$aXm(Bs|>~zVW;Fix4Q2Hduzr4b$PU{Xe*ZqC0POAfF7QA#-Hk&cw50DEs zUIEeQEA)kNKqP`^=h&Z$sa+R&fWgh9|oyKM8LaqelrOdH2IZTxyssXKY zG6wraPp7~mm@(z90rRCu?HSMbR3Tc{FfLH$NX$H!A4qPM{Aox3zS*D47^ghzgQD`6 z)2IeCv2;>4XMEtckKy;>&rY>lAE!}Ja}_)V4X)XIpix;xB?_;kl3RUK8t3$FZE+i~ zrhz&u)a{zDYcp*T8)^4LKa@IIFaa`3wn6=+zkD2+Ou6yOf&78Yozn|_TPSG0C_f#; zceCU(W%|=boz7FI8qHIJsEY`@=6XmeUx&fnB+|+zysOgj@zCMkj3BOspxQfpza=&l=={qmbP0 zgUrR4F2-2EN#6OSJeZSP7KH81Wc$<%4{(tJp6DfJ8-`llCVKF4^K!bQ)cwl`pmdT} zi7~e3=H*+;wkDe~V6F4kl@AdybW^Mu1rO-43!uWdOSLAjU|mA6VBh09W~79SAeQRy zUIQNu{p!2W7?(Yw3gHc$&=!q;X31im=g7`ODddP&hgp*1>v%ZdF|b=kg28bmS+R2i zWN&*K__EEcCZ(*_K`X>lo9p)RqYUW|&*u)B9Q#EazI^v>>hj(CYY9T!?wIb%QV@#4LR$8_SM)30U1KVdz??rfwu%Fe`6 zC#4TNqJM{p5)vAqUOcZ!ugtlTLP7fiT*E)$HM1SD0}AN~pFGGa&pBX4jcq#*z%(hF zZTF7Cc{aV8_6(^2jpV&q`f>bO|!>T03 zDpS!7afxwK8ukM4d(A@$nh;rP1m4F?&_x@$Klfh5ohe{c)xu4F`p_1&Tt6#-Y9(h+ z!UC~nmPM0;&#zwHK^-Ew=}=UG_?#4>(-3s$mm0`_ z7j&Re>RWePXnXw~C8Onbb)gq9Ajqh%P`d?Y78v>~89~rVDh(wvKd>N)mz@bU5QYJ~ za#aVv;^BTctq(B*&-rW-X%s_2f71=r9*kRfRqGY^#`i3emd55c!c75Sh;{u*eH*T~ z<)fGK`7mQWC(L^ES5IIm_X4ZUB>d41%7o9@ssK8E5J;m<;u#VIS|aSfNj^K`Gc#=d zwoh-~rLpL_4;D08C&s<9Nns=p&g>D=9Pcg%C3r3^m`YW_->=YD^gv&6Uz7{|%anT* z(`Gr-=9;qF*{-3iZ<-9i9|g-sBP3h$Au!os{X3GXtpUlW9-;cs*yXrRmaAOq>uSqi z>*Qj>7I+z%9}IZ71F8}Xi2nZCSxh~x58+ymT0|3B*<_(Im0Sva9A@9=?|(RX*iN1>MNV6q@TA1#xHq-jz0U@Ct}~+3w(zcH zGMi-1MdqBqi7&d8omp?LRlk)LU^Y{zxe!%|wxi3x%^f5mTPdZpQ-0|d9aAsc4!=%rAyqs24$*Jd^hz54K z$P^HJ*yGBk+NZ94*FJvG)#WN|%eHPNBB-H45fD?LF%(?$^t!N(Mvj>d0$1?j=fm&z- z;S)jQ^x}4vH6jRDVpUs4N<=geThJO!H?Pf28==rU2082?8275f<0DR3awcb zp=Dl1XAgJ@a=s)H58QJ1naoprz9(0H^aOCi4KA2H$lNb<6$zohpg@R2)n5!Z%+`+? zsJABIU4}5T$-fIsN?D&>D}kZCmjqy? zxyyB-rKF%?)W2=i_>hnO(C?U-sbk&&@eh~SIE;u@SMZ#aI9eTa0W|@?(ET!oEWGz= z>$Bq8=^_^TuFDaN>|Tgg*TW4pp$H&Qbl6cxTR*{bmuktmKfc<^U@^M%bJ4p{_^s~e zt?XWy8V<`&m^(~7!K5nNcp9zBMkrX^{c>V7w2(}X`1?s=n=e}tD=f{~lg4CWmm~-X z2n9v_=m~^e3jajiHS~f^1X5W~@y-H*17ZxAjyIShNFax^^JQsnmodezK^@?)FKWw7 zoTB!#LaJf$po~j~|4qwhZU1&oJV)@vL78$cw^drfPuc&Gu)31d?3`lw1$Of3TH6yvj9hzP*d!ux; z#=Q)4nWF?1hoVB;&vd~BQVs9$o?arw60#cg&-(&E;{OFq#&OYbuF zI13NQrg>5&xwHIACV22oGp&%DN!Kxn@52X-}05tU&&@$Zu`=}2Hwae;)WP3&bS`Xmdf z02B4I84g_9g;}%N?c%Ge+3QAc_~`dq3MvXDIt~-p|FUah`+qUdg62OBDyZha#P!S9Du2Q1oV{;m&V{);F_Nnt@b4EvuvE$RuO|_uPA%AKj z%_1{6gG57n!psdIH+GB==yt6vu8CDOE1H4PqI3Uhn?=U$Y3=8$*7tSTwT>yNoc{E1 zD0k4UBxY*p!Lg(%F3@`o&s({oZ#Im@Dnu8HvlS=5&=VrucW*^H6{SC*$45o(Gjgbu zQZs4hlsPDx9OtK{-sE;>GM|&o%$+OIoQ@$G?Tk4kJADk$jpbuT@*FmS4>+ZM-^dzD z9->tBu-b>LG0rfz>5rHz*e7jCFP0K_%8_~r+o9f}evW&-93C$>WCo21rVOO)(IflC zOa792=XDZw@Jq=h^B#Kfc{rw_h(xPqP|J6G9k&51xAU~NvyjrkL1QZjV0r(nJS(rx zwmVFG5fA|MqJNCIxXfLzGnA%Z+kH+D*?H=0K_tcUuOxN`bdA`_ZK)x*4!~G~!O?)VWv6m=&5dyjUj?|O23p<`zEBX-aDl>-<5?P(7*aRzJf4gvOix-37hNtW? z*w>RiV<9}(d0XB!E~K_FrHb~jW4EJ`!DjCBDKf_j=6!g}+ruh(G#LVmuW5J=R(DTH z@0&1i{f{DA6}*`-q)e_~Rv9&CI&Qi%;riZoJwEfYrd`2amEE7nq5+seU2!dXeWInvCm($&Fo}83y#m~ZF`IjP#R3oTcne;`Jg+(~{#f@vEm3Nn_}dzKxG`Wmme8*Ye^T&tJT*csUSMQ`p?3 zT6o-?*`v!|TFD|OE#+Q*-kb+_GaLO5BCm&BY|~>RSMw1P!2?oW2d&h3j#SdVl03SF%=NQ)CI5t^9=Q}= zKV4SPhf%ipU_~$}v%mIM8E*=>UP3W)oUd)}c8W4x8sQi9Y)?n?U$%Vr4JHvKh`iivX@ zFmUHIl}%m}F;T7ZIGMd1fbj7Vad+%hAgpl4E03HAc$8o8BgD!7<$I_OarIzk9cMsh zgjP=x%~ni+_jVYlmL8%BsC8VfjnGVGWzcNTk|ih%Hnd48)DpZpDNC>1);XBmi!|)4 z2%3Fon4O_urwE*XFjI>Q*o??G%N1%M#wnz*hKQpwQO9u|9>B5QYhULwo1}p_-VB^&|y(9y9m~ULS z-2aRM<7A#cEPdEIr>p{EFpt^D^f%*u4lMI3WbO=08kHgJZ$`XIx1$)mR0Wc#Z0){O z<)TojpJIFE+T#q%9!m|cO5|3!*!E}sO1%2sjXvv$!xWB@G|<@iHT|f#(AGFPH=8^TC8SWTm{yH#TPIPrJ&lPt4k&AKOU5BSZGcPB7bFujQb?w-dIa;7->PR0{J(4OlZrQFG|~#wlt2; zrh)wsU?BJ3rTM-w+9n0|21k`Qs!S#kVO@I}J+Szgyi~8^T<4F{VdR*3w;vAOZ~6`@ zJ8YED{C3V!9=6`MF^AphEks4Pv+1Cxi@uL57htK-d=;IN$lqF~KVA!$sogFBR5UmP z)Nw^#Zojoy0Dp~Xv%e&wgPb8T&6q9cA!o)Cc`7T1!~rUNNW)i7)4ybA zADV&ie~M+d)l<07Z$7eYuZDJ@_FUR@b#0qX^*LrJ&)cKCRX-qAQ}CVjz?guT-gUb7 z*B|?K?j3G!rFSWx9MMObMzK?0UT}M2(|k?EeT${(faZ+-P6NanC`3~hX~%kLGcCs_ z?F3KtHxGoT7d(meSu`vBwds1?Pc~VzOJr_6#-8~tPS4AI=C!csv$3zV9p~q4`Yyt^ zmq98OxVj8^qKN?~`6Pb;xzUE@ZiBXV{0!^^9KYhuO3gsns(QuLx>^C9j#B!}$^wiL zgIn1IfSVR@zqC&80iMQ}B478WR64@~G)2gF27M3oo$?fE+A%y;PmnhshkWyvy8*g* z&yy*grGe+#f`JkHP-9osJ_a+@@+Yhj1(G_FZs=_b8berwO6|xpX&Dh z_r2rRCytnquNb@Fu4dUL%nXeL=HK20;5*m8mF=@D4=5K`dW3eL;PV%{UjG9pN-5!n zKupmhfkaH{IDmu!=V1Q7d-sWdpdr#B%+5146exsW!7Eh_DX;F7Of4T@&!@1>vb!wHQdLFs z@%Y-BGc4@j@<%le%JXV`U28HrQh_G!rMpX{PaC%*c&mwbEAyr&aNzw5f#>|4jn!cu z@x6mb?~moCu$8!zdq4K{`%{aW!`bD|JzO?$Du{ACA!i*;l17wyjw6wA8 z9X&>6ZzqPg{%;{JUZ0P<2N zcdw5ObcgO2(}R*$rc=qT%{$mltbt}do9WH25f=x9srXq7Bq3`CDBl5v)eE9Baa)pv zMCs;&(_Va2w3#;Blh_l3%B(JdCvv6XjxqMVJe4~kPJdd2b-+cSpH1SQGF?XeH%wrw zWqKr<_NPBm`Ou^Eb6b+Y`jLZ_@a&w%0tOQ6cb#TVJsyMFudurV&AUxBE4 z>ZPjOy4^4L?c)azZAWJlhC|cc+53BVvQ`6dmH8z71=^f{FN8HdSSiu30`+3QiWi&F z+XNqeOuXyxz@*y~z!G|jqo=UHA!Xo>pGwhDv?6mOS{B*FBD*r&i8}1I=Am7ORuuYd z)c_Z(oXw^azZJ)-5Yot7~x9U4^4MgPjhHr@1au?y13@@ z5y%C^ww$C4v=H$INLgr0upQ%r^J`<`AD$7Uz~)}4iG9nplpL8%D?Ag{W@KJYF-k!W z*I%aJe5yvrxlQNCYn*4T8zM9cWIXE^!4h@)bOo1Zf<4hm23rfbT7Dl)LyM_T({~V& zbSCgO0kANVS`Fz_9paA{UW!a43JWh*dMZisUc1z z*CTX3lB3nu%ae|d$7=h{ndtK~lppUC<3m$z%C98Zn z$#OxN-;JBWJ|+5udV{<06;SNM4)VPm6hX8gaEFFfS*~~q(eCA{xPGp%axJ3nhPq48 z9z_yqo!2JX%Rx4TEALHj24|%MzHwQhb zP+foeBVoF%CeaH){kjl7PeUQtJJE)sVHcoH6Z`9{UWy95V4h17KqK$UgDHev?uzZA zRr2)y8aJR%b6YA6HyjK2p3V$P|B{cl} zlnG-YB6G|88io$`r_|ZNDjWUUo(~KghA^HL8oW4vgYg%H*B^Z6T(SE6qQn>mVTvuv z?zwXkH7|@7Im|y&Kw|TvD3zh8_c0~J{y+LE3@4^Ju|gxJg7h-ID$*##ts{Gzk-GTQ z#vjtBF|%1h$aWXp%%<<`8%|Mqy%O*@t%YSVo^(u4*^B*ewXeB6&R%8c_apc5m!E4VDP z73%%E1O8^pxd`Am*_v+#4C3ChB5G7%QMkRBr}m zi7An|B-}3t93hQQA&kK~;BDX;-E25^d4KqR{xM;q2E1j(D~+@Nlo~2}b2G+&2=})n zsa29ec}Piq@7dTSFNPEDK-drnwt6Pm5c&J2c-?&m%E)(ig(~&t{y1hIMl782bNNSa zq5H3S%AlSu+&T-B10uq_BxInWoU!>nYVDpzLbX9;gY{!Aby(Hf-* zm6ia{A5D&8$~Ypn#Orv)o@i;7h`}iC)XOLOI)K`nO~5E@Cs8 zQBPlYrdodU9g1?TqZB)Lg*C48diK}-MW;2`h+%Q^GRu2X1@GsWcDgK5x1u zK*ix$)3IkZ-**N3Tj5W%u{gX;yb63AC@^~tW1{9vpf6CKvOalmYEtIT@QaW1F&wia zJb;y}d#K})(}r_w6XM5U>OQjvOO1@rJU7X&E(tgEKye@z=l#>A*T58cL|joT+dJOz z+5xJd*EyAk+ve9H8gtfS*S^{w8(zJek%3to5=Tm~z+vuBg)K?M-XJS%(j|uJ8o-2# z3vF6TiL!cu&n z57z`c3>`CiA`;mx4L1KX4JGZX1%P=@3-MQKUTrTi(*E&Wv{q)rp2N@&cXS#;Qpe}p z-SgtQW?9fUf?3rH*Q7dIpU1<>H2R$lwM6atI%vF>EG`&URgUS-R3EQLVG1^1%nDfk4%`fyEYcpn((=u> zC`tQMb$h~WPts*Kw{aGBVhUJ3Nx4>U#h+^Eac=)A@~FP14X0ngobFQgTfAl=(>+N+ ze?FTMZ`>g$nI!zZ1qI7H3eAooh^6yS*VG>*Aj=}unrSGq71!ia7pu2JFQVv! zwsWonn3~axw87?mV9cnJon{|2@%wMVg|btkci>6K=jXkzPGl>ZegTQhEPa%jflzv- z(qH(Q!WZzNZ_ea9KLo@Yy+?aa`f=D(BX=pkc@T_s^J1^nTO6||N7!gU*^D|?>18mE zuKtu*1z#wNbjMty+_q11`~$he9*S`75p!rE1&tf?LlQ{&oe;zXN+j6N;Em!aJap10uF-&@(cSYWGy&7C=_>xJGl#KK zLSiu9k4)vJiE`YgHCN*nMJX^AbKjz&OrwV~&%}}U9Ys2Yf@iBHCsmdlUr3Eh4&NwSbeOKr4n$!805$ zOEu0T7FMPn#CCEwhuC$Oyt3j)*YkgtfG?20DSEp9$vKGsbAN?SDft4Y11dFEqVb0? z|5>)oaemB-^ed^URD-_9_Q(AIMdO487ohq%{oO11kMW$fl2PaLnorXzCxMhg5NOCk z!+M$5rE8gn@smbeN5{+D?%@Nzj!CP6&diCmLrMd;RhgJe<0w}7xlYyD%G-dRf$CEK zTs`~fDPF;J#e&RHNb151I>7n~nf=*4HmlO8VzSmO3jCQEn>~yW|I>xLz4bn%p>+t( zza=m(1ODA58oYdET{9g(o<7ruXm@oi_HQ;QAosc8y{FBAjyi6}o%*#u$N* z81;QFl|6~3m7%;$bD?N5IXykwPI-;BAs1IMA$Xnp#NDTDhWpL89e9C$=en_-T!+mT zpJnu`RxvP!UCynInw3qVX9rotpZ53Ty40AGQfpxV=~GR4>7qRE|HsZV>z(1gYC`xn zxxTZ^nV!zA_YIXZ$mNPs&7(`Z7aaB~h4K>r>KT9HdfT<~ZXjk_OiK`)d(>O9x7Y$n zkH`!YMoVCJaUa_L5^xv`us@VR$5nA(OOKe&m7)vYtR=gKa!8xclNI5 znpo@Ua)}8?*jLax5|PrrAbE!Q_j*Hgy|OiVL-%cI~jdP6N7pGdl2GkA@As^%IClZ+zjfJR?)e49pgW*@?(l;{j zIqPF%JjWGW0)-I-hK{wAy3hWThiLGZ43u~?LMA5=NR$lBGV2RhqFkD7`4DsYe;>LT zGJ@g@Pohm zSv$d62ghE3ceovsc_F2Odz3|Z#t@Ig6*axcpGzfzz1Kp+j`DHwkA=n~53pvwSxZYA z)NYi;ysI)x!lsG@kuH02a|e=0A%<}7B6^~?M!KW_gaBO-<*H_lva}gPdGJahuzdWt zJh7U!2KEo*!8S}*NFv)LLPSL5`d^CkGM0$eBR>{}kWOvY*3})i0VN%5D3+)z}3C`y=;~+S#-=IW$T#vF7Or9mBbGoU| z$A=Ao%B?q{5rJl$9Vp(Ghw==U61d@rT+O3Pg3&cq^&&Cx%80s7HoGJ;p5!niPVf$2 z0=uCV2n=5%n7iGQa4Oh{(`v-V0pEYYuIf2L*Xug`g%0`4;GIoQ-k2&Qg($JtZi_h= zE=@bfCs=>!!$51l=Frxqc4vka z2c;4$agFWpHCp37Ky0jAA*j>ZVAgXO0^W$LX1uUxyai<~TP}sp#i&VvuCeHR|P`U&~wHCcce1R0y3mc@16>eaA#nB%yQ~ z#`=CM*BU5(`rcLR#0b0Kgn9x%XRBSmh`oA&&X06|KCy-*u9DMW1tdWqm{=pl0des( zGNOr79mjSD2z~3Pg1v&*2TjWUAI(&wC9q+)fQp8%HvS(qLhwMPN zI8*+k`=Z7&>a;Osa9Ed(9@!4Iaq@I~fx){JMTjB4Bs$2LNc-D zu{1sKZXKA3!gPIxKcuEDSSp3=W#(lVer;4`(=b~-#Xi)Q89BATqHL9fCPf7!yz=}t zLn3@cp1#3G`YbJW8wPbk0N`(eSwRzn*fQ=|ot^_Byy)~olv6XFK^OZP0B+4EbMCrb zu53|IqZ`Kbeylqz!e*Q!e*Hw8^rj0BYuTbn*<+^hXUG0oesz{Zk-*JTb#Aic#w$4u zCpO*DOcer3S)S~ZughoJ7*2V(lvtQ6J!&2Fp=O76Fc+oHp7l5-(8X*cO`Sf)ptLs_ zidj*JbXlbe)009xT~^Odzoe(?+HSy~UFd!r8*FU-$)0)7_I^;dR)0b-tfpXuhp)%- z!-;)Ivkgm7&yrCtf`d!R*+^FrI!D>uk97r-nYHQ{LX_Qf6gD*tc4g%H#hN@ zO2`y~L5&9=Hic&lu9Ys4$%J$Fg~#ft*R#(K%sqhCKSdCv%M$eZd(f?9uS=}b{RCwn z_L25<>qVl?s`@<@$Lclv$8N{Nddf$ez*Db`jQ31Qk#gnxH6o#bY7 z--gf?ncrvSlV#0lrlGG`Y1w_8VNH(xsz)&^J@=Iz=%Xp^Lo8>KeFPgnX+ChJH-6Hy zPm(2qJ(PNN;6-at@_nNyBIWS;?-;QHYVNKymKu+2S44)%mb3uK=f84$x$FNY8Zfan zP9Z{pgal(|{SQUWJZ0bmk^`KL^~ZlS1L&WSoJee}|3#$zM>4?1%KTr+02?dIeRM5z^GicYdk!Ox>q|l zALD@d5C;Yk(Ffeu!}-;QT^CEw=d$GE_fPL-NmR(?F`d^WBtVSH7W3Cn$AjeqTrW*b z{mE2;{RJx8)F@tklh}xm*}K3RskuAt*M|-w%KB*`=xz>ZkCTb?1`&Kb zi6hzCm0{BC&B*uxWFy$1Lov2Lf)?mivucE3aFicfG&~nP#`26_c;6) zb2|R*FO2ar9Op=s9t5*BB*jRf*1xZ86$G+$HTE(L%w?3x1?sy3S+-@&N>^e#zhGkh zb_Tb-RgR}cHzdotnmJnvNe`JfNcnYJS;igju;>y3@Prv`Zt^Co{@cl9^zVoh6Zr(# zy;txXTwx@Z+Hb-_WCzTc1<6cr713nnFcR)|3dKP8~=e7m-L@48QnYKs2ctYQj)|U8WX)V z3=<5+KqzJPgfV7{>EAtkZ+fpBf$N51$~MZWJlK^HucGA<69n(YA*Hx45cs3#ocv+& zNAs^gncks=lqqd`a-D2EnhMosT7TrZEuxKL%DJkBB<`-R==WYVr(C5&H-s&t4Tl;I z=4}!q&q}pSp3Z9=I!xBx`K`6_-6NxTq9L-~XFCbMM?tp4seqHfM6q znX}om&+h(`3h)R!-*anv7x@{2{gWs?vNgXF!~8#K4^TG_oGSXVLhJx zIZCJ=O{|pBn;W$HX7{>qNk%BR-ve;hTjua-XQ;Hl-u-V0@t>G(;egJN=-}SQsFf-G zDOx|0_bnzE=(OcQX6?#3SCGYX*zz0Dwp+g4_R?)>vuybnXtp|BW>U^jD zfv9z@7tOT8eH32WF`YRyF_W<_{e+t1>~Yn@Crld<)fnl~@-fb_<3i%`-*)5QN}R0b z3ObON0=L>C7L{O+RZT&MhIXs)5t<1qR{Z{{YdFqq>%n!mAE5+qLIBjm)6>MtwR5jV zzVu`zygLI;Ued+H)I9ZM>15b`=Gqe7&-d{Bg#!>H(CXRNP}0h$@9iuRs^k=Zh;o0TP$>Eh=X&yi)W@^DVH+cUDMSXcf}Sn zgbU;rB{E6;8G(b`B}_1l{#{S}`JVpV+#&<0(>(1@o&Pq?U8azsQ3f~He=l=BCU_N< zdy;2@M9E3b1stLOS7j8Jb$5NajtoiFUc~$4D!!}sFS@*)<-RuM+f#PeGW$U*=#%R7 z)yt)-)W?p!+#LADl z>zGfAWd!0)#_yiL%<*J@`Bnn&)w0T7quV^=cMnctNl<;lhu_M-KJsKJlKp)>k2mE{ zK3ppt_Z74_$VJtRYSc^prr}*wH7GQ9eSeATn*}G)G}b<{B_VWOMsfZsleCn+RBJM^ zlwkVjfq8F^ipCALaIyM`yP1VDu~{UcIa!E78Yj*+q(3;4CI5lKoD7;4Bfb0x|b{H{xM$F_1^G_ z$l%sbbO2@V1>Etv7F89ib_*@8;Xp)v2Zg4F@;lk*>Qr=e(Ky4;gCz?)ul`^;3K|ZU zP#xZ2nvK!z5e6M4+E@Qr6j!NIUJkdBJRS&W4%1xz|0sna-ytA-m4@a*%aHp~V&z}g zAJ!lf%@vLgyQG8z?ycVli><|8E9YgcO``peb=hv-P4M70jK(@&!l2`7#chxKKohq< zBlQ}McmJgK+uyKWZiGXW0aszkZ#72jTypO8CEu;t6HJso~$=L)u_~ z1nB%wegKN=wU5jK&3B9BGG)QrQ)5%ID_m zhwdG5*DV)%KfRb7#-U>|niZ@XnB5n;j-y8<8dFrib6{lVTD0r^-wXU9V9HE>LixjA z?~G*cmlaa~Q4Q?_%sYSjJiBBJtQP+|fr4e28qb6GC63VM_YXsU6@U4XRC9Fa9c^VS z!xMvyn~AgE!C=LoNAss;6m_!8SHk4vM_^us(|C7U?FU=Yx(=Je#foRH2q@?p-U>VTtpWa{8~!|kR=s?} zbXFbJjEudD2b)&j)(``ty^Rj%B0fS~;@gU@B#y%4Khcp@-2oyt=NoTS*Yl8Q3sSaU zew+4OcLh!X9cFzJz0y_36Da1qhP}H3r>;HC?aO@Fy_Pc{6B9&&c@^F5hIk04n4lK( zjl({sA&qqc?1b5_ZwM7y8c1_Ndmo-T0+7$B{xi*bPDTwfk$!GF(gw7X$J5-9_v%U) zF4NyNRCSoM68F}Cy{=ar$x$zVd7o~g`_TM}(U5tznvve-Eelzrp>?=*3pm&!yGNHD zr;?eV`d2b`a`SokV=(adYAxa(N4hQG@qt~kYY~U8=wo2StNE7mHF+J-2ca~}BVvEe zyz=FDST4i}WC?D|>bgApGWH{*%cURBfLzn3``E{zfIHZZ6?ww*-qF5n63C8?S0#o1 zXWN{=%E5*2=+WE?w@>t~E5y9|Q8!*SrQNpY*#z)#2HQ6k3Hqr|!MS>yQtgF)`$x~Q z&r?d~&WDXb_OZ$OBsm!LcVu=dLz4c{Ou%O2>~ts=74|LmNlb;E@<&tauM!j0znGRh zyzcCN$<#fm6s@V6$*L`>6k-{i7+q2s24Wcoq`l^5K_ALA0EvFSO?c)^;(hY2*ch8; zM0mSzJ@LFe9dcy^=N`)insJ+_GTU}JTTf*|D(miXUA_10wq&kF^C0-iG+GWoCWUuI`KBgvlror8Q z+|%#ty}oXZ?m8y&dP+ZcN(Kf#%{WXWp5}o54QRi*^+ffq&yMn+$MB^)K)E`1th0kC zT&`7!$;YQdFUhS4^#8Z!^*9G+)0~&Jsv(0YVy7a`fmU> zs?dK+>_p>I2@C(n^8SBAOb7|`{ZBq;wV`p!QZre|npxinsW53DI`-=HNO;a)SBV*i z?4Ot6DFLlUD@H3E+5k6D@Ff15cgh`k#SK5m(CKJsg5U(UBOD{ietsbEN>{*=ZVXg6 zMOY$>o79xjre2^tRz(>2eVi_ta9>u0G#46p<0qzG3Ur1~zvJAHNn56^)7_&!u{ z2OYEN=#=15=Gx=y1Yz}qG-A0_mrMhHUDl`i=UPjj7UKixlX^?kD2!X8mDhStH+oLp z7_=Y6z6VFqc9=@u`dq;c+^C= zZiR?~z}m_qTw_j8?{yJ&)CFj`Iwi(H_OMa8TXy*o|A{70NziA!48?C+fvZv9(AHYL zn$&oX1M<(hIh?eUtW)3agKOVM$Hxh4N0&S(tB;2JA?09y`HFzG5;}F$v|ysd3f^Wk z3#ezX32So^CwKjMrH@Z+UE?A&H5I%UBJm@W8p&mu6mmPvrWfPKbRK);11Ys=g?lKtt#ofp8Hl10 ztHh`%P%y?5S#()tMV8fFlV%GZR#SfuUvA=*7>7wk8i`H*c`+~b#~1A=FFHh$a%*w2aMz!h znviOmjl%rZjUM1z@+q@dVQ$O7eCbg$X}ksdYbCZ$1_qLBD}1KRBOe`%==*RJ=LztYra(SI{*ZRpRO z>`Xim^hGoSJG4`OvqIoR+y}~xj#>XRHvA=ZS_t%~+CNmS4hn~iAadHJDuRe$%U@R! z(Jn(COm1SWB-%$ZAYZc07KTgrDk4<3ElHABT-ClP=Gzk2jFn~^T<+GVuMBpLV--Gf zwYP+wL^Mk-joJK;D3%i4Om5zG=WNTcOH z`T6smZZjcQU1i=r78W86@k#1i5X5-)g6)B=@#t(XneuMS4vT6(iyMwg$CI|FG(tuN zl&X9dK;)Ea*)R!{u#or8BT_uJ54ZOfJmzNjS>+OQt4;adPshH@rU~TX6tOPoeca~Y zhXl#=V1uS{srHwnB;fljAS~jiTl2!|X!ZzmxZg>IRndX(Xv z7U?=#Djx!%E;W3S`;D^9Q35wIC7WAe`^S$eXd@^fN$Q@0EEdy*>cu1(q;@j}MseWz z0vnzWD}<`*$=@Yq66F3ldaEO(B2OMWPprb&tMRe|r2GQgU}Eo6_of@iyfUnmt6_-Bx-}Aw_jR-5Gl;6RSqe z4gt+GmL(U(JH^4}C5&FZ733Vx;FO*D#E^Q{UdQ5!(XoYh@8p!Gpc)%NgvY0@0d3gZ z)sXI=7)O&`Yn5MRkbj$)Jj3D9pI@N-^aAOMP^_aYbDq_0$a4mifIOwSNTfZ7Sfv8! z(1|e=-Y`iwb&s)LpWIMkGW2}7x!@@pK}MUapcD@Z51}yLvv3N0IImqqS8UBP8*^;_ z6zL?JZjv}C%yNYixmysl){cqmnKq^t-C(_^C1VnF)#mYYcp(t`yZlynhX!7>Qj+-B- zWR&9E^d9Na%zXvcFeXL&z~HNWxr()n@o)T8B~hpOLvL4jO694*DzWc}c@MB%v(%hF zBfqN7dJw{@{O{xxD9F{bN`9UUSx0E3x3}A~BcKsHa?7*t$t8xd>L$#9I@YH@dQ?4$ z6FvOI?FNgz`bKi^)}#m%#sN4p5ZA7iM|w9}X4U@~Z?BML5s_ZhVpQS(!n7V7p^Kvm z3;)MwiX<6-3NT2~#e{`L{|A~iR~^FI;t_d`;5^ng+EW)D*&PxP*?fVT{N{To*Gpb& zAo?(+atyQN$B^L9K2U~(u68fDj*)-7Ho``9q>M0b&~X*9frv(IKnR~7H`nr?*7`jp z2_AG3y*sAmD<0D!(e!3A5AgnL-lvVS=NtH%bkO5WoENz`-DZz$a5ZSG=Xo<`%2;uVIHj^4mIW|)G?1nEngia#_erJwOMxD z6VBX-F>a`s_sOhNP--jTY^3t#`aZd&Cy#uvJ>cv%Mmnec2wI!Hn5Hyhxkm#$>hlEl zF8B%#aZ?%EJlSPVv^)EnmO0Ft+@_o7`Ia5DPe^{&8?xDb00|zrWEv8}t+H4BgjXcS zHw}0}ZZ%PL4QoN z>6OGyJs;{V5WuzMCV`;C4blR5DCZH4Ht-=cp)E7;DKT!kM~bM=IMJ=X5^R{0+TH3s z)V~q3ABX;~T4>m^fXy5^uGn7rw5@2Z99lg;&Z5qXRxD9nmKY=Mpm>)oPGJixizhwT zcPS|AzKFFl!`gtQ`ZeAO;?R|brHRHZaQMF!|6LH;rKw`;0?uWcF|{kjoZl^CyY#M; zb5d%ku&}eGj5*M-x_kSY*NfTwTF!ZCfvwhupsy|>xwCm8#|j~V)v-NsKy%1>DzVZ2 zd)FtitCA5|4S#huQOOTOg`hn$0HT{b&m{1GZm}@&E0$9uKHnUeAQoV$8XN}0tQc@o z08MN*Wm38VNR2)K;8w-MPCdCyr$$1C>5{IGq|DiU2lz_H#$? z)ZT+b;5i-uzw|wT&`)R)-D{P3-fU;m@(K)Xw5lF%!3&7xU z+h_3+yHiBNbkpalJ3q{Tdi)>{X4%=S?6KX{`-2$5wrz*NeK+FN4}Jr{1F)ZEzx|y9 zL;mbE$)<_Qwq5bP0}vCaVC~FETx1$X@TnMrU>M4kbuWW}`HLaBo~pqvySc8x)s(N! zCFqasa|}{KJAt{}9)paKG~VJ1ANQy9Zi0U$7ft;`hIfk-Ie2%A8?oHMwA*o=*=(f1 z+!k4#^(NLc902DU&zZ!MzUsQq)d!w11FuTx@Im2iUT5BC5FZy{2lUkd8WRVLo8$#> z-ua#xZOBqV+NdEu&G4W@LS2f?BSP=av~#Rhf@ImLjn`xJnUQD^%u$b_$_}h)Sj4sZ zoytN=w@BxNv031$8Qgf=?aM`)JVS8qiSn#LJIgc-EiwAyvbQ!F7BXdl0m&s)29)WR zA*Vqgf(F!C8s38e*|+z%Je1=rE0W z>ZJs`jU3EwfzcqQ0qOv|F+^@R)L~!449@JU0N^dC^{<{qR73n_&I0;!%{pzT>inf| zo`&h|Yl9brrSEV3`!~^pm+v7H*=_#0em=Ya(47%~`x5<6Ei=g#!&AqC*@=S>y;xs6 zp+3#`Y}-Cl(o7JqE5t%*=i)Zi0}Mw6w-Z_cJ%G}Gy^}Zzq(yEb+BXo=F7kO}FY+rb zG{}B|4h~=j*b7I8>^r~v(pl;`V#j&?pZ(x^HbaN6yB1F!=H zUl{ZqbgqbrLY{=I5RQbZkKTZ4ZXTX(VcLuvZ@|$JwbHte2$K@X{A%_O8l++F!maP1 zK$I7`TTH(jP(KTvN(6Q)fO@6=>~^k3|I%)P2R(oM>lE^=g=#UF@t1R1yLD4vtn^MA zwxgmxzq0{2ZkO7-sBxbP+NKM)uLb#GzvR|IKWzL=$oe*_(?9F1HaLrz7ho&qY^L8C z2O}fE!=OCYMp5jS3u7xX8?h^(&DxA8{=Mi}bu`9zZ8I)#P|un>r$@T`swlG2Z&`J6 z^zGje-QR8gS-2!dzusnKkG}c*w)ag2E+1wyo)j|BzqB3^dDc8hQP{BqFg^oazOVVD zF6Cc38GKvxyQZz-6V#v%lI$)XV5eW~o!}(DafxOXpYW+^yvGn{k7;zE}qBZ6VoRk~_Wdv*V`~ zU1@{dw)U*1S2Y`KDGlUXqg#d`I_2#2Nxyy{pY)5!rBjUM!y#O=OFr(2?DiyUZHc8o zMuBxzfx7j+Ia)>q&*1nRxS%NrC9$-V%+tmk&v{r1bF zp3Yh1b=kugwk;#5^_jNibU+6eYxUwpaSXsn4SMDcqEO48BTo^ThgnY7(8i7YnnTH= zc@w23GONtWOC-;VmlvsKi|#>@vPu}I+*)K=zs4UUOqr>6IzXP%tpQLAz>ZMU_vX&B zMvqamt?}#Ic7C(0B(MMy@3yYDJY#3Y`*me0Fq8HsC5AItpshxKbHax_vzR1Sxq5WO zx*yEM+O*4*BCpC}=Q$nu6(v(By2O=L_}sdlIjW;~`!InxlM8qE%gC?Y9ueOiHO5U= zR&FqehZTbNh{MXA4yH(<@WE39(3tXxb6?l@Z4bVn$n2!-#7F_99VqcF>>Y-U{LaPX zb0Gd%pGotQF>GX?>=u%{`Ozle>q|AG&z0P)l0U-Cwd<((#2;}%W4*=ImG_qC zn;bZ5i?5&lssZG`N~DmFRp*TCXUTs@-qo9&*ggzxaAmD5$}yid$?eW+R8MpgbSV5& zpVjygw5A4S2ZO?pBdYoEyD1CI%7C7NCip6=lM<6HrI}4jBn3vIXm>u1i>C(NfDYMu zUd>sto^0uu*IUx3r~yol!biTO4OyPQRI~N&vpmOD%X>)@s_cUHjt#4nl2803yr4o* zJb!5oz*jCxpho^1*jHSBRF}zx^}_&K+_eMbcc@Hl24#YK5J2C3>m{|t$LaZzFIvW4 zEiD%N+hn`PMLv?=0y+h~))w)G65^Dzg&Mvw9NzWp+l0AQV2kn;+;tr`NjG%XYSQ1f zp3N@B6UDDk8dPF=^wM7MW6K+IZ@_emX@BK zr|xmQL`w6tLcvI3qsLhuN?Nn)Ajeu=qm9d=Ngf=>fYN-CaVap$ae=>Di(A-Eo*Pey z$Qn8?{zrE#o{W61-V=jBpzJF{D7pYOoH*G56+~-$3WU?mdg4A3&w8#>H!uu?BkNKz zvmU23(o)1FQ>hCzN+Uf~KaF{dhEHZSVnL9o#RW$7!P$L~d#X;y_Mw-nd%9kj@QU|> z{#%^2wx`RDIt+uie=GnCsU2^=1lw4ou+aa z#JRM;&rlfaJN~fWt=XyCOTcm%#l4OBCy^!A&Fv6rzT+$<9~7b=!fZ{;{Zf5PRzV96 z1{{T*oN0f72LEVT0OfwZn`tU<5h^UcyJ<(WvwtOPt;Sdd_*-I;wmXroDU10KNydCl#{{4sKLyNW}gktOJ_Ctrt3#@~WG)uwteCI&PZQk@&-sqM`nYzcx;J zNwIbc>I&DRKa9QO`M@}Rp_j?BDUb}+G5oj+Kx^;yRaeo`$%@?O%(p&P9ahQaG`5a$ zrw}Nq;$kapq+RI>Ie7#xh+)-;Ys5JnJ;ffxjj4H5uizsu zS&!<&!USfo?%PwW0|P#O zCL*xhlgifPQ4%Itu4MUUyC8;^iLt&Go{3eCx7SLVuyT#0;HLo}SfG53OU47XNe(F( z74}$EiERJzuoE zn+U}-wuv7PNPF<|u=(JniRR=ojA^r--;$xmuH}b}=SnL*cTE|mKw@u`jh;e{@y^5h znw(uxPwD6d(dhx7nFqz1iJCAcE_V@F`im(v*bGz)as*_AB?L-u8c`V{d%{Mv|6H7Q zI%z3@nQ-!S(QSeQr1#i9La-XRi(%mlKehvS$e|$e);tjr<>4llHM%gTF{l(O%+@UI zAmCJ*5oW6-`(dPWvt3}Bxu)qC87yt=gNI98&a2jDxNH$XFvn^CGBkv%iw?VqtIHPK zcV9SVHUl7t@3N1~tL5rCzVIrBsm&V(tlZFjgoq!jLa28;N?`FXxw~p+{q}{eNL3-6 zyP&lanA+XvfQjJ55}3&ocbER-w396_PuFhE-tju{5PWo&z{#Q1V2e|uZgV+oO2+QvAa%S!sZf@;?xj0<&5V zsP)w|%&B3rzU}J3!Rx3;Rs?3-FN*>51wBuT!k#*Pyj%v2mQ0NA1J@TTX!{8H@2hg0 z4kk$+zx93kikzWV7Ff7p0&;zowgZGb6>7NgSpQuTMQ;a8xsf9805$_m&INlf&9OC1 zjJa&kYq0#d2QQoZnk;!Y1BMBiVDbo-wq~x}SFLo*qNGR;t&Ps2<)*xq4%F6s%=sR& z(v1pJUhvtL(4rskv6PK-;xt2pmB-8!-yOGVs0Xa-(#z!_*K4SbLbFDv%UCo!frzTD z`QZCOF)-eQjWsvUDX^QoRpVt|?14l7Gb^mz%1pzgEb;b(3EQz#%z;CS4t((x-EMvH zaL=qP=)4OJX4|qJO$lWO-7s8%xZQIFwmi2No%dv0MRM*AO>$Pob{Cy*xYrjUST!p$ zqA=xKf^7)~h@D4ng1-lW|EVO?5Q7u0;`k_AtnfV_HICWVZ{30v4HtU z8%d%5rc`#F-GWHUj-)!!SGYRREIA8FxSEBmjRZdjr8cOL@=^x{6T=S$S^w)Df8VE2 zepxc}k9@Dphcz)MiGxgcvlJuB?QkcmX9WQU$5fD$Wh6aE(rrPX?E1-B2g!Ri&oY_9 zme)oDpYn|cKAOMn{g{u-M?{IsH&uojkJfF2`TA1wU`%KDTFC2Eh09@kc{8^%h#cms zT7ts>>fuUyF`B!>_pTMj%H=J|Sx#T?Vwuw7Xs7;~=%`8@E!)K)+&YI!KhgKcx;fkz z9=%=4XKtgX^yz@BF8VWE@hDW^uIvJX zKKw_E-loP``Q5Ea>NDhwIegRcsOLKPw z9m$C`2^qSl6aC2)wqig^D5bl5I_0@TqosaTKu$?%`$p`@!SaRpy$el6XkdJTly&F$ zdlC(9I5wN8JUW}K4-uFmIOG?QeE>y_422O4Tan}Cl+hGBFl`lYk%aUqGKk@)s|kpJ zz25QThR%>^<&7ybfiq4bdBoB*-+{zjvS`x1Bpo;07umEKemXndb#HX09-xXDd_Z%r znQVDcb`%-aYOF_&zh0$n+PjdMt}J#`nb^4jMi}F}`|iJ7Kg5kkg@y`8l9x-(9e-V~psWbQSbp#?WA)petT$@K}{0p}_x@WEieU*cot_%FoZb4<>WpBI6CAz}stfgKq2t8u(GpMl{BcKD1~n}u zUr9wgN6w0%%%+Z(8d@X-a_BWAS+nmguQK{&*!T|xT>2Ll)nM<2mcsK)H|Yy-R#*+wAR%4oi4S(p2CXH3$uZHy=uIT1$?ELQ~+KEe)r^vCUO@!rY44l ztVa5zjmwKEZM_rx8OV$Yr$V9#wyI0PlWNv|2wdlgCZS2d7`rh)lrN+4K@N6hV~R(* zuVNN~{D0paZD-hn26Uq3i6(vRS^cDr8cz7?mO>P#pghs2{vx|cq?5PCYWKY}Bnw}B z^hx8&wV*}qy-qyi?Hz*vX)_Ypgw;Qqd+t_^QlqvQBMwn|JhR5Lb z&U7mhoN?Xzce_ckrE#krCipSuNgRgyJB~&BFZvY*Z#f|2RYdX<)-H)AHT-h?7o)%1 z^)@(>Qsoc~`Q$&C@0&4vwEkSWjj1(H;yElLalD$CNFrO{4bgXL0^+!@hkYj+uf9vSrpSLh zqNSsruG=&r0F{^cT;cC8nxYpc>^DYu6VO^lgf?ML+_>t?z3EM z3`u6WosnRi5nZGU`!N6g$$wnf|KxjAdlOlrcIDaAP20ABiP1U;^X6AI|_8mFu zX69xkRgjH`|4(N^r;qplWdGLLj2K*Y*dY5D6Ey6na&a9Z@qO2a&0^l;(r)xm!GU{4 z-j6n`dWJU&Sp&Iw@ynQ5-}m|E%DBF_i@t*q!WGS?b(mCEw6PYSceVyZ*y(ve^a>6F zXuhTn+#^Tb(@oCy=RgkAvB#?*+maD~R#%zIe_m6Fg{Gr_OO)cY63oNV%JSy)_h45u zG3K{cj+5nE-`l@(eEVyPF;pXdWCK{?Uz;t+jDG#;QD7fp}VJ(Fa_ z{&}dXr)&O~6%Xi9*`WC3v@zI=+RFpF&*=`mHusPjm4+W1qC!TEdP3(G zH%*R+;gOTP&_R50$gAN&al!^euxOPa1zauP7>qb=8x*H7h(mk{9oUqz@}U(%h*d3|}t zV01UAq~j}se?VA_{xT2ICsrnQ@=_`W0Yx|s0L=mjp^{9No6ApLgy5FnXDsl9c1|d@ zj{yjLy+z@WccODbCAf2zeo=EM<(v1M1z}D~_xn3XY!Sf|9QJdAi0W(TG7d zokbNT`5|v z$=k3}F_@jj=JJI-I+Sw;FxXbqy$R|(r{>vjo4!rYL#VZsz+a|tNuSKh?*rjMZ=~Ts zPt{4uUwEpMSe8PY%hPy2@(yNPJOIM<-w}&90F}whysuua|G1%e51WJHGkF7T-g`a_ zG9EJY6l**qZd4{aqi!2)S7UDF*8zaEJm#Lrd{G|g4q+T3;3WTg2NsIx1K}HZ9)~41 zi?>V-E{MR%tKI?KkE3HTppPGZ_5vk}d!`Z4Te$$JcU=jb0*>ba7}>d+T%r=U(-qpg zB@@q=KzYsw9}uy69`A<^?Jh6nh``y;_C%2*YAlA|eltUpHpu>dAtji}D6Kh7a-6!T}ShfxdDcYfleMwEN{8h@A6XTL*jdZ zaA=lT%e%nS$(H;V&?7$Cf5Kv#8Le)dxm=!k9SB5y6o@QP7Uz$o*qScL#}u5RKK*34rjI&sLVpHPlq(SZ#HJ4+X|kyLC+88l_#?I z$CYl+Co{mDXiY%KU0F1*Ea*(rEM$OO_KD#3_TMVVDKT~j_Uba33RG_`c^yKF0u;9?PK7Nf21FN{jT&woPcg_;q=|lht%Fz z1`xpG3*Bcpj;?IuV3L=0KIC|!WR+kpX*qKDa6$3iUcYPsBIM?5M^lsWt`9%Bs&o=| zy{ip!+iQh$`mZGlmYM5QNFG1YJ}+u@9`q*7t-~Iu^8KFH78kP#szhC%@5f}95&lkb zb{_*m#{K-*RHB%efx>iesWiWB?{uKCY@sB#(@mzlMAYG%Qa&91Xn79aqFe;3H4})s zIVG;MJzrf;=F0@#8c_38-Wu+2tZ}YuKHq>w;Q3HL2NfvYnW!HB!$CvvHRu6Czjk%( zOXEVVx%}h?VRY|VUFVE6MtddLZiOidS!3}YLU%Pyx?Sc68N;k3u9eVxC&OJ<Mwm zvobGHjExp0pJZMZkM*=azQGJ36PX8;ySW8j+n-#xLQ+|duQg7tRP;7o|HJBnSdOnW zWEq(pYA~b32SU7WB;__$2Z=)-WKyjqtKjjPrv#aQH%AEMo_0us&^2Tho%Rk^Ha&ct z@=XH{u8Cw_yV+yt%MNHkQV;zC4>tirl1ug8j?^K`^SQF4m8l*tFGoYZf&`GDsUg8T zG6OxC&5)Dv=o_q=UzcguM9%<_QBY*aoQGzJKkOd=erHX*bL}VcKzvQ@J$X0Uvs-X* zU*-SlhE5$ zZx(XDGBzarK+y7+4Agw$dk3u8RkEqrTr1-MG}}Hbjrm;^0`8;vxrso)_KX3Db2~kg zC=9Rr^ycG=O_tFLwCMVD%hY(6>nw{K)bsdna_YgL$Juy13E|TAc!Krpp((w%0iVgE z(5S8jZRKCbL<@bMgXBcR%fJ&^jsW5K^%n{t3A}9duim?i>FqDta-Z*|Azt7@c3= zNsW6i>iSkD378Wla|_CvS`uocz@vxw{(4BrF$_Mqqs<0B_l+zG%{(>FwdLR-L|s`4 zSVMAbWp4L>EeRpcr4#1VGJ8A(Q}Qzq`T`faxDQBKnIL=)bU_gW?ZA_=pUj3EN^ZC4 zi6yTC$a3{gOq4ogxOQSdw)v>H3%}f#bvio_gbL7Wv=NX9ii9UEfYWsgebA`oPEU3|=aZ!Xbi22Z7q4SH2 z{9kaDiO;HTONvC@_Xb4bTA>Oj;i}pwo(`l~>;VoHZ=#FeQvMs%t$(Fep9@+8a<>&T zQ#%)KN zbVvP3{rsDnQ_AOD6pEtk1n(e9Lt5G1bd zZ+az#*OMOICJAZ;Tt4yC4S)U`QTIe`_}z5$8Kx6d^_Xz-I;8i$P*#n$DwqGZT&T!(Rk?nt^&iFaPtRh&OH(KQAb$P_W zel8{Wf%v7vSbj=mjvuwl57<(r?0UZ4iNo)_Te0-o^Na#Vz4`$&qD{9oP8x-LKk;z7FSfRW1-dIqG!((8#9xNAW1y!8w(K zB#0J!h$MW2k@n|MDNY4`69OY2bV$>>Ja%L=V_a}%yQHBuep00n{7#V>%X87U^&g19 zjEL5?6%l586K_&Wi2AL>JWEZJwKHh-+x_$d`19@ut~N^;PDHxW`{cZw`?Zg3qyG36 z%sRl-FX>)7wXe&b3fp43FsXG9?%ovbWIw)z<)ZWD$&wV`_*0PYQ1Er3#i*07G11}p z=&hc;xEAbDsgcb08E+XcA#uygh)x9g>`d^PxpukfZv)zXj_1gK?IOZ>o$ynHC3Yz^ zsm&x(CQ6GkqA<7h1DqrZFP{HPTKF>VhQdde_+ykHJWdh*U^@7V69YGg_qB?xyx2!# z;ox_X6w_~W?Ts(o9ClYJ(?s(o?pDF?-}0GEM;tE?4?10m`~5R%Z2x;0G3HI~s(tnI_upHm0l&Ksu7DPMb%ah>sOGP)A}hp--t99Le`%+QZo+}yUyqR{?3a?TspWgPX8(4Qr{81V986aq}Q3J;LiG6 zfqGOC!MWfI#z=*Q_RI(cJxpNC*}70nKS!B4We~A{=}6c0qE*{fy7L$X1^vFyGXBf@ z5uJ5s?qasB0N#B+%t`xSa`I_eRYRe#>Dt%qQbmUNGcn13T?e28Ri;uw$Ls;u54s#U z^c(Ri;D3TW$5vk{mo19%2Kz@TM87t2mK*+&lZr1YqV{1$Qb*)#iMuH81k^*b@(X;F4~=ju+%Y_*eeeje%nJ3qWWIgm z9b=WWW6q(72`5+0UQ=}enbRM;AH|#g9#*5?8*|>A#_@922fHy;3YXbDNK~D0G9FYY zn^_G%p=f`rMaNvV9@-wy-L*iJN4gPoZqsXJahzCZpB9C;s778d(G6W18J- zMQW#|1HAP)O*r*6TJQr%Q-vbn{@X2ZCW;88J+P1R?O%+_`7eTRpjV9{(aF-5U(MCP zz5?$ygPE+>Lepx$ML8JiDUlF=CYn#*XrW+!L&M`Cu&aB%MW$#e zdoUP6E(GE$j|$6=WRLn@Ae^!PDi7$I{6P-GuF5>cCkr2kot!|5iuwJGX*!(|bXi5I zocD(YIpwV111APrip95+eN_v;>lb9K3fe!_j1K2{??=*3a8x7PRm*$~&}hG$_*}nN z^4va2`sUiBTqb6$BP2QWU7#$QWPpTSXB6b8Dw_6sL1~^f=EX6e;FYpBlf?HFp?3^I5NWE zrV!)%Z=($Ih0CjE<7($^|3(yPPZUO=GJ;kN)HS#Gh`Qhg<8{*J4^&GAqxD9`rZFze z1YhM_8nv}#(W@hTL63=BLUr=)?S*A}ulg@0;`8%|>L%97DW&>VeKSJdf3)QOB*%{3 z7n9uUrR<3d!E23n@ojs)y~N$Uk1xcFtX-K*I<;Oz(PWcdhP+;_k5(!;P33PLyatI| z+4!wB=&l(E%MX5Z4P#fgCzvOZfA3{EAQ_eV8Oq>H+pl{g5cpp16+OA1;94r$`7+Pw z{L;;v-I>;*vWk%Ihvq`}uCZgIWv}~chX)bBAH@wF@kBNmC==X@q{BoiQ)SQ zjXU{B%}Bp*&&3yvyHGEU*4I420x}?LdP#y(UdesX+h8ZdI50NgmS3Tj;2|Iwl_ma{ z1%^xQg=Bxmf?k+Kyu^NX^F!&1yKdQMIlzIINQuLg=8Kegd%QtRI>nh4O7zG-CuVUA zSC)dtk5pEod`9u~g{>3auk5H@13g$%E+sC%X2jE1&0xdzu(ctu$lrhAwTV_q-AHu8OWzJ5* z5HjO%2?Shb3V+!B9wZ8ASvjy=Gt9`Rv8m*lSqR7t5K{e3I2UiMQXC@`2fOIx>iyV( zh;N2F->-E`+j;)R-yl!TAp^Cv3F3>1YTqi$u;M^bQb{836Iqh4Y0(;4mZ`-^6)}fi zkkO)T)rvUXbX~o%S)makxgKX-HA0GJr9gA&eNogVC zT8@i4W<3jLurW6*+zb*T43dmU3k-EsF}{rN01O6nelcXN>HqEx>G|svgB>!=Q@GiP z+Nkw3YZr-*8o%C*T9L9)#K2F|cf_WP+NlKEsiqF9hD7_*>8pXppZ{W{*Q*-7lVG$A z8H2A%reygYFP}ViUrWJ(P%miUL4Ba<>t@^0f8$G8IBYXcW{TCL4;Y5ROvYu_*TsjB zv6{QXq+PMp=7SOIb}zWfEDPgaut+#q^86Yq z70eIn&NgcUjJUn7r&_-1tdrb=+3jWv{W8tR?1Nb!|5m$(z)x%$s-&qpIJb_LGaV4q zLi|{F@QWK7PlV!?%E-jHa|jz}OYxI+KpwP#7I-I@cL*9i;f4EUoGWEP#+UvcjZhl2 z{^XuW-Z4bC^jei$K(&zfa?BC5cT+qW8}6cN5WmE0NJjO}Oh0O9sGGJ3XyzI1z|NbH z<=u`MyXAJwI2SXS9Bg6k;daCQf|{yeNI+l9;UO!~;a}Y2Hy5R?n-=Z*Q`MK0^9E*;fVSuEnjx4uZx>Ep~qyJ>-tt zEqg6!>A(iYvZUCy-Cv9x=sEfbC%*JnZpYRp95F;}b)UfzH4iKa1#)}z#+gz8^_~pl zyTO{EYYG&FQEn$ZOjds;a@sZZIha&sT3gS<8idEVP-;ei*7~c$A$qAi!)*x11siA{94$jH6DMhpl$HcU{hB+fu=={BI zT#@gP!%!x`K@twgl_z&S!W`p{xdGw;zh)i_Y75m1kWTYi5cy*;s$oz${VxWvD11-s$qdGawmly2iAZpL=JO)G~u|d&ik`Xsxo)~4E@I=K0PH$|?Wn@n0 zW(zl?rF^vpQxAP+XYOlyqNA)letn&x@S2sQBx{O*GHY~~cIxI6FY!VK5q5w6eEelH z9)SQ*@~++a$fQj)Ub;^MMznQ@M`11_%ZG+wq@$LBonFIhj*1|RZ`)c0cS$*5P*V~I zaeP8PGAV`ps{M|cxq9ji@vi+195WmRwNC$Q5arCc#Kjr|6=lH3MKb(!Nu}4;S$t(N zAY=TAZCuhyM_oPyQni&88FW<=b=1lp#}o$WdT?F&>|z7z3j8}LshRIzk#3j+GeS=7njB6f?oh=8e>z@bO`eITCcMr#*&!P@wEy#) zZ#NXY)J5U(u2bsAqLQ5{G>oj;jirA_L2awQy{LsE+V5l;9gG-5d>k7O|63RTM%bFb zd$n%lOSN!!s?9JDCjBM1?&haIC~R*)fN3IO6^&;;0fFhuXTu~rGcXHY2T-Giq^au7 z!)^$1X>B+-{1Z`Jv(}baZ~}nuvGnfkYw0>MlCrvL@Pzkj+vZ?DyA5d-cB)#Cdzd>W z2UQ}Xn=PW`J~Tr~x$cJfM|Y45Hq<+9npgERGge@v^zz^<<;vh@$6=f}0@`t;xmn>Uk)-=xpy1D1jIRsrv=hoFa%#yVjL^FslH<;oY6CKiSviHi%`@X+ zi`fMzIIrt02AX)pV$2JV|Dhwkq&#$4WDIE&b8|wHK~%Cz(#QBKC=O)T3a)7B`^zFVkK&YmN<3RGwMrFkOtNR*6zd*_?3`F;?3T@Kx5|-*}Gj3kcPi) z%VxB~re}N6?|Vwa4TTQmFHtB~QY&_22{(;$6fhfWi&|dzHQc^sTNJw_zA6(c+x>$i z)%}CNHmmT%V(Z%@`Bg^u1p~83O@}%g>cxu?v2od%!8x;}8K6|LmkEGq+`SkrqXQ79v5LYUu2m-*KM}N&^MO zMT~y2IV1mtreKgY;C@eoGaJmIKV;$i-BP!A21g;^_rgYYjqec%*7--cAwlB6v_f%d zebRk8ROLhrmjb*orl&+hDAix3L7c*E@V9AaY5zIaZL4*@4($<9bh@|_w zc^Lt9&Vx~0d;&2##N;xaZ^W}6v*)>tB~C{hZ}gDHt{!cun*d%%GbAF0V8WG4_A`1d zf|-%zVOGIIJQ)A*7Tg@}RZ;5!FGShS(>^IhRn!22n=Oh6PQP zJ<`I@9sA5dL%K#HhHT*-6>CES@R*-t^PwG(R!yQmDUF8%?3}eYaTtiN#YX>Z6v2SC zKpm;2IszHZodNhlzk=uQfpptBh<(q`PydU)|%5RZ)KjHlIp~7K0N))8)8lDsyaw za4-8C)31R2yxq&$A3BgfOg!=dNyk+-&JHoapF?EkOQl^;F z4Xl7X--~{WD*Ig&>TXFRMR=}l^vZRNa0Y~MBug1- zxX$v}8F|QvWA5)_cn$s7mF-7RkAL#aI5;HhTZH&R>Me-Zp&E3#k5n|Vi9|gK@m#E# zTzlZODnadJ&`ba@sjv1N7C*GDq@D+A2RtjaV8-^qnEve7obf*)yKxu1lUSqDQNVTx zh=fNoZp<13qo-IA6a%!G(Dv3uV+Gy}V6~_83PWnqk>@vtW9g(UDh?HAr8(ee{Xa!H z+|-?}zF(dzNN_#4nSK(t#F%93|KTwkK3iwUNnK!26il(D;pC5CT^JHvlg_(g3u^4# zNwy)zDzE1J8LVGi#>sdcAg`vv=56I%(!li#!e+c4sCUj{1v#pfY3`g~Qz|ZN!vr>1 zFoWY$=f6Uj96IEaNPSbWuaY0*Vxv^^WYE^9!mBiBwOnS(P?pa4Gh!vt$7l5foS0uF z@twqVM8kF}QdAW;RmjGZbgY75v((XBEJJqD-M^#n`L4m(OJpS|IeI4lF6Un0fk@W{ z*9Ois6%7RvHFeD7f5%2{4%Yw1Mu3x=Dtd5k&o;Es%OWQ$ObjHdoRYAJh(Gb1I7~e8 zA_}Z-haU_nDYb+M4K0$yNtaA5EJ>WKs3(Oq>iBIA8?V#hl+TsdrJh%9*8?wCE+|H) zZ?xMKvjt+00wsVGz*T>V0zbue`Hc+DqcuPw*|}=xz*4VWVVNHaV-{GObUAo8WjZ1u z$UC|JC*3DXlPaw{v$%4FCGCu$ednI+hOIqAW{WyPQe~UYv{UsvpYb8vU*g;=z^o>v zj@hwkh;C`-b+nFjkKlwq)2CxAVr( z5h;-_tuM3`mct`1aYx=}YSNK37-}?fc!+@O62KqZ=}!nTuijuH|Je?g%(30Zc}$ww zgA0UI)*4@B%S8j;a&v?>qP5d+kUJWU{~Wik(cB*2z#rm(yhVJyNX<%`mGD@Rf_H{Y z_$6I(T(GhCA>@7sr8tC?5gQ#RiKaK{8c4sta)ggXn%$SK6DO0ZoGx6GaUxzgHl6o7 zw81AOT;h&_f`^?YA0K=yIT)THG(T|C_IiYf|L{9_wWu19AUT4=)@0M|{4GL^ZiOun ze#zX#0a}SA%g9Q;#1>M8ti{s_m<@i*r^T zzkqClZvvqGBw}lmYtuk_KvL|OrqFxF4#%F##;0MMnmr>I)qtO@(zg!fTr{u=Si#7KB_edbT^<2iwTGJ>XCYLKAkD>D;3REB!C zQ?9kWxMF>vnY+W(RIQPWK~V?-LJp7!@{c<2 z0mm5|RgXeWB{zx3Y|@=AGKCsu)o9nsG{Un7O!u-5jk)OeAHRo#PSl-1BDDR+4*gMx zLZ<=THBVLBOWYG|U3RumifCTaIwi!I$FV@vD0(Lf&uY0!8VvC`LMrMg9vE(FF*aSXSp86-lr7s# zo)x_{^_99z5{s6oTxh!+-TK)hSNy1uTW*$kdH7d0!ZNCh?KUrB;M+?Ngyep_rBMkG zm1+tnTx>_Xgd{6!wT;Ha>%}=w#z^%C+(8bKAWIA{JFwnc0u4>w(P$6M7K{a3t^)iU zRbT!RL?nA6EioRD9cEhoyD;sZ_8;q7TtX_^UWHTnKSqGr*S^VQB~IRq>|W0l4V- z)QN$Lo{Mazi!472u9Ws5AW=&&f>*Ea%jWt9N_87HS%ASu>Dr=-ZetZv$-UnfgPP%c zHrT1tt$TF^{*2&8DqrVsHK&mQl?hB~vRD|%8Sf3;f#d*- zQXy36r73aet}#i4JD0ydAHn*+z zE5~W4DW@E#sr7{N?>wCxuc?3nrgg=`i>}qRT!yaI+4SDayDNohX|;>d7KYUxhWAXT zJRpOy!0JWg`^nu#GgTL1cj9U>Pc!fE62P6D&g^Ebz4Pg+(^+sDm?fk)k~_7(c$jrQ zwjFb4JJ~&nb37pe{b%~*@us&yUOSR)k2{G0XC-~}9$ zvA!hBdW=1DxZynT4Swml^RMBDD<_k`FOcjgm^kI~+;~ze9&Iw+F{tv4buK=T{z|4* zJt{Ju=GA6(9I-WU)t*#tXwI7%56MJDzomK_Me)J(cJ*z~=))kH`4dT<-mfE~bt5nm z=ZT)oI&eO)-oM1E~zX{Q-FgCO?1HAFMG;GG5Wf}~U3iyVI zvrE;;p)yy`$0yRtk(O=-ouyrC>36Uex)_81ZL=6fMf0 z1#dDJ4pr>*&G__9J zs4c=i2DY7RU8>>6JSBHzx6vd0z4Do%Ez{$;a>FiGm3;cpw4IXQntSjwf;$c=%*wHe z15Gr}DBM`{oZ`C`Ewx~95AU9PNLY}VCRLD%$6g}83r)`_Gq5Y6wB{=X4t&mT>;tS0 zD-**(F_eh+0;_iGEXzw_=2C;_F2PqdjyD;rw^{(`cT?w%XKklSp8NTjjQi8(=<3ZS z-Eoqf${-!eTI};=C+6v6q(A%Dl65)SomKI`Y@QH!iX`~;+9_ImPDD?np2r9Ir5LvF zQ0x{@O7KbhDAFA&z3XOkfz=ZY9{g98YJa>2OF();!5NfQdsw90`Gl~Fj7?*_l)L6) ziozX#22~8#g^r*(5bvMKf1UZxB#kRStRGNOV@Fm-B)N~Gf6Jw>$1LA-9sWGlE54j3ncmm_~5GGV4;Z=NAQw8ujCnm6jnW@I5C*oI^rNtaGl$e4$6N(hZJvvZUwH z;?8bW0D!7cYb6`6K<%(mcbAq`L|EF;UWJV?tPRUKL)J7`VMK9jyza+0@rhbjeF3XB zBdi`XpUM0dOqTRAzz*kmLS&x_=xXey#wwOzdxe~-vp0H8+XuKE%*U)gw%2AES3S`? z?*ot9dhOgA2dED^FVlf#P|2;sX>4V>G^Z0$T+z z)Hc7XodDF89#+2>MgCg&Qpkrr6fh=n@{)}2Lzh|BN?-Hk<@b?nA(?hnHk_LOR4aT= zdJqIVp3JfuZ#(BEjZnM|+c?qHKl;dg9!>T5Tgh**NBp=zzs56NgicTh$X-jp3#(S&7S+@r)#P#=Ov zeQYAz?xs+?{mj+ZyuX2!yrShxkI$2)cLjIc(LN1_A)t0IImKBk=+^;=xzLYFN1dA4 zvQPc7Mz-|WAw^boo;0^sCoJ4}*dN^YN1*AL5r1VWEAqymYP{RQf)#SHAJm@sXTjS` zBbRcCUVu!Wq7cv2BBIZfkG`2i-r28J+Ug-bk8k7A1hnTa#1=ls|6=PHBAtZEK` zdCd><4J2CGpzEFie;1j?o7-rOs%>RDlK`jF{gU_m6%gJAvEI}!2B_fygDV4W(H#~= zGhxDcbx`9^P*S;;!)O9HWpFML^tA`|R?pEVYDY;9wa{QM{6anPQV<_PxlK&Y|B7&+ zT&c_#5ag)_#*ozity_KtI?mL321s(?Qe>*SDgor*ccx~o#fqX(wIa=UHA#zZRQBVuu>eqn1wGe>Sy70pu}XB%GA@0X2_ zZorMv7{||{+!guQd4y;E1_I=b7b?hgQ56F`s zvzb1d-d@>gVc|+Y-tyRJ>xFqW9o6qKzh3CE`9us9NXF9jxiQRnCN5MwvOBsBh4>CAbMd0+>5k9U83#9g>l+J9?_j)QnyLvt(Mg`PYBZGE zIvZu5mf%hYW{s`%MnRGw4der(P)~I}+&LA2S@Bpm`_KFwE5hE@LVG@rywmFmwTy@H zijJSv2yR{JUP6<|wui&&H*{6keYPSG^z!Of4~(NR&6+)2R{?ZIAM_vT{7#h90em^? z6Tw|+f6z{KavX~pVuo3fdT$n;xEymG=!9O_A2%lKHH7VsY?n#Lj|G6~kfx?B+impg zUaXr(5?x2-01%8)CNv9kWB?l+2L$}3G1g66vcz=#!(SO+ILM9Y4X;Lf@{@zNkWTg6 z-^Kj&tgFa?o0cYTrptLq4|(gWO&{3ty2K}pqcuu+--wO6^~1U`|4yJPmps>kILKPM2Q+Im z>?@mptBem#Pb1XEv(Zl?+cDY_Fe$~H&g`qyQGr90V%6vs3`zV3zFRkw*%taAo%=UB zG3E!fWDkv0jpZVi9{VU87{02DXy3|BGpTS{dPcZ}{m3Y_F>g22gE2I-$kGvhYUI$^ zD@Llo`#YVRs2>8a#^t?KY?Ip%f89)kcP``<<28B$FkyFP7&=5xeaEd=9_Ujby;K^V z4WN=^N?eO>sl6xRG@H-WGS_liBu;d4o0pcS5g}|0beEW>IX8`+)V9rUmzYf$C=Sv8 z7#N`NW~TY16Z6@kZX_K1VvDsoU*fDw)VbwYWfpo}GARV63G&RFJ0OC0rfzx;Mp*0S zxCYBptQLis&1JVf6VYgh;Z)bin1FoKw?80em4qU${@tpZ-bGUB!+3Ld$D|M{StMkC zA5xW`3=SneJ-fgTw<}+pVsaGGG(YeGeUp)~YnT!kYzC=nL~rSJp*a^HOkmNBjW3t? z;2A=Buzv@<>&hmJZ1D1Z=uR$lnzz|XL7MDg;9~&Kr(&9P6*IfQr%AS=rfyR*5Hazp zeUZD!#F}qeEvh63vm7H9r*51FKi18W=XWgpx1We;Ua0B%{)vdsP4;;SE3tFm0|0(kBqgvuU%+^VI1v zmaZhywjS2#9n?Gkx_zVRGIw; zmsyc|hSuCL_KM+tN{DT^jMD=eBD}+CZ&1d&O2Q-moh(3jiM5vhTHW9v8)5!Y_6)D3 zbAd)cd;CK?ay0&3h#l)XDrgM^uTTw`a=N22l?t$`rs=pbW{$8hO>vr_Xw6acd!GVqxLO8*)KOaB$*C7 z@Zg8~#Mh$DYou2z;9wp=r)Q-4w6SC4mRW6TZ!W5F(m_|7K-E+N;ir`PaY&?EZv{$8 zcoD)QoyK}WkvJ6PN6e2Ea3ci~%B)fvx1%f<9#r*Q6m9a{JWP!v;9kV+1#w$ z#Q&b2J(#R%H!qCh_ve;Q&8LB)bG3h&^!oSgPvSVZi0yoQ))=R-4kQQV<~L6>v*B>v zInr%Vi3eL(fSFr2Pt|}ZwF9RoCqXn7Bbt^>X%^;u`%HMeWHG56>xngK37Z3Ir+zRm zb?aYCPD7Jw*sb54E6sm>gWx&218HP+`))a+fby6>nb+fDQr(f}rwt2!DlH5`-D@UW z+h`D3&w(l~Th%2%EtzJk_9*f|j1@PwkRcfbgJ*N>*+J^5v-LdAv^xV#_d zO3ki3k&mZXN`{C=(KJIDKgJVJ%~1VaLNm5I)+EQYoFsfJsfAZS54)$zgbBZww$0~7lD3FH7J;1)R@Vk3lm*X2TrOR!PW-Vir9&- z-C+)p1f0@s7m$}kV5~VS{1=5`Mv>H{+r{%6jT*Q4i2UaFg|g@fT!u;5c>LGkn-Lm$ zSiDaG_8KP{C@+z_o|6rNpyf5Q5Nq8B2bKnJLD39eq4X=%xZKyRbHlQFa({EsNP+6C zF~-!?@Q{#Oi`$!_%&q7KFcz>Pc{2W&P0^owKEqRU-g_4eF@lG3OW6)XO$^;I;<(tMruA+*gwdvpo*v{*n2m!*V#k0x8kvYe}r3Yn%UFrpzio3j4xy5u1wU-vTg!?{6$Y!o;rw-qV0IhlKoawx zDJPb+gNn_CS^Sq+XIJriSSo2Abh-z*nB9}O;?86t@? z0V&PH8UgoPKW3tTH7}=4ZJr1{wl#mz%(x)>{AkrF*1Q=qhpJtKNcxpx zcbIW7Tx!-r-MhKWkn>!>G)9P@e?^M79e<9&z@FK1LCI=!R?eO8Q-1t*9_>v>RuuAW z2MSOTsb?9^YXW8AIN?FD*MYUI@j^v2cIHN;v+QR4wm}Hv%IyB2N^<_yk~QrUZ(p668!R4UDr>6M$WI+IjgnaL|t-yYBL^|{FaH~u|EKZhV?Z^uGUk%&1& z4}Z^;<)yJ~+;?l-7ypq89AP;R&k8||5CE)2LQ3=WTQj3jz+9Q|Y691Loc5&|_S&=X znD^Gm$;Qw?)D!#~T8?$#&KXQ#`kC?l)aYl{VnEd-6xy*pFGHhY@~5VW2T%=5v#08% zF|zT;HAUt4Y~BdtjV}idpi~tU*Ln2Shu)c!y4Ofcfi2vH5FMZixur=Y$%=choD4i_ zyk@R7!z%Pv10?R4=)A?>I}zG>rawj|no4kTHs*~kA3PyXxGEz?RMhjjBy>&anL-aQ zpsy&rHgi)}nmQvu`~Z`MkRYqCN0e!R`sua%>=f^Vz8X1N8a{$FnAqsZT?w~3&9n+; z>P|~MQaJK0`R7Wlq}7SPVNenm84?i8qOU)g2Yq;{8ISt)?aU4g1~s5B1#NHD+}V)| zCvKi5-ET7U*xp?Rv27!whQ%sp8Wo8e41`l41}3h8MmM=FZ&4ML%Kyi0KySR6UxH2e zCP)Dff%p+o>Tmkjx68ZSDg(2JQ!r@aI@_%ffHgIK}o6`uPiBq(0^P-c=;rE0?exjrFE)xf`3K)1$#>v%g~4jz(B#i<)B!kQPx4^BrfQZvs6A| zSyR1flajfXQ}i}&ic4`y7taG-ow=S`Aspol+QLj8g|~P|96Iw|!P-rhg=ueN$X2~Bf- zzqxSGEnQn~9pJKje!LtHzbgX0baOCO*;ZpERC9Xy$1JsovW_s`{FgM4B2#!M$58bO z*O#9$P_u$1f}rdU5k3-uQF4CT{({{XjbSG}Nr5j}%aXs$)GwcoeA=dS+!!X( zkbh3yd;C$N9uRgI3Z$a!=UVWS3k%xS&67t|s~a7b%{)|(3V{thT!ej_KUUT#6b`TJ zezY-f88)Bp?gHp_baWCjC7`_ZMiHgXNzeiTe7PCA*%2YC|l5$EtvyYj$c< zdtQC@V%*XqO3RP_Xt#6og9NALHY65#***xW)!>~rg&#rv=*2%CmL@qR6rljW8pfY! zd-@r*DapD4@oRPxErVMpgtRt|+q_i_ZLWNbZMFw{STRyMPtOdbzmu;*W%jHetf2{y z63!esuYe)Xm5-G_C_$E+xo1bWPxKs4-<+F>5v#QdAF`_)!V!jcdU*tG5xVWk+N%xI zUHVyn;igy=AGlR>a%?hTAa44UuQa`X^5=3}*?9;$Jgi?BKB`d0-(@IbkY#eYW${97 zC-M`&$B3D^@(KYAW@5wGC1r>*Z$)tTAd!SDo`B8xBCzSTbnpezqg>`Dol$Qn^6~Yd zz0BmNLA{$qvCD4 zJpu>a#HEA4XyWty+SVO6Y{;RL(FF$Tvkffng9mvQVN+5@F8zx!cw_XiCt3^6b~iVl z?K=_VT0!DN^qcWep6`k0u5iU)=$5B5D;HcK$Sct|?++$AFPOlj<88r<%6dSRAICw( zP2v1Z;!4QrJ!F_&F1E#j6=J*Z!hS>8#aj^hfXWp;H={!p`JmBCE>3DY#vcE|p+lQ; zI_LR-8q@E<= zdD$#DC|V+B8-f-&d_#SOV%_WOS_gv2&;Zi^ef;8N{WnPksRhXf#>4iHis*j{Dx9z3 z_+J1Oju_AXKvQ1?70>@bQ(puX&;LMEUj!A;|3Fh;1Qid@|0Sq6nVJ6~sOYgdng1nC z3itmYRP6sjsJ@WwHQ|4d>hstanefeT_u;1nfryZJ%VKVHv%Axkd} zRq3wpDrpj$c)6N&+YvqlFhd4m@KbOtQW29;`+bx32ShM~(qoVp|IO-0UZGTfy~-V` zTcI-lNgd?a!!4Mzp~l~x?cM0c=KlTi(!WnEfSCh$L@RQ1Y4z<^z&A3ErT`*uL7a?O zR+NWYp@|zd{@WyW&|C<%ZjXG9yz~OThI7*jx3pkKeB#F%{zH1JacyUH|9j;$9vqSD z=yw`#nsC@sm%GJr#lT0;3*%vih!j%sb%$nKOCT)Kl#fk zY0s2_Z$H3;!D!7}q=;!zn`%E-ULu};Q3|WMFgbo_7U|niN$=xNRyKY&U1v?g%FqXWlIjL z{+tuuCPRP5A`F`qoUQc9ejsqk+E?6cT=iSNUm;U_RNf7bz#4NqVnrY!t9w!~d8JVT zaI@INR=i{>@%we^n|ikoOo9%=f;TrTV*Yuj9GkxMoGd&DuNwkKuc=#I7CXLMU7MGk zzy1i_KD##A*S$WBAxLI+Iq7lT4h!XMb=oaI4DRWp0Ca?EmY-=Rn*ldiFq1OG|IWei zq|UNK%7K4*%0Cnp87S+&wI>!>kouSrAZPfT0(AWP^K22S@bmE0Pn2j$s+fjd&lMx~ zqZbtaklVI4`&j+KDnEPqxwvieS;^~HUH4bM87%3r#g&IPQ1FsT+#&|XYlvB3p^-9z zg@vHnk-9=VyL>vj06b!)r<{b>nJx8VTF0f2lKfE(w_@t>bY#X%(px9FA`==>9XduX ze`E)fR`oDfIsC;BO(I!AFdv+f=+{fd0O}cqE%xySXJ%R?w%J@YZMq?0jnwDV&Jf$} zH@fkC@oVlY2UMLcxS2O**czf(U~KIF zJU=<`3^s@vCCIBA&U!l{o~{p!o=T`w*g*3!d2|RSFpT%rl<3F!1C9K;uP5yzReLdE z&r$DS^jn7oQ6DT5mELmf`F$rXGo^jsI*LwMWn|&6H;+lnQBrYl)htudcykyz;t5E< znfwL*9UpH9+Oc0rgAUaRm1i{ec9`>#w+mT-nV?d%h1-zrFm4qY6xo{e%}_BHvJOKB zb3d`ATC!oXqVV>)bmOm9ZHp33P-o{~JL0odmv7`9W`_reu1Hu{nk>%VBa>f|qtUuKvfw@V03 z;{Klm1>h!G{6#5Bv*F-Tz#=aP(kC#%5f7i~o}44Hggu{51^4wBYVSqoPNw8TLkDZG z6;XLd^$-nz#}K%(J{N}@3fl!ZcFU5JbvUGc7NA0wpJcV%sV+%@82ThuN4=z&2_{xs zaV07$9^>$abs;y9vcjprn$iTORBEVQ{1Gm~rMsjpKjw(NM$dvZEqo(BoU^eZB1L21x}qv@LX`Srm}|Eq2_m&{QQjH=* zmW$^yn#2foSWU{PE4Prl<fQf(W4QkV z>=zaU(+ zBe_wFrFKlBX$dIU(j$c}jN9V0IdS6hfaG8!c^Jf<-j(H@H1pdHVcNa-zKLjup^8*N#3{XesR|lrr3k$llGXCJeuz^ zF*9NFUS2tsRhzlo>m@#Pdh7Ay?N>XMOLBh{+jujsWZGr){ln-w2AOoT`wc}%@nXbZ z%>?m83x#0S9a!JA2k!jEuof-Yfk2M&Um)I^GYOi?2Mg)v&?6(vAV zJak?{u&n;zRDq>9MZzI`-pJNT2MI%UPVVYT10y!g5PDAEg5A<4tkJbf>IacdLrFZ{ z(#8RYb$|PbSGeYo}L(pB9xv{U@=9;q+CRs zuol)_>qyHsp0Iy64^5-$u;pS^+8)6%%HATS$l;4CF%)>Z?2V|Ttc+dA$W>*5;rF1= zn}Rs!qWnT8;J_L%pGVxSs~g)87$d`n4!6>yY(>{|8LjiD`*zzST*_UZX%DQQ=HdRg zv0&yQ5nJD3p}gV1IXM5LSI_AL$@aCrbNpwg$^CWB`H$cd4kz=!H?j*yIqxD+i7B}Qv820SdgwedSjkKU*h zEI|KA8^)RPZgdzK<=EMedE>AQN}v=h=x$lQ!l8N%*8yF3)M(`P`kW~%Xp)n|C- zfBhTJd}A>*%JFt-mIVs_?1A~YHK58O(o78OqqfBM&`BJh1tVoRh~Sn?CQz zM?z5h5;g0yDS;rKywOSD&U6q;3=^^Ep2Z50H-Z8L_P~jeq6RL+z-eN$hTE8fg^K}| zK`6d?h?|d5+(;Lo23|)?u}13*sTPvVhmqiD;8gTR{4kCT5ASp5E-erpAvd$n;?=|X z7Vc94nOyV29=}p5subIqBt5ZBv54Q!k`R`^Kv=}Q&s?aWp;0Z@yzI9&Jd?Shy@I!J z#?+KIMB)oNT6pad;u(2Qb1ItFD1Z+X1qb$;ii$=JJL^-Z`jI)h&@3u`+V0s6Ld$G|U zf4q|LW6P1Gcl@58(VtV52rB{(ok^PLNy}g~O8{04%%pQnBE*1V$Ob51TijaE>WB=< z_>pgeDYmW?(t?9%bBmjyroOu`QAG`28I-0ftUK-{DFF}$=-Ius4zQX^^E~|0*%hQU z&O5^eq`84X6&!?MQ%QsrTpuw<}Pk^T@iAFX8tcQh8x={u`!9J!iMgg3VuE9yGB~$ zXmc?;`=VN!B1VdmZUIN;mnaC5tV(lVL+!*~SL^}!aC3OPl}pEhjE*@4r7tXRYU{d7 z?JWes{XUnf?ugwA7c-EcNbT{#q019?Wnh^GAjr*t^fMpc+}&Ij?gW-;c(T27z|a8x zHUjQDImna6`|!)h{7wyysyoI|Ux)9#7%x985C3tVjTd$i*kRLd=$+hjxV^Ob2d9a< zc{*urbmIp!b{wjXqMyUleVpNMlBrH4gk#B{YiI=y$Mq;friVa1He8&c1#^5~JQM6% z9K>V#zn6UZoJgIJLH^ncJ(ld>qP~8+4H;+Hi4u+Wva^GuEJ}+)Q?{5-M5i>ey(@7vXDreX8?rmgL= zSL=CX&nocPx%1>-t6(eTI%Fgq)<1{i|EIXg!pzR|@8V{*nq1;KC&Jg!_gI=hK2kCM zgeKjY?W9^|a)0I_Z{)^!D24OEw)6QBQdqg9#Wll>zvUT3cpLn0ylOo|#cuzuZli_( zj7nI<>sVyLD!KxCb9uXZiTToCuDsg8pGiC7gEJj@n9<}qgW#g2VgtqEZ|7ivLfh+t zkj?EWjyk|!yfj@jsakYP=OP~~{nIxpC+ueq-J8h`P>BIMLmfJ_#VcOezJXVM(D*;9 z;PHAPVCS>Wo2WG0zgSs@uROZZ_Q~h68Hs@sMghv)-?7)ep8<5Hk(X^x4cj#})n8hz z(%Ldij8%e!pfQhoyKYb9w=!nwQA})ST)#u()qen?zvX9dPdyj#fN3G4K#<&e%lhL| zBKVuHhH^rBBJo`a=UKR7Pq(e<%;_BFz4EM-YX1XUt)pgHRL*4n`V+ zUJ!u(#m_w(CokRk^ASvC%KpdV?Q!=Y?JxMyn+z+4BdrTL2ntD8NodKctZ^Rry$JR7 zojx{Thi_PZ5ckXw(=j;VsdWnY8L{DYXFA6;;W!1p#8Uy9;?r$&65~1FDVz>j)`>}0 zwAcy(ms?y$4Azz-+5JXnDxdUst%in~46_;zsWG-Y53YjR)Hc@0$$Fb#xfP?umX4a5 z+!Zap>m_Yw@mIz?_qt$*!~ILVB`?6>>gYqDa(ON?AH^TQD0HR^d$0>@&aALvMVvdf zU7tOa9>Ggt_M`51UxaeX4p&KAhety8QL6GhH$5KLT9+Exj$NcOI`x8>`*LHdQmKd= zXQV9qdBcRRhQZTttr#_?U*wStZSax}P-lgkR;Zd)Z2XXq>ZrRQI3;2x5q6eEW;7de=CA>;P~pzi0CYY7F=T>t;hkktb(t;;hUSi3O#*T(iEnOs`w1ciDsRg z4{?lt)q*$l(^d76Ptji?I;8aUlkLL}%css&pk`ukVsCP(_ackwsN_1F!m+z}Wk)cn zlGp~6c#O#*N>kwvy*kGFE1)#Ybw_!zZ{F?*ih0xQ+9ZJ{yZht{@ zZo8)cDGNFOT_IbgL5D+ffOE34{NFDqR@Q%gL1A;Uvi*MvMQqGGssE~=_2eD4+rIVm z_KKu|jg#p4OOw~ac0gkOnCFCguKN{~#eD`98$$_-c~yIZF#b*QOTlVLA*>2sK^VtD z3PUDR(HK+=!WBgP+wxpn(+>-f5NA@up^?fLRlAS{OT z{`y4F?f=kn_vb3Bci&+PaQk+YW(dvG)myel_50OTxxT$uk2BL>;04(q<2|oLnVXOL zESNp$L8QErQ07|`;cqmbw(`Hmvmi7QstEBZpbBv2W!JOqW!2yO@voU2m`GqSCf?M{ zZc?5_6vc_t+J{bqjpYwn)3n;SbuX$M$s+j}ib_}`EkTKbUF3*U08u|vH9WN;oHJm{ zh`TJn%BLU@th6ZOC#ZLglQWjf$K+Kcxb0nOU0^Hg`NT<6+JjXG_`a+ea$unvCV6uG z1BI?HMXjjPbkpUTjk2XiD1w?(k=0I=(I=&QSczn!Q>`cA}2 zh|;AwWitIH(2D{;HLsSJ3@_R-lS;=rViPrHNlJsT5e)=X*A1Qw(#s(3)8ZJvh53r6Q{rrkhrjJ8?O|>q)$(S4nUyhWI(cR8iT>k zOo8ECp$AMyaD{OD4=4&e*rS^gV`%NT&aPXta8UtA07E)AIBha9@frDK1OhH5uG+=* zAXW5Ms_E)g7uVP$>{ATE^KS$}QuTtBeSe>R?Xl9yvH{zzB+be0z%b_;u-pQf<*KR{ z<1SlVTTr`@!0xpdc5Zax#)=Aqp9TEjNu?G7dRzlC*@;|)nbcQ0Fzb=aA44)Re?F_Y zu8dHh0S>C97^}AyMt^7qxkYN>T|op|9onEQ1$W06 zbdv(T8Eg8*C<4iV7efo74yc;b?GAoljKY;|hQ*Qz3A*H7q^5?F22ke}HBM#)gSb!| zaZ?mY9;x+v9N!st$k||?3ERy+j1}S(9NRo&fX{(i!k9T@fOQ57a{Cm8<3OE;vscOz zNm9h}JTt2iq^X-!pcO$Pb(2DBx)+eX-;*Gb*d&Y`u3i*%=rUf+*j>$0#A$^hrTxkq zHFG0pLeB(0WRFn6oDuX2ca~hmD{UiV!(ffI#RC64z5FyiHamz{0mGhZsd7PLC-#;d zP=<)-aC-A~2tW!^!jk~@_hK-4jUp|W2jRqEkSH$}k4@e6!MDQwgl8FTS; zN6A%g)L;A}**M~<{F_s!gDL+MF|gKM{2 zb6t|C?r3bpM$#M+#jd<+LxfxrV91;6xS>efje~Z4^6l9=-?>|Idmtf$T549GOSj36 zlErn1sPkD4m)9@4QRkdmlcOnO#e(&0Ohxo@;FqaVV4@43-|Zs2Y1}t`9Jr>48VoF{ zIHbvpgY_TsSYxoAEhD0}Pj;Ed?TaTrm80E(#P%&F@TEG))0rQ*^F|gvfQ#RHXBo-q zFby-yIA=Obd61IfJWJn$J4zPWCC@zRGYT7B5(oYWhM?N$cbb8qp=Ck2W!k?n>_X;a z0Ogb2gDrIFiNh$}bj=%0C4aw}LkTQR*)VPjmy4aKGY1E#5!|OFmo9j1VDO`r*!=P$ zg-DMmP_qzA2zZT65JPRB0#HZIOECmc{?N=kCZq}(%U+;V;^Q=JjgdVA! z6Cb~9(LE!Os}A5A) zLUs$JyLzOLwvbDsU$swY9vu~j3$;4W@k?Y;wfpI6)7!i;f4Q>UEO6!i7?^s2t2+0c zXW=Z%g>hj|jK|{$fFewE^r=Bmcp=Da$GB9#_#6-kEr|}DZ^`&Gc9~-P9Dm_COH9d) zUZQfda|-(L8785(x^XQ^l&kI8RmZVR=P1-BrE^5|qDbGZT7aa2zDxpCpCRx~1VQgr z8Sxzl`MC8_NY7yBXHF!GW=`45U>tKyXcjWJHk(;=oR-=GVCa_M0HO1hziUpVT=RC0 zDJ)AdO=vsY%E-^IOMk0pDR!`h&a+~aeF??&h$V)QZPV0pP0RGbZ`{{ZDCFT;>j#V* z`u(bY01G4QCW|vRih0Y(w@Z3FO~GZ}ceWcK#pU~a&VAP>cjY$U30)OP^&}Us zwe|s**2Bf{6On&F%J#!s)%z!H_sOdEoj?Z2!LY>U5DDGvR)7my&P_B@1SNipruRE| z47Nr$P{*hVTs8Hh;%6d(roaOIPXr;Bt@@@nedz2NK#7&qZ)FUj%!wN;^lw5L1uyS; zkwkLyKK=@B`|)JbU@*U}O=GF9qkf(dJ*49=Esp4ci?Lr@n||t>7*qNqqzY+@%?17@ ziGUO{KoKIZCl16MLYAfv&qFgE*riaCYe1{fkPI5Yj`}v-^j0_^M~kcP8pn>mua8vX zE25Wu2vGU4FF0?#$t$wz?{#;4x6wKprb%QF*cpb^`tlOTlz%Z8wPX0nyZOFkS{q}C zN^^4y*sLQK^i!fOw}hQev|$SmOQpFx$Birg>u``Bv$2SBKk+pRTj) z*Yk1JJHF8Q$-7vpEpbpTT{*LJeRC5zTsn0o4Uk02)jAnY8@MY>eazuL84epb`k3No z>Or%(iEuFRrJUME+vxf&lRo=&gIhOaWW&Q@AVA)8wze0EhjXF`zxvy&o}WuEg~ikD z<}YrIYi=0ePCYQf{^1t6M~iwvrp)^FbAI`>|A(^5$@)Lde^z2nwtujxU;I_Juizg8 zz*oTj6|nyI8IG^upW5Z?hVv_6|F>ZPi<9ji>%sp+6Xs(6*L)V)zmTgA8_WpZUy^Ve z!H7z-+2m@|MjaU5+_EszAd=`h2z)A%KGfw+T+2HNU(MAT4`L|BTwD<$jYMR79G7ZADy3#HOoMENn%C9hu2wW$;M zJNV|WwL;NOzPpTUAE0RDOF27VFO9p1;p4UhUu~;=8U4~81jzf1JZMvm(x<~Ws zCcHU8v;3_&zK8K-4jeusY9`w4bqAG86Qr`vXd8xSz9vQjP%dxGrg?RN=TbvaU&o$?zWE=hjYxtKIo^@cDJV@UQcZ zXn<#)|BnoA{jZCSQ2NMi`K9ylg(^M zGKve**p-E;7Is2?J8Mf_-Y{vqGYj34XluI+v6%bXh{r8h8D+d6ZI#W^N{cUH_7#(` zC)Xd~vo7##nbpP}g|%aleRjEj$y^D429q)tJu`^hra1*gGZ2KmmpUQrA`=76GEkh( z^RKuZT|W3L0w)<9ZbB3~LsT&FM>L3FAcW!Yq-gOZUfNChwI84qq-hc~yA8f%h=ZR> z_wS$DSA+k5H8I(~1}pwyr!xPSXtVO_@((*zs6uGny~I6hJ(fTx!{#DTt@Ec*xk_*Y zO~DAHIpP~83?I3%{4bJV4r9b%h{T_l)N@Kd_<)~Firah#Q$W5KpWYjv9G9FEo|En$ z8H*`*-8gRiY;UX$G`i2`uig9`Y2QxRp40?+GI_3E+sHUp9#0fC@4sI1*F1`+tOh^X zmX)KvNPN9?Qy-O9zDgu>V;>tvGv^-)vD1hz&1>C)oqBXGd4fJ0thH1Q{t0{CUixL% zU2eQLL=&WWVU2ywLb7~puQDTg(|BADD$OMwt>)Jwmmqt4o`Pu$2~#f3CC?_GJ@jEh zlRV=fI@z2N)rt=-c7XHQ+?zw5qs^(wnxqWlPKN#LS-v#37CwjV*;;#xUr+DfV-^}O zbw0}?x1&cu2xpE3gcLCdFL~2EX3f0i8NsS2KjRi*b?VbW$zN3tDjglnv^2LFr|5Yz z*Pbhiy8uVt26X(jU51r%GW=Vt_N%ES*fi(Psrdqb2?+d`3ZmX-4r#&bvDPMZ%>{ za-o$k)0V^GbFkhUIhK@NVX~SN*6S8uCM9!28dro+&1x z!xNc{DY3uzr~nvPRm8tbifk@Ji?J;Pf-&U388=&#N@vm-d#I{u#G|Zey!9I909g&wg=SXh)*@?hq&wP^ zRm<5xd`VnR6bb+f(Gk_(u{dIt?292aZW(cnx8D1qUcdtoPEF=sj1Nz5W(P$Uc)V1p zY}S^0a@o*SBR|R4uEL@0BY2-@{##6VU0`Z)_t2rX$%i%cVR;PiFe%Xl`d%&HgIfE; zhjX#49I^@Fb*zZ=*(im=S{LYda1V(^Ft#w55yu`v{ars~p)|z;YtqfKs%b67p!J@V z0H;}q3_wrDq6Y;|Vr;WLA87p_NfOyt!`j}&ljpYMVjnIPOC`yCc zTPB9u92Li=FG!mn=a0d(Rp>Vv!^KCtk|Yd<=dWpZ-pMoK$8GG(x%n~c8QeiJJca3$ zW72HPd?YhzRowJfy^=F;=~SK;6g?N5L^(z;)WQAw3dvY=P9MeTWtw(gDuUynpNXxZ z%`p1-e6s5kX^rJppbL!Vf$YDe_p0E#r}}jG(}od=H;)_~ucjjj)`+=`1-eqQW{>a2 zHUYq7LBXX1bEm>obiG5M_Gl$^W04!P5yQ4?hZSwJmQNs~PVHb3*avw>?^rz*T`$mi zX_OYc%-)LdFZW-Z1fRtPGD}fse}&62?|y01sCSkY0lYiI&{tn@P=`O__LybL_n3Aw zoG0OCE5*aOZ3M3<1@(FoLlDkmh&;q{41mEkZ!LO!ZMN=Fbx-s zcs3`O7Vlo4_*UM(GU_wOgO}MYCHR3!ONtEIKhivHfr~VHqSY5h?c&>v*fCdDp~AI$(Y)-ORtc z9IM5=kgycp4{}d9g9~Gp?0M^v`Q%aW85I-KxO|2nV?|S)tj>mLPVCOWo>GY8b!9%v>@vvif2*4vVm+E;=$dd#2(oi5<0mC9w*1%5)D1b3lX%L1dRux_j zFvVUW^;^#WTv!ifgFfGeqBgnOFdh_e?rrbf84Z^xVeMRU>mqa2sPDU22Rt6iI z9=@VXcTywYR{Kq4=ZoTMnBsRRQI&cpXUH39!TAvR;+xwAycoRi!%UDLI>(C(BREoG zeplvg9{CIE4^Q#;zp3+UUQrJ0KS?vKhT)5jvZDJRJk7e|oI^g=jH^T`*S1xu_xyUS zsX!B-`m#Ojd#vdm1Q+Be~F_6AP=>2JT`*=#WXFw@4vRYo@jZ#Je9JveZH9jtW0`+ zT)cbJH!50unOOLfJNkW{1s%4=9(GPhX0yDWbqv~auqatg`ELLjrPqg8e`h{~4tt1_ z{vKQYoeFxnf7n`UAPLJ{fUaZ%c55?ahO8Fj)c5lEaGKwETEM!IPrfzWE3N$8$VIe# zXuWtnNa=FssRfQd!h1br@jI~rX0s>R!?D;PTRXjeM|Hice+`IBGMSwl3F>F$oAZ#^ z`LuXgs=6ROn*Gphv7hLfU2G=Hy^(*A668(9*xh}$3;_bIH&R*(wY_<7`Z3-mavmW3 zXdDnM^GyW}Qwl5D<`Thk50?s}k<<#gC2nt z9=rw71-KfHxA$i1cF`UHbOWK+Y>;(9{yt$9$L~VIuhC&!bFk=*0rO!d0eo8xJ&BiR3=!MuV#Mnw!r$0bOrPqTOIXRnn>> zW}X^1J8(?ooku1U?r=~gSz)eB;h5t>!1{x;SA|=p)vUL)=VuRqVr!CBFO9`^;LjWm zB3N4Jo9w>jeNBmy7zDazcTs452U`4MQ7x&UYkHUP(Imoe=7sxgt@uTCD%R!w=9L5}>4= zSRb0elkjXOAk;9H6iRkKH7Ag2Ml|w)P74YY>wgxWt>S?8LF zjdcNlc9FoE3SE*!*td!n@JZJDrb)<*(Hv13kDKG@2FLBJ-aj5#I>wr)s?VxDhs%&GN#M@%DDfHK zMP-;icqCN3p0}apf`s*i^NpeTB2O0+!Ki|)&quG{jU63&w)A0ZN>1k?6yk&X+_$8n z^S3>Yb?Q1|mfBBZjkioIuUJh+RwjWl(r7`exaYk!gz2K+nTK}X@rlTy$d!l_c z@m2UomOnYuS6B}W#E3PHGpY-LIL9XddLjDqk1$R+_}pc2mpqhqAgm-S){hL_CkQir zQvQri|G1NFmv#XG%?q0mKB%#atacU(uJ2JJxiyVH&KAs{n&n3K{wP90x21qlGFTli z!E{-gB){1+ewGwhejnu_{+4Pg`?mDF{HW(eW~hPns7UF=bpmH@Z|36q^$+kPt&v8_;rT%Yz^rM8N1aPdtI;4uDL+(CM$uA9Iv)g}H!o)Bu{V6=zgCQ%MFN|dc zZ+<32!`Bp|vI2%G{0R8%%a-2t%DThH_LAdr-OhHS) zLtaT_ap1{o(J;-H>z2gB8z0cv?#Qw=vNB~cDc52!o3Fd(L)A%S259Wp;x#aP?cdx; zcpYqS6k7CQTTDLbeArdG&Y4}~TYDm}{G7p3eAu|4_fHdVZG4ka(4RT0$0w5dD;BFS zzmcJLHY*H(nmy`Tv*lq;dx~qgw$Che{)NZtn4{N`So5}Ke{gl# zo`nf9&YD#X0=||oevmO_Z)m}JB<@0(aO92j$1nMbsWCW5t$0?CGjKkmvhC5_in_ZG zo>Bip0JwJ@=L;RqoZH+qtcE#H4MonfdU6y>pVvZv(|QD}2Ec%Ix-Nco@&{xutFe~A z!lATSP8fJGm4vmB@dOaB=;MgLijC0Fe{Rt0j46ifR)DwnvOU0S!~NCc(Bac=JT1ZN`xPK2$wI zfiR`q%lDLXbR-11yO~$9q184NL4AwiPEQPRsp#G#lBViObFrzry{Bsqi!#d(GPvbS z;B1yagk)5qPd5(>=*CDRB2OfMw{QKeCV`R`XdIrG0%$|}t|NU28d2o`nj0nItYUCC z9$47*B~@{b+uGeK3aZ=Nqf2T<(L@0CDt2;gM^Yn<$fk6l;HOXhez@Z?6}2*9h2oNW zpaoFw7w)F%m+vVfVdKM>vQ|t^F&=BhQ&3@uZpI$c zOi2pVS93=9V4n~;<#&Dm%u;M>;hypG;>ar+!IwfqA1X?8HA=rv2A3p3@?vVMImFzu ztTJdp&y=_Dy8O{&q7XKI-sOaTA-OR>B&)LBmj(*r(m*1G=3u3Ddf6h!PKm>0nWtjY z1Q>5pVD~PCG>Y0ek_ByF#p_rGkE|!q@L`SO3wBOfa8O19w|_KH9^0oln3tjUSx!iw z;o?Out>gr2uZu5`Bd9^HK4-Y4+(E)*t7|PBTrN)&I&>NEAF~7;y)}5S@kJ<2&Fv?H zpQ&TIq!BVf&(r1<<#bJ1R2itmp@)c<0HQNlk>eRJ#^29YBb-c0quF55-x&5O;aJ{UDrPSU3@GZ0lw$v#v17tnE zK!{fM?Kc%0wg;;|qima6E7If6CpNzbmQec9v!rd{{OyCg)!;85T?NVcJU?v5h#BLS$k!8WjlD=_1{<;@~A(vA2jz% zjUJVi+IQ>IU~($XGZ5%dShj4%Dpz-uM6;jgL!mqDsAP5RFK7LDwWieAi{c1Z2Vy;p zjoF4pE>9ZrP!aWo8F3Kxu=Hntjq3e~m8>HW)T*V027H0xhizqZ z62|(l(g?9IrAA!3P<%BBIW1@02FVk7uR)m8sVQn&$-kFHnqKJ@ol7RT$PpAj@14`x zdTmFoWNzR!)=}239;YD7N>bmZpSxA=SFzsFwwleNHE zIJy7D9VG)$teiMwXre&u`b%sG>9tfwBmN=}>QB#38smwoiV7A61w}>u9*#gN`As~w z#E!5(_gM|t(d_Y(opGI$vvwu-LvrSNG6I~(ojc6lEWnMhozYwJE#(Y>TBW#s?yJ?i!(1lsyO}VnYXu$bG4M36qy*&%E`3@>%J&T zVS_1LxxG;Zi47!md=iLRc}ryE`7d_2kl(c#W?+uOa zJenE>Q;Jk1SsyA;p}Y54Ys0ypu)U)(Eyo`qel0WDEq5`i(wFW0F3dY?xfO94Xz6DE zo|1-k4;8YJ5ySnb4NZHYo977E-x9QB<2w5oQE^W@&Gv0^EAv^$yTXJ+Z*m_``}9Rc zV{2UQNVHDK?H8;FAbLnIv(v{Jmt*tlCAp4Mi#+AUDs#v%mG8Whay9L6`yIh<_EiO7 zvh2~twNbCC*PNQ5x7s+yn@UN%Cc)b`M*1y2%f;IA&z}bAYJsa6Lb!O@7_;o`wEd02 z_b8WK%7cecZo>S&j-m3db$ZHp25W~JZ@3v9ia`1b(^px!UZ8peG#jcyyV}HeBy^2b zhW0i}E~U!Jtq`xW)9VAendzy{6t50|1kdhHTGH?92^^LW$aL7@{E_R&bpW5s5Y<3Z zG{ICjopueTz?!Pgmmo<(MioUo*~Ook*`Ly0j5(o`VmH1_Z*Ih(O<#I_`JH+x(rv3L zHu6bE2 zCYBVPIlZD>kM}~}pBeM;yy92C#6qFy(KQ9J#NvaUi9z+)RfbmwHFDqSw8jq~F1BbI zw=W*N(sy2mSsob;`E~idptC+Ce&1ZP?BDm>)2{7B@#s2=m9f`(`M+^j_R)?rvPQ*L zoHnJPYm7v$LV^C;xu4bTAXEl``xVW_WwHd1BEz&hNj!j(5+;3WFW=)r=mP32$&cOTd(CPKj09W#=Z8egmu&|oXb}7w` zFq1@wcCsfMWJjqqBA5u6&D6eK!qB<7m@7bik^LDzSsai{#pTu`%Nv%lFS%gz)tzYT zgB$6ML+O$jZ>=OP6kjQ7P6`Ssq>)#FVn~~KecsyNpQ(hY?uX{F*S@_ zOEKp90L_d2fXI0?ai{(;@jezWE>fbfqldCCXyNx!oLLgMp(H4Ph1crAm?yr8tJJC> z`vkvDzQNtpeFb26b%q^1DLqgNigHWoO@OE?Ve}XXOuAPwPrz8g2af}*8aQONLBYY8 zPeZm@U$kOyIK0nqD9H>{7?%=gkHo1c-iof)G%*UJENg&;~u!4!kH)|765)w z{hK2RrS? zm6b407ybL)bd4!6oilDqay7n>EJ1Rm%on7&<`39>R`_BzuHC^ZS$f)`piRdD2R1r! zEfOu!;JH0<=~yw{>+7m-AmC&UzB}xG=frUO&9o=nsO23w?M;yOKS}K0^Ys6(((7v% zBFlgJc@!%fE*k&dh1jy%ZB@(n0iA8lw0=(Syc$MLS?rrQQUNKk zGYm@^Y%mBcY-v1rFdNhWP4>4y5@A$@RL-jv11+#^6xwd$rtBJtwj7~nlLyLvBZmPrN0`PV0D#y1?sPWm%Koj$3 z(npf*$aN%lV#vkoTFXEnn+y4<`pFvQLBjtt*Y~COb>+(U&AsPAzNeLph-hpo)>j1z z{oH6LlY8d&+Tnra^Ng&d^%QYMg@T+jG^ zGuzC~_yk`Bu7$B)xf&EjAaeI4_Gr5J>nJErXF#s?Qoju)3^a4(@=+}1Oi>ip85P1} zlZqF}@RqYAnoDBBbu!%)UbgAk7jpwLSNB>=DJq4=^3Ql)eQUZB8f%X>`FNZ-0hA$R zD%|4KaO{>bT_Ai>?)W{%qk=GvFsH+`U~noi#@}bgvBi2+>xqBP%67tp}qXj>J zkl3ME`7|*I#K*#f7n7$!8U6rLCjGHRz@(!6ktA8WXcR#?v6lgZ%_3u-E}wdSwG?O< z-K{-CVp);iZ!woT=Kd``uywJ#U8I_-pU$Papp>&jm6An{}q zQWYEuVa0L9iDz%k?QYI-oz|ukQ7oa$7D*%6(@>QXQyi+N`-otQ#e8Ee5mKyJRU(X+ za<^nU*^~83w(N)uim?vF`Bt2qpmx1r|SZY1os=Kq&o(VU{*2Z4X67@`Jm4#~n2?C|U2ml~~x}|l0{fD!wNNcU0 z*60Ie?z$4S$;f`38zjGua9dpUqB2THWJz2NZjA7n=(;PL~ z#aoSpViUoZN!~ zmu185dycBT3jpm?9aA=ncC%KKUv01%QP5V3)3Ywar|E@a8=CaPRJr!@^IbVk zlny@0Nb~Wsp}}h_ag6M){iV_MS3BplhNJ5b+F!=!%|0|K=jX(XP3vcLikGtEGR0l|?HK7g(h_#l z?mJEN;_d+^f4JX%4Q`B1>fb)4V96Z0g_2E09!WxgW(0#BX$Cu^qvb`8Y80ikj71jO z;Qr8v-AYQpa*nV+jvKe;&RQs0yL=Q(sT3rkJS1fa*;3kdHHTVD2};*f^AGy2;=4>w z5BW;HCCCpK%<=P#IQ16Jhq;w-db-tF#uksBKY|8u(#KEoadT4V;cZ+IT0#n+9cV|J zP3&Vyk@lkF_KJ&A_DTS^txsoh;aVAS{Z^WT3*7N#~(_ zcl{T<$B4S~KUtLR|ECm$VBuu{4=Uzsqp2X2&GM$w8x%cCVjHooBtiO%N*~Y(Gy6cOLMMayqIDBTX=>o zyRqC9LCULnz;m>x;tjp9no_9;oNzmvKJ2UXWlPK5nl*DZB&?a8-13^*Sc}#g?+xc? zYyGW(lsy%*vUxg@9v8zdu>jymRH)yCZ?ZS4J!bDR(0(${qY`Dv=*Q4(ZN$=oVO0naCe6F6-r*ynO zu?c=6l`~^JPV88v-EUw^cv8C4U?pe^I*QzNpF8u-Bndn05d@v4X7ZxjEBJ~^J50Sd zZx^B)1|l-wk(?wOMHm1(8ezF3nWZ@zMGr`uwKpenRct(7rT=~`qM=w)Oui!Wr_P%x1t^6gCsB8;yB5S;QGmjcqyu+YuhA ziR)US-Bal4BAticZ4Af=$E@0+9B75OTf!dJh#vF z<$0Pq30r;rBaexf^3o`8kD1q6qVuPqITG61z+7CvtvKUBTnDvIr9!0DTm@crUN|X! z`wteo6NYBouH;+SELA}g`Bq6a(gm8zyi@ z&8Nm%f*ibtw{JpTaDr+Xa6jMZ>p(9hTneKm;pBl65tQh^_& zx3+-%6J1vwLKApWQfzjKD>u~ue=6GGPpKNwj*rBS1h{mQXMY9>Y zkVRr2p#;!C5T`0q&7egq#M@4M&AV1bI%307BkDjcUaS5vU%#MiF3bFeCATL$Qa6O2 z2CIkMV9v>>CKIKtTH0DQIwL$EifU%y@ezWYQXziX;`JP-zI-DNXvoP7V z-z(P`6Ns9x2?^GPbl73K0D_i96Rk95fh95_b`GdiJr!w`16N-jZfEe3Qhqm404XNR zT$~+#R&d{$jOD#XxhK}`*KFvxxn0|c=1`KdRXf6}JMgU8^Q{eLR>MLShon1De$%5( zf=Igd9$l#-g3X$ZVD+W^Bs+%_^(T+URG!EY45!=12M-2^X3U8y2#rBPPU+M6E+Cr$ z?E;wfuZAbyU&Oj1#?h~GSG+I8mFEnTa3P|vAp8x0r`|LW)P)O1JX`ZtgHS%yIWZ#< zQ%xU?Y1JNzi@7wq-kH^is;1d6*>K(S0uV}B=;pi{K&xdG$Iv1uxXMU^?!SuJ%LF}> zYb(ef?4spdm;8VuVK_z-swr$QEVY0;^agZ*xcD4%S@su8R#kz~QPm{$P-YB1jX{X# zyNaA9wHwUxH;qO-Wr`f+MIpr7Gw?IDi>LYvJBr(!x<2R8JNbKWm%sG1&1sZ;lqz8jz5$eGzsMa@3EM0R1-HuAV*WQ*0ORILrG%gb;8{E1aXoheLX`4O zvBR};T$)b92(#?r>09KZw~*zz(xnc5g&=7r9+dR@?+oKOOQdc#b;2ZPkPy8h3HC%a z<#Yqj)>^R)_4R%Kbd_63R7!AJxVs6k_0v{PYGcqYw=s9>%=puOy7^c~4ugJfTr-}` ztnCxsZg1c0{k8lB*o}19!^qdT^>N{8RiE$Qh#g@bTD|>~JDKM;U$gNAaLH$1RTvq} z4&I&Y8|JNS^z}L0>Q)(6@!lR3GW=M7trDkeeB}3$>l_M0IQLSBP)- zt2fj`H>ILw58U%b(u*skBm2!Lsxm7UG9WU%;k4NYFnd)t@_VQRPg(1%L*i+J-VE9V zIy|IUMu!&MA=D#Szqv>N<`Zt5YzqoEVTTeSQ>q*JQ_UGIFmN|8lJH_NrvDPNaF5sgMU-EjFm0x8_%dbNlTW3e`29T;prq2cAx#Ks*zO4}rEe*v0;#{jg1dj0( zpB6)N4?<(niwZanE(3{MCzJbh1`2)M(eSpOfHb>#GUAJ_%+{>7B7N8Ul_#Hu1E)MP zj{((exB!QimxYc)iG`#OPWujC?YRaOH;STdl-k|z+xqLFPM+=(P2I=|Rjr#}B1G=* z=)2uF3mgi(ekchT1mBa;Q)eTbZL`C-p}hv)mygX4M>^NdTMtTP#fK-BsHw3ieqoPj zK;~7{8JA3NYp}`v>JNmn=TR|e8yH*C^HuQ=?mcx^Bk0f2{SXlx3uW<(?BBh#P0k&lH-^Vak}sf52`nN5!`Kn%dDD$7fjuEo zU0b=F5(c`<5HI=BLx=>3ebyts`jrjV0B5tJsZ__!O(bI++DPSfauW5k<_cP&F;lem zuA6F*XbOxjsw~kRA2^{gZa=8Xn~H&W*6gMjw?c&V5p;9erMnx!dh6dGg#L!(4`JKG z5hf#2*7TXlLb{@3r60-Cdlg5c=PfUVkMb9Yvu>NVqITI*qU7_ zs&j2l9=fDN;~kQ9*F(ax;C_P;R|UZ!&RZp{?=g78O$;q%2k4zwK0ZOF1u0v1*hvyJ z!doabam8lYw3RB!@HBtTrF4fb0BUvjeD795&|4vh_OloE+H^M)^N*IUFU3u@N2_pC zZ<3EKnj1=V(EYxlm!aS!c|GHL@T`qmkEblWVkyCD?~WUI9FfPg?nzYKMJPYF|7b{;#&?twuxXU5jN(RVl1!0PkS#Bu$dqg%^XeZ3(5*XQ17Ame4u2wa5X)3#RGrn~WPQ*4K((3G7{mB+{ zVkUQ5n+RIJ-6c!!xgB}k<)IR7!B6sTxR~h`i0Zm>hQZ2@D-YZy-e?%#Jf|7>@Yl9W zwEBj|*E5fIGEnvjQWNj#>neHs0gaC&^`3OE$Wz%vuXzeC)2MKBug@~JIUt!A|d zRSsgX^(Rcx)LHk8{q}Mp&txYin}I3khUwiz)6nDp_I8@HatpAvz!#2$(9FKzuc0%t^69 zJCn#*O_KQIm7C#qZKR%U1W<2_^OElnqBUTWXbzxJU>ENQOMJZ4Go6hHsqa#hBP7IB4uYINW^s}~o}H?;1Z1_(oxvA!mMTGPlP zr!&({iOVu#84){VzYK4lx%x@4T9*NF&rx!V3=xyzz6O7WBpsf0PSDJA<;hIVBFrgB zDCr7Rv7T=XBoJ^jw9|rDv{z~KSf@PA_Z9#PaV9wO1qU^Uz%4Gx731UG z6M^hQisSQybS*~I2{b7NwufIPuQe0bW-A5$MV zCP2|+sh;q3l}MX_$b2q=?4=ZYJgE8jUGt-#A<@Sl{YZQ8s0J7GJ0Nm@I1ktV zL9(K&B1;_<&M5}T-{5(ng37}K%%&iEaN^XP1veVmFgl^2u{_R>9;ZoVAD*bsP-{Gm zgriM-%_B5}3b-zSJ4!?m)K7Fl#Jf*7l4ExhZBp7kKBjtOI#EF|#*{b{G%ydN@1n+E zIIjdmnT13Mh}_;ytYc~?e0?(JF%ZVZLmGOjqIk+e9gkMPk5nRvaY>jFxDTqh;WV9qH=G>#<27N<)svqV#ry2S;kH*<@-0>Hdsme4?UKDP-+Nzr0UUVcPm z73;di7y)}z54BW~wq{d~c{E2|QzU}@?DIGzsEkhi@C|%0tK6dZ?YP7KqBg3oJmal&l4b z7kZ#b0AO3=!%Tv#|0eFPRg=Yf^zI>zTH%?)7$5jVNt8gz0tJDJ998U|t&2u2Yc@A; z(`YYyZr=|^k4}<6Djp3wwsAacPr<= zGBso?s*vyT4PnECRkzJ(DyNV+P$T^=v#286FoZu1CKfhN{G66u_yDiJ)clu)zNIE#C`4f2C(Hw=Jw8bAo6CHMM8%ZWd65`~# z06<*-un?xeRHx4SiUjJx7Z{J`Go$Ovo~`G^gv9sKj_s6}L1dC(E!o_<12q+zXD1Kr z45wMAEwn1yL5P4|IiWU?^1~aoMzk7yc#?WaC0zig{YPMY+TT20wTbI;r`#Io-`pWJ z`LM|W7kd_hE4?zj5-6szq%1^JjP=w3w%x(tN>ZB zz7#<_B2Ari0i;`Xh|EYDz51~f?>_NPEXh9w(xBQ%>^8O_+}a~lU5L<*N{YSBmTq!Y zM&8#}2cX$mVlRdR#zqHuF7vPuX<^NCOG#8)E}U%QMDwb-J3ouD2QRI^o`KsOqi_3; zF-%rO^#XF6ksnq_e>>I&T95Zl{{T=`Nai@kl+HX8EMe{BU7O6n5ygeKT^CJUJBy|GWx!d93ycEMZUH}oIP1U-{m?m3NXnknBMt!Lv?jtJ zxY1gJ&YMmxALhp!g_4~S@mbi}_NCeRxY(j!KTCnm+o!n^R+5Q{93Po}m8znKierAS zX{4XYLQbh=c@MKYAb(X(BL&>In?ExBFZZq@Sa!|jXQ>~_Ed?YG4qNSmgM4u05mbM@gFOLbaT@0Y4`Eq#jpO65&22u*tIEla zyn&nDF!LzU042H{34+fFa`EQ$71Q=o?VvvR{84w6#W}>V`u98QN{A1i@nhkcp?Pdm z;onzlEVh3wEdOtGHfF}}7%eb47$smzQ!|#N1=%l0e~uBUu*qS8Cw433zxyhszJTOhT~e_Bl1e;?d?|F8KH3qyL-kqa z?1LFHdk;{JW8_WDHiN|9C>q@%%EqM&iLDAzi@Fvw55cNcJE1mPK=17gp3A_eZhram zQFzVwMIpyVk|EoD%*D|{8?XgHoXXZi0$)xCcK-Gdsx2)e9t-)(o77rv?8H%j(t2Ga z^VCSDWVtL;o>?CaIMg#>Uw_7jfwfBAHI0fzz2w-LqZ2vb$;UT>4$kJeSn7tLTegkkyQe~4IWJopEn}U zr49+blqr0ve9Vj%Ab~Iy$|{Z%O+gto99Sd1`Q$dThm3fx z2xKs%7VJ1AxKU?OnF4TrLmP7(LZjGY7$k+XfGykqm`|_hlNz63kvU>0Cn> z^Fcy^VUC!eo>E_Z60t#&%C$RJM|BkoWh^v2nlbIVOXF^0Ir;4#u4y0o%dkC$En|El zG63x?xq@pm6|ad>?C!!NXVGmQWXM;ikb22*Z{@H6a@(x2gU2Vn9DHAr4StoqsL) z>#AB5VhS3tNhwG7GLK`GS$>-a*i7gzG@C2rcYM$RoVjBBXbrn6oB4@N-{DDJ|12lh z&zVpA3vTD(+XLp-@U{&%e?q-eBBk6T^f zQYefBWIsa(4ogALwhRy}>?(r0qb8W-R0}*Rzs>h|7gaV0(G*OW3kdV6Z;L|cF5TEH z)@dXp)CLZ$h49q!M#UTsUuq5#yV32NmM26WAYqr` z7Au>59UZ$V=K zCtA!n)JjwGa}&_TI1n{OMl>bS?po?4)1)4%hlxltC)mG`fhb&39oVU-%uAR5YBCHD zy(V2sq1G_0*tX_#B1&Rz!(aX&#a3zy0>L#HV9sA`$9x5Xj68&$<^CtjUtd6eC>2Q{v~(~?3ve@-szW>u;eYe%@e_8{RX?LSNAJk zcop`8pWXoWNM^-tbkEWDDEEO5yqeJc!}au)Y;2w6ox-~8Mgsz1O$DyuY9_S`U{hC; z(!Y7+f8B9C{xzi~C+N)Q+TSJ_gYJ{k=u=vR-ya%Y^DAJljdFnIJ}g7kU~-kfco?$Z@DePhmz_Rz(j-;PmZX*2s;q{i6Y&94EVTeHlF|7!WCHgUorg0TI2 zByEA^U}Nt>XKUy{XK80_@8o1^>_X>kXzO5;dZ7nO3&PHx_$7?~KNTwINK8N2+5b&8 zVG#Rfp?pKI|Ak8tVQ2pjr3yRye<)Sh+5bbS!p{CbsVc1B5C2c9$~P|M8-ntWQibFH zpj6@f&P@N?6|ki-X1BqJ)OAMl$AYlmp;h9L%~F}w0=AK9Rya=hbW{|CAq@YAMNhYm zO=yCtR2~>1jBt`^Jj)gQF}x&+4ro6rkNHjpgMgR<5Lprk1DYX-2EBxwp>DsthS_lj zVGg}P``D~Y9#V70$>qb5?7&&Zg#F_Rh+f0S&)n-cB7jcD8I|z2R0ZSCw@!Q%O%X)F zDjyk%EDsNjLJLozP=p1-(3(hWWTxI6G&~(q4PO>BCPCc&kJgtB-(yybaZYE|#no{R zCK_L5{T17@5(+2+gwgXGT5)!M zP0{$#xQ z1WvOeg6<1rwtASOS_jf6H3*}>3&o&L6Ee?l%dNW@g9<%0fS>V2$*(3nhjIIp_ zaIz>FB0l_-NYoJC*OHfisu1E^wq&wuRNbY{fsKGyQ3Y4gHxJKzE?YxB(EPXM@sxbP zf1oMZkW5Dc*JL$gJ0A25@NHyj6*XK~viXbVa8ISas+*UwkOkj?>N64Z3Z2h9R?a0(hN6avBC#Z>h^D8&6#g|z-e}0Opdx80 zm{^*au<~Tb1ubE0&zVs-KEM>KIr#ZoC$y=k##$4pT4+Um``Aj?P7mU#IJl$BvqRI< z9pIPzG|^C2z}u*9#<{-s^N*|J*Yiqf3*LH@uE#rwjM)xU2M z-*dTt+$|jc54l_ZB}MrD>iuV2=bzXAf6v|WPbvNvcgz1M#ech7{=?n!uR4AISN;Qg z^gqh<-|m+GySwGTs`qd5#edc9zx6HufAuXO-^Zr^uuT0|{r+3z@_!MzfN-$<6FmW8 z=SZzP0!Ia7X6NGgr#V)orD=~hg8c1^#=>G`X6>x%?idjuKoe!P85F=n^^DIlXw)I- zBUr4q5Pf|pt1izzA(*C@vuyWX>mSRFH}BlPHD;&o>v`)TdhY|*TwNJr?o=`xA|dc(`nuPZy)$9COB4@m1m;iKJG?pyQ=HXS-ou250$ z9kSSWZfxjG3qKlUZi<*4BARsHUKMg6a@O^e>#m!-IAiI{Vn(?Osq~Amt1nt4iyzMQ z9RLP@?&8#z?b9$mthdlc|)+Fl?sV1o9AE>a17!MOipl2oe zLnG%WS13b~;smSj<0N;U-J-&PK41&1Ql8s5QWCPlYS~G5H5b9U@OUZ)8T0iH(&MSq z2+PyNT)-S6--3H#$Z>wVrP4!ZyYTw8?t1mir^u^Z&D^mXZ;oef&&gXT=>ngoNeQI& z+4=dlZ~iSm;#y^lZ7eluY9OGW?Ja{G@e3H?S*|Jd7WTJ}zW|rqp4eFD=DC6kj}{Dk zAH8n-m7C-}s6m-k%NSU*V}#XDNLgXv+12d=PWZiH@0f0ySGHlnfN^;_l=qQ=5tvW} z4#2Y)H|d;)caGt4*Y1+8TuDVDs?xrFo_$nryo{Ck>D}$H{)_o~CdAcYhUYUQC+81g zpziwqixol2VVSo{YKl(n;THry^y-dTZaT6Hkx)hFb$f*-CIv4Rsr1O!77rY2eyeJx z2U9oirc&mO1%9X&O9fGNkWkq)OaVMHDBxQ9dyITx$lFVL5L*ZS*((h0ub$+7xEh)$ zjwNk{)y#CkCJM_EVu{5doihStvq9K=53f1OX@qi5r7Z~-gF8#mO{jM81~av{ap{k^ zQ@P4*=e(Lx=Sd?9l`M;6!HjGZLtjXs5SR;JC3D5nu@n zp6C~8DVW3gSr1~ph;YZYvJ^ghNGjz56D%V?gIrKA1+9Tfo=Hun$~gUv8g!m4)I}pt z_qqwe6r2aC(^ww;Q=kCtOP6W2dVa4QgW4E2w^ZU_21IMt@oVpmwYv@Yfibn7)m6Z+ zYAdYR0)2yc$hbiA<-i9@P9MU*A21Lnx+Vun7aMM!EA*2}CE!;=lJf0*wQx$8qL8S! zmfiu=HVh=VxGNkiTqWkf;lOWHd5;k^y9^y#$7sfOyq;~M%6@u67TQn0OP$Cv_|7I6 zB)-xTFhjb+hTh%=@OWBT&j%qv4a5{sOBYu-qsT#&@Ju>j$P^L{n<@fRen6&{kmhb6 zp_25tSfhxfU5(Rs&=l0NVJZy9ir8FJD?!9SVe2HoKh98+3qf$k?q_}=xw^nFB$s`B z`Jb~yBu1fqg{sV!Fg~2-yZP5^@ZXgY6gI_l&M56%L4)?}vz0(sI8rie#4*Nva>iQ;<`M%DOvVI!NCQ&F zs8UVOGz3~nw{PF8l!U>SvY16c96g`D&a#)GveG+>qb6m63KYw&5&(?5yQ{D`nA*ry z;<rnz>=_HwYQ>{2$VNS$_!FXJ1|o4v2j)Z8xp12qKPa<|YYT(`?Y+$r z%o8^B7!u_M%r)zk+6Gy8qbY;P@9r`NYm|!|G zH~pY};E*)V@SbLIbG5EeL*sdrTw3Js-!u1%79}D-gRF$bb_0~|H6HP;V}TluS)uf{ zMGhT}6a*%ARjnRDsMA+KL272gZ$9(fhey0UDAX$r1te<|`}vASuj(M|n){8$WM?$3 zPLG?8ej+G?_a2s8u@vokZ1rwiowFli&nheph|re1A)a-n6(?NOR-Cp(v8uY3#3Dz# zZl|3gEY(+?j)WVzIr0Z=y72N-7D8|pstz#mJ5RN4Wcr?C^owVQ8Axh` zf4n77x6{%FQ^dumL2Dsi=x2sLW{I5Nao0!uL7t~DA< zu||`$6vjXlc9&Aw)1Oh;mFJh>uX*y-U7M&7NgM)m2_;2Xl7mPIBcnz~Aurj%)tK(+ z6TYADx(EOnC*Oq4I1kjc#HhNI%;M5l6z0e%YUGOG#+m$NZVY`WjQHuia3C0m*os5w zB1yJO)1;Uta%Gs45obxdD;AZVOq+81t`E1|dm&?k$Z` zf_Zoe?t5IJ;A!6}-ezik;!3dI;~y*;iNY2-ktG1T$I5#lSG;M4g(;0Mz8y{hO{<;) zw{3ZrhZ1%?IG66{T-*s@U@CzV=5Bf1q)wkyJnBq=VR_8==!6_FlkBJu`!4`tlKCE)k?go5!2yjcRyJRxF0Va>u8de3} z!$Lq|40fr`wgVJSv7B+KZv*^T%U4+8r+St-Ux%H>F#S^GrK-85-JeA4q^hUw^0PYOF6idJpcKit2V;dfY}nktdrM z92Y2!>@oqWLd{uw5ktw?ULigdsu{!S3m^MWS*)y(bUGsk9ImaJ&T58AZMjP-7iq)J7ba$Va?>3Dm-8cZ4b*6w;F>=_?J{i7XjUq?xz?cj-eYt zCZ>5vf-ck8Q9`?b9g3DsNYj zN{mR0lM3*b1R9{Ni;o0}V6w&TWW3tGpyf;k(|%eFs~OQ7U40*b3UQ_#5~U77rZb@6 zU}6E9xz;48FMb)NW-G{X7=>cp*uSe{%`Ar%V1vyZ1E=R$B3-!xTFi>bhYP46RyQm} zq%?#QP)g3rO%}v1MdfFPqx`1*D6548s6oUl(X?_P=S7xD)R%E7{eWe?@+V7$T1;5! zwb2O}r$9bvW339?Q-UKNk)uLT{{C5EI-cTDu|7Ep{rxM+<)k73d)a>|C}*Imhz(`` zo2P^{muXsDgd9O>-q{fef&tL>m`HIX!!Cq^#N?R}x>^QJq=+*MBt6sLVT-?_aIR!r zx{|ekl^I|})W?QcHxElK;5~Xzo5LSb7Q)wOjr z%TNIUp(<|MP=Go8FA@4q(6x9ul@FCo?@! z4!aWm(L`yWQ_x~yA@H0}Ri8Z9ZW^doJ_R1UUwX{9K=oU^%uK89_nVV~DZUP^(gZD` z&PC_5x#M?WCK~4`wG8MJ{$R|MeHw14IS6#JoKq}5o6*J8%K5?hh0{wdUtqs+L+Fyx z&Q9%+Zx83&;2-s^nmd5{{Zk!yzAhx5IB5l(5z`JCC=Zh39ebUYI})m4M-}K60L^B$CpkWJN6vME#|YGYc_RS zwnXtMQ{6Eq1E>_RAUGh0k$-(slWo5Znx8Ig(+;L(>1VdfcNc&wzd9~`a$$M3lB7c= zzG}O!=XYN3^_=Pj$jcHJKfS6~-m15;9r!1MA&nap^0`3@xm81}A5HnYXix|6yvb~XpFuPk>JL4-2nG7JCXUo8Z zYHl|@{2}wI0)wn*ZZ8oBn7psGzvO$`xb&c3E|Qmf)T#QW{_CrRmWss;0r!s^6_bPU{}4Z6 zWlbcqpawK)YbPGCyM1R*@}!cj%gXK2e+b4+^4T)~oF>&Yx84jVrcG}j z>nPXTtit3sEzxR{oRh$Os#)#Q>0X(^4aUgG1!+?fD$cDbiQ9wc8Nm ze?2~`YEpG;3BGv){&asksBc>Pb(Jw{FgK z@G9|qMaIc)X7&rb@1!z0*y(n5GqlmuR3|oUNuKCv59)E<<{r6k!-G0+Ki7}t9I=A@ z;s?lc?%Xi;al_O4<4fo$o7>l!3#Ct^e(w4AcLohzFA^O`1Zy^mUPhShwX@fpx;eIl zx?T~uQnX6}=SJ7yV7__K`6!_XTm6Wf%2jJBaW1AX&`@+JD1TSjo@sAAK>Ur1^I0BZ=2pt?%;Epg9wrIfR-_dmZ9-o`(-@?h_+|3=Emb zK2zzqn*KKn-&Es;`#N(l9{p?_RzN@1F7OFeBf0{m(b5ufMejOWh$b1Nuy^2hb}oAA z)Q-NA?r-qetb#`Bhkr6qT4yX9dLhbMr!w>}jLl8`!0p_%O)>p*kw(DdE!DbgXOs;{ zEcxF0sP^uu78a&;sY8=#bxa{>ld)pD$LL5t{h(Gqz_SaB1J?KbdMjf{9zgEhVq-lQ zKyqUvOFB&cm=>#87AVI-w#sMid~gUhp8q$&G8g_Y6LpJht#zRSG59sc9NW=`y+phBC7{Dlizt;koBd zqDHD_Y)Olna=U6%_ulq_Ce&bjp|Xh>xgA#>*Pu8W7xP;oZa*YI;4Q6!>O-@)VEj!l z9LppoadPV}C)_o$eRVDwGXpj;)`c9G_mAIZSrWMs)FiX4UTZ{JFko_3PH3c+>`x%R zq*1)lhYgK$C#$?%zs5ypl1r!fNW-{jaEBhDAz z_mc9o_(JQ-NEKC_737gKB2`szCj7!`a_0dYoV)fC;E#HPHiJ$1mH8Q0_bA+scLn7-hxVan2W7Icf6Qgl914@#B447@QPrxAK8^Y7^`gu0xkFkK3 z&oOZUk4-tO8S{Pa-82jE48_nqBuEgM`)f(w)ASAwakEfNMGcce=wk|mKq|gI7C;5=v{4R$fH>FT=Hs zoFZ3FQ_4DUu;?AWD%uCSBdc^%G^iW=k8QkE?n!e zF<5-zItRsoc8{Rst&Zeo(1wn&H1@tC38$4~{<$vE15hGBigOvz-O%yTr@tzQkU(=7 zr{$$Ork{rY(Fjxad%TFem6-p!?9iM3(0jHu4<1E&+>SsoXkk|N`MXH~nlI;y-+<%s zE($pLX*ZNV;U2qEIN?TaPhBS~gpX4Fr?MwiWd$P>{;8x-o12UIgse=EiRfgmA4fA5IE^e1Q5^;mscD}{le)?*+7 z0@amE^iSE047C9MSf$3k5D_6ElzEIYkiPGe`zwrvr)cFZA?LdUHrT6iJ|foHn<@A~r19aII)x`3>#>oc2mv zFeka@0r)C#wEBwK71<<_OzG3H9Wm(`cSeVl87V)C=|rJt2W5*;P=qsw#fI1znGrx6 z&Sr^W&9IN#Cl2^3appzBYS$aM>Y;&2kOJ>%5dO-+{R3GX5#!3gFqU> zKo6gdCj0uKbj^9rP1jB*v0f=~@|6P$NDN6f;^!KUn%&v4usI6QGyyLfjfQb4wSaaa zRl|07$t`5GH+aV9V$M>?WE#qXwh^Fn!Ii3S9L@ZbPln1`Mw6j!KU*~tvx)HTgWnp@ z-AnWmZ9z#D-vs1w#pb!FN)RJ1wOXdcd+3IT1{EY)6w8Io!^`1*V2Z36>8KldG&hB` zAJJy%&=vjQX+KQnmuu%U6W@|N6Q}EL4`1TQv&eNpGFmMdEN_v41TvW-3Iad~w@+h$ zNI)6*m#PE}k390^K&u$Cc$5$Xk?#U4YHfOY$+OtL))5bG<^ z=ej9wNb(K2QO%mzPa3XL{3HL)pTl%HFoczu@7MA9$pF8?jv416c zYO@aGqxy?ls;9QP6EuU#ral&LA<>C@=da20_*Qrwg=d z$If21b_f({PxPWSG=`d@!ol3#!q2}UBQO)mwyJ~8_0t(zBz!TYK8{yn>ItVbw2k}w zXyMpfRN)RdZXp1b3TU&IZpNj&{3JbJ$HM5<`^W&y7Dcuok6d=!_G>wm9r$flg4#2N zeYVh(C9ZyVoGaf!4V}N=!T^a1#AK?wdWQzu@=sgnKn4L;E=D3Q?e;DR+uNr9zz^=+O|2a>fyj7V5~!xJ(JlkP04m40T#~dYP0-a8pOL{@ z>yLF$sE*$0pi#7KvB)-<-CPlMe5C<4KEufO)#vM*OR|Rq489kDuj?$c7H};>@Rc{^ zBv75c%fct5`NXsvN}~hCQPb=PxDHlK4!R)0k}#16YG{l$*y79|P_FNZoTD?Zv3)}wF4VxjuqQA!U z>HAJ0`hH`(ayYfW+}c^8<8y`Tzw$hPd`HH&@Q2K%&c)X5ejfb{G6%2Lv5MM=%dc@5 z)1X-LZnR3`>s)0I6$Ix^@@fa3Cs53#8R{m5P6Ci=xAJ-WS1>Q@z<;yhDC{|$9wqLr zE?KADxN$ARTgQV6bIvWE;kU?3BQ0t6*LZYSeRCVlhcA9kGbQ)mg%oy4?R3rcc|Bwa zi_S^;RSQ+RJ+Jpv(fZV}kc!0!QoE5L4Z(=c2YO3@x$dZ=Ehl1-4K zHvrHk%^1&WOGn388Gf4AV%>A|(q@Rh*X!QhNKp=AXL%J80mI3zt_zAosxlgKgKBJ} zr_n#WQ;a>@=yp05ol1x{%#;$M&|{Jk>H_pNA&<8udM`bF?t^#gbZEK{=tk)efG(v7 z!OIpyBb65pjRf{Msz4hbF*i&G^4R7*tZyaHt9Y(Rlzt%0St)%&6%%u*BfWRVB(Q6{ zqRsCE!H~>gu4_wAsRz2@wX3Z)dzo@ZaMwj&L0dMJw-^6f#G}}Kmeaht4A*ZA3kI>Asnl$xuHTfY?WV`q&(g)>Rfs+xaF{ zM^$uPBRFD4>{yk6KEsfGK46w#nH{^VM*-A+N(Tbh z$8uQ{?DgG!B67+|85RPqN*16%yRwM-GgTxKGvm?2OyFk~SRFU?jGx$HvKIr+UKShd z_CF9~u^(uJf0{JV-7Tl9fa1fr(c!Fy&@4@}%tQLn>EGE-Sf_CW7FlPL(xs-V+o=m4 zSd(Im=mzR)YU(Ny7b=<7!vXYbu0eL6(5fT*Cj$F9&4>BvFD#5V`;+iR`cW=}Vda$r zoCIryMl9yN4~O+wa2Ms;y_YAY0`MdF+{{hcVmDuM>O6dqLzGVh{Mo=&VvToYo{uKb zpnl#SRBrY~KM79xhyr2c&xb#W*K020=6Th4`p?9hBTWD1wzqM;ApxWw*Z2hV&&oZH zMC*v}KEv!4E{pNa6w4td9H=S&>ZqYkq>?IU7}4)qE;;Eh4wytVO7}qEy(8yy{!m1@ zqf(@RL1@hs?V0Z-6`dkIS@1k-K*y0y@ZNP&a~YBS<29>~Ka`pm;@A0Jm$18x<6x^U zC8JY(`n#G4Dr5h5Q!*+c`5;GJK2Pe`=gsBR7pOn6q+RNb76dBRzvYGh+sY6t6VrE7 z_+KkSSz6lm8)6uKGc`!BN-(upP9cYJ;E#QB>8!R15Y34O@mt^)lFdzZ^fXegS$yaE zJI-0y^Yw|eAmWEQGnr@Z(~P*f(e-l^6coqu+1(bKlr!_>*0*ynN0bb*R`phgQmO3C zcv&-IISGi7j@dHq*E?Q6vo)*}IIcG@e?ET!pm1$&Q3xCt;oo(mD2Bdu00hfBg{}>Zm(MbvsYFc+NTQ z;q#48y{WTVDy*j3U3B@Oyq`;vc5Mg-dImwzpvYH#7tcaIRc_|e9aOi#O91xe@UgTC zSR6xWVMye^_KeZRPw3WjynFpd@1L(Y`)xNW~1am813ZKGfoHIUI5(QOQ!y!4>3(r}YqxW`7Bsj+DMU%1o^koc9_4L_e=qn{6@`$qHsp>KXC5&oVBy*~NU797rm% zjzv1^BArO=XCIMfyuIiN={AlKtR%`}jyK?Hnvh4%8s1;`!fH5SkfJh`W}v!DPOfaa zoWPrn_V8R+TSYWFa6*&`#WJq~ZMykSs=~ zF%=pVXYcL0HFq6?{~EmoY3YDIo75uBl}&qRjV#n}7IF0eE)qrW#tXF*w`Bu;-r&k38kx~qMzx~#q=+ligaZI(b5X!DnnTJf%K(3++$wyiw`Lwu7&6k`Q}Z>xa^t$new||u?OoO zJCK3qJo_(M4i&meqx)Sme@LqNnxTAzI6%AG3%liVuovhej7FdU3^r;L(?W!bG8yDg6e;j1gf zwpS29-wHIJ2U85sgA855UBP3vFS&F^?!2h-s?5N>6poKWzffTZ{N46FuSFqP!rU+n zQfZkc0t#jBWqf)M3qlx1O<^ZrO5cHU(k`8c8I$g#J||m!53*1S2KNSu8iIl4F|KbM z--P;{?`)pTbdJ^2iOpLSDAcQJ>C4HRQ^I^5TIgM^*~D4NVxA=TAYYvsc;l!R-NEh3CK$ ztV3BCb~34wOGlA!9*jy-d8urdVR&w<3f@s`Vd#HFHJ?s2olhid(L>Q0+?|JMV~{#v z#oxXG`QvG+>~+y|=bp|Q_Mrs%x(@OxR8(SA3ZDOvv*-B)=rg%R7RDMZC$YB`%BDP{ zNRU^WjR$P8fiGVm(__#gcnecp4mWd|i~9?wU9dXAEc_KzM)68(Iekw{Ck}8UeQ(3r z#e4*gvXcc?s+Kl_AGV7zX3b&6AY@!_1USVSJRkXMXsfAeY4!a$Oj&EBDMW+nIxxSw z-EGS#pH<@o2n9#%I*gMPpOIaiC45*_`s6PC=+AHFtdLehDy`3s#JG)Ou>=>onB!b% zvYT05L^N@oL#AWRFsdCEE7URES0(=(Ed1fZGk*_DxM8;0qanH(=bYE9W|NKs*TNHl z$Nf?@W4LL{IF!e@%9Ijk!O`@KFb4V*g1mJ)g6I7LAeCR@xx3j`rL2vR^78|x;Wl4V zmhU&Ka%_Y~@1JyXmu;(t!VQ+k>ebnX!ej*Su-`6B9cFwfrsW2)7Qd!`@DiekEYDqZ zQuxZeBCRropHA`bJ8vd{1$vJsu$9)rx2a%|?a#KIn1auDq%>WJ5nG*hF#PF^h?-`% zG(^OQ1TY4lIpf*1Ea)^Ge9BSNN34F+NwOuSUeb0{6*usE-h!ljehe(Jaoij0VELmi zp}qII{)~J~tv&5eO80+A80|bz49J&{-2uCg61jLM5{Ufsg!M-Nl2)UZYM~t??h&5y z4wLkB=MX}DivwtAj)E9>SN{d*oy6yYbYxd}7y!2wmeyRMDsWGrfXcH*&2y03nE7V0 zh!3t=gY6>T?>#9p4&!3kn{DrskUfOaD}DkIsW3_myjRAJY%PyEJD2D->ba!;9nGix z!BN%}5FH-Ap*A2p146Im#wwow8g|n;MQm$?>M?yHlzTA7Q9^7kbeJyD^F#Z+a(|64 z36ST}Z0f1S(G0-J$&-Nl0wkjKgin?JUEKg{T;wY*QY|dn_F}Wo4CSfqJ|+$!j%G1b z^+-qj7E(r|7wIjU(xT6bEj zxfUEhX|~18vx<1{!&8mV>rhXMAhH@G07$2|H3r`mNqCjXrF%Cw?*9;_Jw=(7+^%sk zfM8!bpi6pM7TXJx{-rHPhK+;h#EFipEC{z9ybGRwQ!@HXWji&fY2LkT*L&)2Jp+@ z;FGRMtbxX{Fa`QA!L2K>S+cBXSx?Guh!rdcrLN()b;%WHUuP=6I3Y#Bkb=MLYIxT> znB{Qs45|AQG*xfs7>fR0m=ND8$j(8wmB>6+?5?2Eo9z7Q-Alt`%Ti4?gA0yrB5UkI zCQC)G{cz-skE=YvHWuFDiw`CtEdZ`;e6Q{gFb+?&R_AsMV(OEKzIgPfNLoI4ieYPu z(znb?$aF|}^HWqe9ml)qh2Y|M<+L{&f&mO-adyB415uo@&yl=T)CF{9iWD#eu^fHj zOVWpwN+v05TGi=JIaeaj!bBtC5I|RQSo9!; zC*&T-&-?A#CISk!>~wDy|Et@}KUCnPYoD#=W%?KQ5?`SY6rByR6=Ak9t%iA^0gHE%(-ajIbFVxzF&KoIwdn2Hcf3~ z6Yo@2s^8f4+f7&Caos1~C#oee&WP$_XV(Q}>iwO(ol8tlxb?7jV!#Y^xGfnKnNnIZ z*ry*l%XGvcwy477>8Sm5>w`K0&HH!;c_!pN+I-R`1|zBG&XAy@T1g~W`?!uw8~T|o zY?tWh^?uR+$JIGSXBKqpI<}n-JGO1xwmP<*e6j70ZQHhO+cr<{|3712>@n7Tja4^Q zbJjE8Ck1C|%Ah;lOf8S>6^Zv?+bu}V^^W&O%9@?pM;4#`@f1pL8sapJ+MXY-OD2(6 zUK}Ed!bu6-qiG)?wDH0I`4Ldf|EPBzGrBjA|H@cx+-)Wv>5^3iM%%f9P8`A-RHjmP zKEGN<5HdfU>Zy>Zo71}kHq9^2vNap6rXtrTba8C;#@`!ON%uZ6AO9LamPUqEHVLg`3yoBE*S+oT}CTnjgsR1PfgtE$^2Y3_fBVPIJS?i59}Iy71f;b~%FV>Q0`r(APLhVXfsyUC zjh!aw4|f)lK+N2mIJRTjXjZm8fFdFe8+{DVue^mRWGq?0hpZPoyC2w@3Zm1VIq4;? z{~7~yOW5RHwcR?e+<)AyCVBJwIFV1wF36mIm8LzUk5OfM`BnF;S^&GQ6mpwl;4caB8;fYX&HCGsxna0kv_{-0XPuKw$ z_6|XJ#B{>R*eCV1uh^PhJ%o8Yv3e(ngt?vL3tEnfAUrzaH(2VZS^zYVPn|p1p_OTpwxiSEdc^62IRj^Y@hbQAS}?}z5u=WjgV%+=E8UT=|! zr<>uI_`BE!<;EUhfEsRw|E<550O$J_XV(AjOL_OA`D#!yDm`OcKW(8SF{JlRo6yiF z&YdR|T%dO+&i#tT=beE&N&tAR9zA2Og1f9f&Dm$1pb%}avJjtm&G!{LKJ&(nCCFwo z5`DEIXbD(7*~7D6DuLrquui`Tf%XY!1p;C2;}ssAg*O;bC5f}Dt7U=ZTvv%{OR8%q z#jngCBZ9nDIo1=q{lk7dK_mK5PP{FKFLqW8aq^BHdiU^__yb(n2OanyS}hfK5F9n} z+YAGdff0sY#>Cdl*_@Dx^`9H*KYfL(|LH4mVEDe&>>fM@6Conn1seIuHCRM$$wfBz z>ICpnG$*vWR94a=a(!I9a3|o7SkTGshr`McSSz7O`W$k8P6ywwQ5JZcnoyP-R1#S! zaEZ!E)vQ7$3BDaFP{pg(Dgrdk7@_hQ zwPRr@Fa9bB)(AL4c7aHfQe@u5#Poi-2A!aSMHM#&F;_e2+Qn|EHhBADtW#OQ08f zAT*4PGk3qPU(Ix{E<5dxaJTvdD5#I>CORsLTW>q~cW0$*Tu^YWTy3dv$EKnO(UpiSi&~N3QE5d@Nkt*9DOt)y zI6P5?w~uGJ?tB&FraJO32ztbm_c~<*vj2CWxp%a?`m<;e$o*x=G;9u>q8W8j7J0ammW^$J%c)E-1zm} zr|QMocRhagSMDCyocZ*4w{ZyYR2wY0Dv!^BxJx+jE&24W&=ZKbA|6pEC*wyRB{u0mDw1f4KOydMYsqR7Rph>?nB-@ z^pB0UVoQpQ109iutiR;Rj^3schaOuqY30a|@+r>`Jhgo2LW-i;`Kd#fMvjEGDMp+F zaTnN0Mx2XqANVC>w%~GP=dSvtzRh^PqSlrh6j`|$G!Ov~b?%y-RM`Zp&qfNJ6(P}F zwK2~BFfhmT!CTS-=t9uQQso%)7a3lk@`#smcSS&{YF&vh$1ehU(`5Vjh# z_5wJm{^_z|!?+Ka3JGTu+Xq=jW`qENvC-U5FGVKMx2O+2D4w??dz$mslMfc(6wUZz z;0TLyIfivN3do+HzKy$EAN726r>IZKeUx)|k90R$=B&EqIAHnH!;hcsl1d-Cd>~9% z9Z1617_TGX7e9I?#qcn+hs!&(G9YD(2nFbSVM!2y()0o;2`VH#KLRsiQvb-7_26jR z^(>~%_t&;yu9`K-@Nk1wg!u=t5fxB? zU5%hb_eTVlqb=T?fw_-Nl zRUaM?O~#v=?vaFg+xVQ%zhXj0EW|kpw>>WQkb;+DuH{WPi)~!}kak^ZK0$lrJ5HoF zIFxQ|sZ`h5?b!S`Ru0>KUX!UK0uo>k_`d22cVM?aooDet|4)syi1XM)>Sm`GlJw9k zy0qwzLdyAs2= zHzSapF%N7Wmk`Y(3h&gTRhi5>S*+MorkTN7S2Tw9xC?^)y{&@DiIv`>z7 z60!6{u8s|hKBt(Y_cOz|B=1q-Y^rWEF`Y7Onf&j^FPpkt9WU+Q>FWeA!^AyTC#0c^Xtu~1uHFR$*Eg4US#Fmc3J>uCGcKy9UK zh3L7HTFsQ9p{scs5C(~Bs+5U#*FAOfzJT@6%J8{`F}Q^ zn{qzBc-^Cx&>@F_t>zPr!%c=lJj3`xiq^J~etaA+Pzr+cH4E0NW?6Q8LUVA4)j{KH zxahe$Sx_OHDtGKvsnrX|X;|gYSFes1xXHSyKanPJK+3mkx3Bi2P%CdvJYX-2cUAV^ zF;iEG(+*8cU=M?wS6~1bTxdJB$o*ZsiBz`^l!QRYsoVLPJL47FrI=fZe2#!iTps9y z)6D|=dtJz_+A_UKr0eL6#J3cA6VKTGzBs1@S#aE2TD-B}hX(`5vBP8R3DP`4vh+Ku z*D#3x*mtJsJAs~!3+>xa&ww9V0u9YmTb=L-k`5C6^lOlaD>(oz)H%&gkrr4NS#3Uz znI}8PF#Y+$Ibh?tVYf)yyAdpKt_hNH2NjO!#GGyXrHVa~9%6S)pp(_!xpq%@OE@udhJA4@KBxQS!_f(TmEPApqwhrmh>xl z^KJI0My|}{oQk#dp?7|Nsfn2Xjh(lFN&VD=R;y+?o`A{t7 z-xZPY*-#jLZq!4)ocP&)IPDsL{o3!|-E8H95Yv)^U4>CKc!LG;VdZ60Nh&3)64SaHG_S@uKP>xgSTp2Ysu5@-GctSYh2yL=4<=W`vw4uO7_F< zgQ((?oM>SqRRQ`aq;i4^(6uI0!_?nBg7cjU$$-~h(`eTy600Hqm|ipDf_%Tg_`dw% z39mN4``C9hqB>Q=)NtTiYgxwF>AS%D#Ng;m@>b%4>hqcV?#)%tgve| z3azuwfpQ1V4{0CH{tH{=?ht7v9P*y3VlOigmi0&)^3eAfi*nXNZ;_aHT)IcCN75{y z-jl70|KpV$QN%t8$X;`tqd&oAw_{9-UGkL?3RQt7g^is1Xek4svLRt3ZTnaj?gU6qmkDPU{mXr=W3$+Z?HvdQ)fa4 z1CwFLAPv3`Lp*Y)$dvVtB2U1PR!1+44I!9^V1`0lO%4&#d#Gw(fmI2XqQQ3&YY3;q(leN^P>$&FrUFc~Ml-r>E}Zz{<^%og4&Jd-YC);8KH7ky{kEhUfZu5u zPwKn8veNccXQrb1m*XS7)dB1t{bs3aah|5ov0ia`WwTnMI|P;DV9R~Us2|=#x&$6#5r`G4(@gbAY;wRnb{rk%LX4G)XA}rHs+mGp&vfPx=9K};2aG%+ zcLR9?y#D=byGC&rAw#$6{el_;dWZR#j;lcXF)(D8eu9g+>AWnYp}rF=^S*ztM{V&U zbQzQ$E(3x1XTS|BV?cY}mz@PnM=rxtO#vv`#U2n)a6}%`Nh7j0@PuJ2s%Sd%<`Rh8 z6VkOkV@DlbMzBkJRQIoGBl5+*_X&+5L)a|9#c}nXkSavbePF zh54+7Lb-b=L#X2qC@Rw8-tG_GdQsJK9o+cmlEKK0Ly)B*c{%W9s_Id;AFF}$b=ezN z()2mPicXzW#3Q;xFPd4+VK}Po+n31rH)D9S#x?@%4lx?3n?-geL@5Va06Q&ber_Zn zhudVEf2g0;Ln$ZiQkSTQm^L^fJ9B5%P4*x^&^=OUi{@Czk&1N(x`v#pkHbe(zjLFa3gbCDTM?0>O_} zZnQX}B}ma~*}D8|w#8GrlMt+RKj93}nqcbkEis|m-*yBFqN5w=?v*NN^}zXCA_Jou z8){T5CJp@@r8)Gk%RC8xTbn83J#Slj9*0G$&h4KY-+dm_)+~6&tCZRp%k8f7a6pkk zfVELVGaESlRr!r1#5v7T6_Q&D3KbyESN<=y!rb4Tkgm-aA9?z#@sL^O<)~?($HNTk5+5<^_WLg2>gzD`GZSGVCqCGm~FC4V2ax|vSb;Vl9wol6gDhrJe}XLltqoIC>ENl>xID3IK7?a?3Bu=}#}Nj=ZS@nE<=bn| zMWCs!4EuAASv0YsTh&x-h4A25Svw5+bc20B<$>XMqV)}j3X=?rxzT~=2Pq)28514G zjnG)}ASH;}pQsVaH(u6t>06muEhaYq^jO?gjj#r<0#J~&=XeVuiAsoqJk?H1F3@6MNJ(e*gx>t7whg%M1SC4b;NT{G_sjyq$Y1#{$ekqDByfi8oC z#8ZDj3xk>$mcQ{z|5Hy3IFS^<)bS*23MFZ~gmvoMD_|A>yhP=!3k6qJ$HlwE3BryY zCXMJXDK#b*rZ+!wqg-gXhMOF{a%Hu4v;Vt6`+Jq}*{?o{;UXY!y3KAyr&Urc&gY5E zMB5Z6CNAbX7gurqnNyTbLC%8E+hfI_??@6~UU3h}qPj%d`q4xmARrY`AIZ|D89~TS znMW#ZK;0r*q@$WD7b(@1hqk)X_z??|qg+(hK4_d{K2`whmrpx}Mp0DM>y#P$>~F5wZvL3Wj0wmXN9kb3wBOeiuXI! zsTr>6Mz|XO9Y((y-iBqeF7a<{>#!II{9`KJ`bcsv-36a3K*;^_4>;Q}`dWhgr>9tH zhh*DfG?win1}An9i%8F#&})>`oNcnMhOYD8Te^WHYV(uv{ZOQenlSC`%W&6degH5V z9nVU>7__THj4H-s^^v-gSX)b}*7&iK+EsJ;u1&5$Cq=Id32O7ok$zGgsthts%|^{W zNkq*(bNekNVAbNoytxMQ=dZvU&kz<0D;|Tx4RYut@v?s|P}RIFCMr2(vO(9#dPP!} za8D2H*L2&*Vw&vVMa9nT2r|$)PD#@Ng06-sF`zqBFNBG*2o`~9l0m%G*N6sc<~ow8 zcgBYhNR?eo0tQ}M+P}fi6HX|4gjPwa1d}0-oCNY`fZQk)-7ZhMgo+%}c0N5icInA@ z-9jC8r>!j4zU}IuAwfBZ+7ZYRSe38d!z(!*j{=~pOZ$6nUi;lH;2H?#Gdj{#m+Z?m z%QL20gN5TYo9+|wU`~b;hLs!?1BG(S+cMdrfr)zk-45ys{e49{X?N1h&anjjp6Kq0 zcu4$uz@d8f_Q~KXT_%vUPWd=bG%B`?#%}r@whYaRx^6#p+^!(u#zo8Y<3@N{OWVae zYs+H_Gpr50?xNW>jcK7za~d?ao1JwwB&U{o`5cp6j#Q(k?zsj$PS-JG$rh(5#rv+u z6m!A+bUQ?2HGUQK{Q)I>Xl?uAG9u(w{%kitfFCdWRuhpcSEOT_4gT*A-4;h!_Do6` zAPanXgO-7Vt;u5@$nCqx>#{Wlt`-m5;HwwtBO&$)SsR zc7nxjJDJxe?_h#uMrhNDc}hFnq0hpW?r2n3IeVE9CrfT*9=2Ly7>io&m@+*%i4sKu zkdLnaL-r=Z)V6{KsaLe%r7TnQ(K7KwHnDC!4_4}EwLTGA4hoz%{P#27`u!V+DKibM zzyjg%EM6M~32p(K!~r58>^r^RGrMSo6B)aiAV|>T1k|88IUo?0bBJGKj0orZI(Ea) z*?m(?G}8%32Zb_+XR5_T7b0%0&r;zT5F5i_xVxlgr~S0ujB9cxlo&vKx-|c8_!EQ0DgjZpE6ypXYnE{5g$Z#> z3tKf<0`#tYSs7cXUF{nqA73>Jv1>9rKzhaJ1*=+yO#XneS7t~hjX#X`}9^<}B`d}uvDSkGI`t~5V;o{~Mmhhr?N>8w>j#Pfb z=f)=JS1dyW_5vu`$7x z6It_I<<#g*N6b9p@rukhbz-gsJeB`?@<-r(bd-sx6WO4WYXT>1m9cQRG-4_DQPm-X zYmZICHeT-0M4MsSi*`YX^15&~CLOfqH-W@e;k|~(SvT{cHHZcXCCY{b6bjMAymrM_ z-tZh2t)8hFSQ95nm-WPLQ;1Wt1IL#QLdWG9>4Y(?9Eii%&@du;8t8POBJ9?vi^$ zq$B6#pQn36N8vpHt!W@@MxPf!IkXvs+8xT4j*}5!em<1>g*QaqM8ACb%r4jtC0G!2 z^U-O1Q?R4$nyPu$hbs5lFNDx)?=odFs~`-=nq%n3(f7eDS4ISG-b-849oJ*?sj!b6 z+`-73I(W3oO2?K7gM7Mzr2z-Gokj4nUX_8|U>})2Cw47>fA+BDs{ID3uC|C(LuD3a z{z0eTcx=CHo(kSw>*0=5QI=bg8*xexP58))Z6GD+YigiZ3uPNSP8;?F?dxtmST+Ia zLpFz$dL5ir&*;}IMn2ZfyO$P8vvce-?hi2#{X1^gekwpF-?)XYHka-@l!|DuCsh*z z7{=FkV1^IyeRDH>4Jmmq|Gr_G(Wkp&9YpyKS#iDUGqpAS5&z0h7G~77XW=`-=j@6h z)1ouq17py#uVwdw(0F$%GvZ3SbAhc>QUpjYjmWvTE>uRwHtuDW`it_L+rs@i4c?jR zaU@QI(^2C#-xd!>*HzXbpO~()^*)}1al6<7BxuQSHF8?ZV@<5~h5 zz;cH4fJl(-0{deA{tuTO&A-1*%`=8311J2Kr1k&bu%`YchHa?*hmAcB2cZdZBh@s( z#~qPeSeF%dMU^RG1|pto{$;b2FnBlc`NX+#grS)L3pWcHez5;sMlzkK^Hgo;Iew-=-ksXcG97)P8|(EGu`T;f^~+}0uI?N$YHp6S6LW+0 zGsghnAFek*-JITkBMiu0C=F^sf7-(^`zNgHmQE;5L#Mu^*R`ljehg)_2v|5m5xZ05 zgm*eU#nl_p->XSd{oG9TILBUT$p5*yAjs}fs&1n_c;J|#y-a5-RJMMkLFRJK;XO_3 z)1FuEl6g|9Iv70Zy1;h+LO#*o-0q87o=*VCy}D@ZnnrSdPepKh+Y9zRnU%ac&)J1b z=jxD$2dabDH~@0JF}{kUf_>)F^lQl#a6Qvbr@W&QeUVZ2{?TX4`ynH~(4sFIs*{Ct zVK_PmtCrIDS(RGM#ZF1TQxW=U{EY;q_eyqc6`D|6|R}b(+i* zX!bEJS?|j%SlW5rz*h|~Jgw0-EKamN&Vzc$#5Z^Ho=26;({y=n(*qg{Br`8@7JRG> z@l^89&!`O1;nss2WNTb^Oz8kP?^S?Q;+T=#2V_6VTy&$xaPY-Ccl&@KaiaSa z9X2@S>cZW03E=>tgSAuem{2b&e=V*+j|!L3C`?kv!!Xx(e@tELfmhWE{rIG#D1JG4 zS0pg3niF76k6CD8eX5!v=i_D#F z41W5%W?&`PzsZ&$e(Vn{m4^M9TrUuyHST8jMaNgd(zsXo^-IKTdIxm%Z^3d9*6oL` zu_+1hkW%+K^H9#jX>j8k#ZAln&5e?tL{7r6MSxN2t5(Kpx+TuY?@~FVi?mZr!hC@{ zL{!eVNidtp8cY=)Q)d7OucAbvwC|!+GX|uA+?3R^$jYjZn&UwQlb;=>eTJq|l?Ert zUp;G$P_-%OG;GJVM^cOW0Djw1xqcHh9s~ljpp*w~wK#z?o*M?;{k^rp5J|scWVv7b z#l;o=Nn--?AI}n{b~7_G$ZNmQW zTfHlkcDTSxqJA~!KrxvXm!fi4Noyy1HpRMeSuc8S)~<5TLI+bW=SGIVyd0Pe*}&Sl zOUh{!HrilRE8oE1OotQ|1?=t$>Lh;L_Jc=cY<;(qD%X%)6E2@I>tLL}B6sD$ zVC+r8p&Bi{Ply1E=D^(;8m$SgTI;(r(wZUM$y}FKr#?A;)H#FU69`Zozv8=7>>)#1 zTZDGY7(&N1&g9ix`Yr_SqC{Y2^N&&{btt6QlKfBPBWVk2U-6Y83>8yn-;4yi`;4Y? z29zm(x`Alq=Trz#D;;Pz5Sy5qSUs_921PJm%-Sk7|9S$5J#H|l1y}px(sD9^<~#*2 z8xUeS2vB9?$@mkWGY_)4wuYDkP&^SFCR(hCFqG+tCn>1v1J27BVY-@IZv7X*vWo(K zK@>S*ThBU1uj6LMHW<-91%>U0#CIvl+=yCP2k^TZTtyvqToIYM!0{LN!#X1MK*ZiF zViqw*F_8g6#+Mn4RGn)nrn-lSsQ$&rKX`vjTB>mX0UyxpZXf#<<;@-3X?p$kGL$|! z#VK<|JmkVOwx4kpD++q8n(ptF1a0e#Ou7pj)O_qkG?K+J;l@O!?(_j!ErGH3aBGpnq#yzyN7RW>M~wz}}_8Xg!5dR7YuVqh3Ao5^<*G zQT_u+6A7a~B1_AFe*^`a0*bSY%z{CZYFXI5C9rmHw%U$$QJ?gqR4+ag=b1)|7EATsOc>Mi2|C@@A z-$Ktu;OB2+mFEbP=}-sc0I+!3wkr#umyGOr+u3CH{d5ajylXnlD>4~uI$%aXvECpQ zQy)WzJBdo`CKKTow65>bxpEmkMnka7157!6kvk0y9PN)kRR!EH@4KdK?^kyQ1UPBb ze8iGXnaUf%BTN>sK5T{tzo8^Tp0ih`Rdj28i$i@V*kEx zG7j`<{Rl~*6UBggBT5`Zp#CzG*3sjD7sK!AQZExBiIoW>59i^$`%x!IEVB(2sQ4%W z26bh)^t8p1Y3Wxe_`)HzRUbKwAXHd_2DmTQhm9+AU~)xLwMC(jrEZ*Pbe;6tZ88`vTFBl zVFHY;!*FcLI+cLocvMqLZQ)?=KCJ(g?}8ThX{^lLAm z0{Z6w=Eg^_wBw#%#Lfz>ebCqq{16mq=KY^k9X~S~O%S}SX^*g$MhB@6N$3Wu&LHEj z;v)iIsNWoEhpGe`uA=2_pHZ2jbXVflm6zsN=ay?xf6*5#2yYMovD#*uaK92PM65I- zgffB}ffYa=&9eM^5$CsFmAoDwJ;tCl>m7;>2`zz?G6Gthd`58U%OdG~+c=23{~*oA zzO5i+qBH!_lE%Z2FfD>+2()y|x5BrX_{4(T3E(hQYZ*xh3b7<%5#lmq?&Be44xdVM zz`#PpYN(rychC9&WXsuPYVSaKheaSB4;i)bqxi|?aere7Kd;h4EIHh$&bJNcCVMex z@yj-ua!p%}c;r~ZQ!%ENZaMi!|AnFpR3x)gtA{O*aCGeA99h zG+A?Orj-ZLN0D_3qY;C1(Aox<(P21oZV!1QHdg#vXLawOb8oW?jM`L4$sx{wmf)Tg zS!k5WTK1#k&=wA_%1i6Y45chP9zbNP|HkrQKY-O*#CI9xKKtnAg$LRz zbqoY^H7{&<@Z>ue?qozzH@25k8q5MxWDYOnp~jROZ#wpdJMX&@qpY+gs*K3(B_|$V zWRQ<2FePfQsE0PhIIbCA68@>FJeZiIoh{Q)-nH2RMA#bz&QYXkvRv_EI?;{JzN!9+ zx}BfaZ3S~9k-(vV=*erE5YMme!23O|zH`~`3=x1Gh=D}A7BzPs$f#8VMSvja6ddzg zKOPUEFXGi%s5+NUSmryvZvR(Y{nI%14QI>kGlM&T3M9sYP{R+|{XnD@RNpD-jBrmM8ON2$vun+F&~tnkhP z)!@HI@J31SRuj98a;8_C&x|i-bI4NuIohzLNj|X0D(%#3g**>Hi}Ozt^0>av$AodL z05wA;hSOdm+bSbFh0#KfN(PD9%e-nT9<+pHW)~*T3+1Yn*nmGa#6BDgxjXM zT<_ne(?OhSB}1$^v6J{F2SAww)hEYrYmco(NL+UIo;(HuvrmSpE*MyP!e*f7;DXD3 z4MEOjDm<;$Hg^fGbZK~-tkXl;)1}(^WqFwh;^NE& zr43PlvAZoWMrvCIMY(^l-85D>KRH|jJ3JP`iAd)#8oVPRWC3f!-FeNn z=gfe0?^Q34v&Z+DKLv;@^}lnl9RDM)`M)s2|JM3NTAFSfV)g&#U}yh^di6RD%GH<~ zY#WPi8?E6hj%8+Kd;mo@4%UfSO3dMl1AHy5a)F)d5+G3o3fNUv)SpycE($J)eIEBi z%~JV4UpIq@3skGDoSfh8N}0qNMN~WNojG~md-~N1etNv&o6ghJnTM&&q)vk^WyOo# zu{Z(B-u@ZTpXrdc%XUXpM-^$mzI<4ox6xCbN&)PFJ$L-#$@HeL-agFr$n7lP-Dxqy z({20YtzR%b;0X4iOVzL7Zi2~Yf8Qy~)Cmb+B}N~n=L(GR!Io{@MpE3a=>wy7_I_nxi(J)|-u^B@vUu^O%WA5XkB;4_77 zNAng`N#PAq=3P=-sDWX<9ZJFGAC9d^L#vb=H(llZxyb=_-(>B3Zq0B-<<_jYc9#Gl z&CVhEAE6vC^Uw7%g*RXNkAtHx3rLSwg&n_bPh_!vr;q)DWAIGI&>K zG?yX)K9wF{7P;Sq)^`&_IBII>rQa~lDz@|rs#UCSDgtWLZ(!v(LPS6IcRvU{1V?T# zJ+QA?y*IQLxcsq(n*vShg$l*5sX%}mDNiM(ZXyx~J7vm_gcDdntrk9w>;hW~5!ir+ zEM5ie-!MCR>Ud}q<7M^Z@nOoD6Fe5%eyBq84}ZG5i9UDOl;_-)n;3#OFf}4rt49qg z`H?kFm%IcEX^PmI1Iw!{L(W$*P(c>pkiDoCypaBYb#=B|RUbX0>}JUYIph%#$DFnVZ;buzLO=0vK5fZyaAbnL}V7RNaTj6kfg27eg=NTRbCCb^ln zDM9GdQwNE|NbLn7_v(Wgs)}PBS~Dfzs0(<;d-zuoMjNB4jv8;+L9+i`pj6W7 z7~MfUwHqF>4izIuavew&erv#w4em43q4rs|wMv_N?4}c};VYZyq=bgpx5xpJDCCWd zz?i?Kt`2;Wb?roFEkgrv?E3@l6om{mDrdx1HEGxfrL|J(bgWhXC{rp&He66d9JHO# zm}J>jqv&($Zs#4XOUO1G^w;~6QU{Pmibcq=m0>bfP31UMV2qVRcYL0E?rQfK_w?+m zzZeazaK^nD3^^M>_Xg+~VPj4~B3~N~pC5EmZsR^YH9@rjZ{-34%;?{!ZW&#!-f!P9 zpe&57{JpU^2)^Ga)L8nB){Wnzu5G8(0B^`O{)1jK_D$=;xNd#8puN} zX>QkMsjel^?Nn6oAtmgBOd}?DzC%*rkNK)%FbaBXmxq#6HKXW~c#a2=xL~iU+7|j@ zP33;P_1vWAdF_DOy{k)D(1Ru@^Wge4AJN2T-DO6wnsX*vj%Nnp43GqXt0xV6Wh}!J zJuF>cU-ra?emFf+sm14xsZoxR?C&%cFJU}yIowOOVDi`Ndm*YU@;M#DE}YUo7yE+4 zpJpp-+*&%P!C&0Tdm1;d80biiGiIL??WQVPeQ5*ve zeA{5KuAZ6eNFHHtkWwnGt?l2fz7g*uDFStC?Y%y7vaqKR8Ki9{)F8=mqw8XWtHP}4 z-hVkz%dHv{(P`1(Q*0`5S$-3cnbs+Uj2AkfjEtoW*a`F^Yp#vX)45k2J@g!(Wj9J<=3 zE?MNABv|=xSrItJx;FGdlye%6ZL;eo>i8hGnN+nl4M)%u0-raMiW3IDs2v=F;?^3piWhnh@mdto zzhqEzC{P;pFCK7zjbuq**g94utvj3~hD&n?5-UsGYKyBcHRT6^Oj}<}QvqV({?HIUg`ggCW8Y8n#RNPjoKKJOu!?N?Exs zbut~8I56ioCK(0M>TZ1%(?W9U;OOVAIQn!<-Mtlz-Fzf=sL=X!rmyAoU&uIM`W=4Y zv~HjiszcD@ubyq=qi<7qEy5+PIVyN~%b&4qw1SIkhVv#v?5G%U54mdrUb!L{j;ve$ zl$f-P`*aK)AZSJ}U8Ra&S4eb2;H^T zu7!1K)}l8xri;?_&1kwc|IvT5u$Oe$t^sPYH*fyOKDgmhWz=-2}^5zS{tK%#?o_Np`MG;X8C=fZwRUBmb`wg|B0pI}WP5MC7(k&qSW zHy=Q?S%mgnx#T+`bsd4}w2TplIp#o#oNBO3KxyLTlfkp4^H>i$PGWH)KEH2Wkf4|I zU}DuP=PtCW0fQyUS%Xx@9+<|!t7kI)x2RjIhib;T+Aw+1Y|C?U(gY$b$Xb)KR)nUC zzG}o&KSu1~q8IR`9vDw0DO=g@hZ~y}H?jXBI}C+S0_#Cu!*Ls{yY=yC^HAYEVbpdy zTnt_gOSUs1k~!=w>?&7D$eN4vQ#?g~%sF5>+TN1anwU~O5D%}Htmb@S3}$SR?EQAL z{#x;gB{9m$j0W~reG#Q2Z5c;3gXzs%GP&=ZeYHo|Yzjb>EkoDXJrryvY5@bxh1mL% zt5m^kRgYk}!8tEsg-j6ZZ!XOH0&TT_P5|2zFJI;HMh*JkGOXfp9rLV4e`Gm=qkd(> z=W?-gG^^E})5OrTp&|Gh?dE5>6B~bao&NCfkK5+?9suL!g*#1UoIp=QQBM6}4>kRV(7RS^ z*!`ybN96rNHzHW2^9Vne4x#jq{O9{|%7}#j+JY;mUK$bf{yHF}H>^Cw>wVhnNL@<1 z{ySK2xMRI$+@MoUswSj3ux6#!Hj7z6)&v7uxtid>#ys+f9ar{+r?Q^-Z`@ z!p1V`%{Qr2coB-jti|RaMDJ?aJhk?RH7HnZzr*A!4=K3qpnrp|dv2k8Y_d(CBX`_V zQ@t<4yk3MYJOwez)mmzrozW3wl2AAs$% ztaM8had`CcZeA~KMW%3%OLExsqD`?hZaTXE$^>UtV!%4|%4;&JD&Mi2zp3v{ z3cKYIo7+5G)Qe#R(=2juwwpKAm9XuYFlC{bKk^H>>jh~x^$R1M^+R|7Si@ABE*sikap>^-VCQQEmpLe?ARy{B}d3JBjf2)zIgFSSH z$a}_OvSEVY8y095BjP``J_sjgY9AK}DkujB!+&QDrnDrT{>5B%f2h|UgQ@*hC}3xC zOf=S161#IJ;hc77QWjNa7^;&fZyk((Iyte44hE5=Ocq8UOh5g4^4i(ZdK%_k;`VWI z{yN{aiGN{A#wZ{4T+#RZyHKXqo!Y@e8+D<(^Z1?h$#Z5lW)9B(APO*ulJgC?x9$n7U|t2rc`8w4shsWwsCSpbVhEv@E#Y`t+QpHIO3C6s^h0M3fO^il~+ zLL{tAHNWrF%|VmuH)abL*}TXS-t|v~GpvA3QrV2+bZ(}wTrx$E`S7#c{Ws=F_8#c2 z;KQA1aBw#erq=ym(}D6Zy%pT71$oGrZ>Z@xD2@opA$9#C*TRIFWR`h^Z*XzT62Sb{ z)R2g4*fORgbq`V9xs#2AV+_cM!7KDIBdN&u?UP65Te!64_@;Uu#(B{S^Xk2Ylq`@E zKPgkaTUP2Ba-~u9`^FxXbFAq^rIK)>$EGqX*=>ZS-mFb_@5NzKyH^5X*;;u7DkqfB zA^xQPhK(M%Bz-`dR0{o3mG9k%Un94Es3D-?O)_2KSsryLan|)WF z)?UHz>>yi`B%kGS^``GqGJBq=#xI`&C@8Z+nk%l=eR;+or23$l@OGU;{#ndAvB62^VKXqn{1dZI$JA1sLxQ%w$-H z>mX@0F8?9sRVCMQ!`UIs2KW4?a)UGGKNk6{!Lw`oL&#&RT%!L1l_r%%j~!9VBgsh- zYn8Wv_Pe2s@?H6-5p6780tEW_ujH~C$Z={NB)ikz1hgT~Axf_+^jOJ11-ab*h=G;O z=O}T~;I_T=oL99274-oi2LN(lp_lwddf3m}PYQ5-D@sQ)c;j%$luX@VBd*t1rBnbT zq=A!!?hcZdt1SekOPeP~Q_@6zP^vTts(Kp-^4)nTcFWPjom4BU2o_ z?R9UoHpoQ);hXQ@3QSX{;B*K;YU~i*a{pqZcC_v>M)u{ak%>oC7X|;+k6EFP-%)j7>a&mDVRAFP;6i|5zdV*(Pm2ysm(Xm zvj|L5R(gN~0b{f#02cvZ=7HA1>V&jUbnV&-J?R<;ZRdR%%uF54?^9P4sow(Z__r5% zi_J%K=9cA3V)U`?$r30069rICqFRH=q>;C*rKTu7z~AE82CMX>Y5;tWZtY+0dVm-?E*XRm7{95d}~4Tlj0g z9AfW*DWKG3-KKH4DrbtDE0M=PV(=adzb}l)Z?H1Q5!wj3Cyb zwalAjZ_dtq2C%Q%z;t42P*%w^o4q{RjAM@xzvCn}Crl_1n7~ZnnF$Xc&RKiYf(AQ; z-^7|fyrIUQptfkB0Ek6#9T-zw(vG-5NX>M`#Psb&1%Hq@*SS(rhI*h=Bmc-@h(4WG ze}S+3XZ!aBp+h*%kCvr_JhvWmAAiCh)x2(iz3e3<1OUeg?aRwt>h~vhY?8#~LSK7I z^`Hc`T&b`geC$1aCD8M(KFe7mPaBxs&0q~bmh7r5{uCN(6|*?5xoTr3&7jj#72-#X z`zb8Q=T~*Bq_j(?@HkdRa>y|hh-Xf(39Kp>o)U8rC}-JuAZJ=kLEHdwteD)91&yYP z=^QF2r@*fr3~m7vl9L|r$4BCV+gHU z0@HUEI#<%`!{2LA%BCIh{k)DTm;QyLvf#An*iY^QIr>Yo5%qqb=@TFnlv)j;DVqNf z$&>NSX0Ii7TfSS-6F4MVvz+$qbfIR1?(-50T)ZUA0e6K{k;0hE=-6~+ z$nn@In{*Mn^32Oxxm`gu>)SN6cSS1Tf3JBi;bX^JjB+DC+E9>K^^)%MsuA<4ha#!ld71@jz&%Ia#CLQ)Mj^mHcoW3;A%37&rpkdJjSSb z$A|qD&SWDYLMLgk-gTnD8^q>E_N4EKAEE_WELUT~v zBhyjG_fmh?7FAO8-Qb%6#9gboG8kQ#E@{%)WflVSo$A3B1?Onn*O*$#zIxsAP5V2r z=QfwQ%ckvbHuM)A{D*KYG2cCxqD+>lEB)V-Bq)bO@({kz8>RT12@|X6&3W@^MeEv;b!`fWXiR4dXCcgeAYp=@P^|Ot|CvWy@+jBtWZX9m1J-&m4S%$4#4)XX zlsQ;q)x#`?$f6*Fw2z_#1~ab~(@m-zZOM8{WrIxQ{wW1tL>xigjQuICw?-^Q9AZvu zoG^jaK7qyr{T93SnSvm|N460;1Zf%rrJsh1 zhV-5}_e3{`<2|tx2AC?X$-5aH7v$+~Liu|To;U4R!SQSvuH*}5 z29A}EV$(Qs&lC%%p|RA*5^sufA4rV%rwTzcGUW|Htp)jM0E=091;>ZD_`kuS>};M9 zac&a|0yrl?JT(eSO3cW1*hx(chp*Zp`pKOd2Z~hVwPU+UUUqrG^k}&mHgLh8H=yR+ zKa7bFN296;FVinGCe9We6eU}i-aBasN$%Rb9{c{l@vx_QO?ME?YC&Me=AX*@@#qbz zPUu9%8h$4G0P1pysS-br>TM|~C74&81_l=V(ssh*nWfu`!IpGoyHRt*LOZWC$z_c2 zL;I0{5iYsk#=;zol%c$#cTAJXj?qx_L5eF{^Hukh$adr-I1g?wzI6Y=2-RF@*lj0{ zRKh!mMQF?uGn6r{gCZd8ZU_*8ypC4u8*5~y_%a1705L{uy@QHG6M8+RMjqqN^ zMC5550Tr=Njwj^uC*%kx_2)UQE2^$g&%z6jEW&K z`pZx+qV;f8ozou8{6UWQi)bI&p`18st1x*Mq~>5ukhkEAw9T=8N||m0QBGiGdA-{M z?UGR25X??opCD<=pClV8!Jp<$bA5Pi9^M+Bo8h(J2qcXwPyFfAKGnlx7#r$4L|lFT z05eZCbmg~tLt@mX)))?MVty<02-y?U|)Oyu{LdLrw6*E|O-G`ofO@D8ms_0H{ zAJ3Iq)@`Bn%xUDqCYzVoZM@DD7LAWW=FZy?nhg97+!A&@#^NCM$E+jmAvN`iw<00f_9Uk zpyS!haL~n{S@0aFvll;gNWewXHZ5dwy2usp@Eo^>ntQonig7ciHV&iy$hxI=*-kFh z*9C3q-0)!(HcW{TrLk0mZdonvNQ8&~&VN-r4ph4|T@@}e4T&M>XTU#Zdg z3CXg0+cno9!#?o!{7cvik~WtPFJ}(B4i2040bFhVPXI1X){_6#TBlTZLm{M0%EKb3 z>qdxC@SgkYHx=H-)s>0$3|YSt9;f34RVH zz>tm;lB16kvTMva5oKU$A05-KF-u14C|$Eoh@{X0J%@@SrgD?TxXQ&dZ|Y7kk$NuU znuah6kC728IgO=LH5@8Q$KjDk0W+eug$#kkw#*NJQp%uSlQcpt$EJ9fiN|19r^ZIZ zN?Dctr$lapG2lj{1PGokmC9$JgI%v6VZewYE>I7Q)~iROjr&UwvNWlj39UNH6ntoy zvnn+f=pMKbXPTUpIsC6}?liFi0~=VRE-BAia1&jKiEZXQykf*qA=p2yafLB5!M6Sqy8=tv#}w>!zCv2GE}AI#`9*3Y+^c+Sr0_ z^$^pEBDb&7Iwk~nQcLLO6;eQI2TDLMmLNw_IIk7AazIxv+OZM-Vg=nZ^jI&c+iNrx z>kw71sw=9^vQ!a;L9EX_())+vpJ+z2p`F+(u5>EZE2gyUt=uH2bb*Zp~xycK&)^OnA+G8trOQaK7-+b%r-1a2fsd8JrAb1~x zr>=H{)ERY*%6K%Z^}u9hO7CUh{cPPWc9YZMq$4Tk2C$4(ft4|Iv2?rvjmGb|%zJf^2!`XN9Wi~7GYI+50tMg;xEcW?l zYNliBs73GP_VG~R`|;Gd!9V2r%lo0eLpQTazf3*4b@~T0HMEY5|PTH~6NToz2%11(%u*@iLS(3<5iX?{OS%0n9!eyqZ1Pcb&@1h7YRBv~BebX@iHjXT@1%{q-iE>!? zx};jlZY$p+Yt?fj3@nPVJSKBpj|%7;T)sm;wrbDpy=%nM`=YjB6*`gHF-m$s$9yn` z0+>yIh6c4;V)y!{S@eAxFz{b?yom=saxgB=wUK$;6gVy#-eC3j+Noj*JFd^pzoz+n zF0>SoZOpesO1s-v{~NKrCr@QHUsO!3!K#gF_$?7>Ogs#|T)b3!2$D+@(MOYzKgZvU z83$43M(anu2y}>@g#MGI02loNZur1n0i>>QHX6o?7PIHG*R$K2T7LB3PaZ_DJj@)< zC8;Y3#G!(5n4dV^5jISYHJ7Q)OsCrJyaQwPa0C^FrTl(BOy{d=K{1=((BC$-# z57h0Dd4nT*4F=1=aLdmcy;3|YIbhbUm;@)w3gWQT*zGzktJWLi4KSWq1g1g>w=S5m zJ~AcS%xrF2CM$il>+o8@oJyF8XhipSOezkKQxe9}eqL_l=z7by4(|#K=G~crlVR3^ zcezGNW#f(Q-z02Am@SG8WK^U(zL}}3_H*7XP45OXgCM44d;zx0Oq3{+wgnZ%10-SJwzDuDp%!+ z7z}Z!Rh+GIZasF`7-|AM0?47i6~2L+wUR2Y#^1-@+86F zpn{SpRtxoGTk3Xgi7FRrk~NG?CkYBJ$qNF-x19%tT`v-o%M%`(J$a%Kw*hYfbdF;dL}Q|kSa0gMAmgcj5P1!%UnyT$ z^{iX#(xad1EUqJ2M}Y;_?SHT5aORJ#qs{zte}K5;8nkz726O(GHj+uk@(KMWEd>AV za%p#x3#bGdPI9v$NZ=5BJ<1{e;im>2=1VBrMEIYjy)lI)OG82)v4bC0tHgTSH2MK6tl8iQ4=*$gwi0aE-P z`%paOuZI~VL&qVXX!7o`KXT@OM}2N>X?}6{RK#PbzVg{xX%PuoM?@?C?9sKX zuL?4zQlvlr#2X?%>-c{Wrp)Y3|EHOgi;3(1YwLvI=3pWI-@;Fly3BvV5Bz_^kE1uK z#pIK~+ijWkzU!vbdQ?#(YnhLx4En$_Ie$L-%2W>0e{P2Pu~y)KbsuCchyb5DOF0?N zm+$?;`Qc@>b9_R6Vo`+pMyaCGwX+f~Ru#uwPF?(%L4|=}`DBT)w3U%6R7>0}G0enG&^kbO0=*yet<*KUa5>#`DZp2!PvKocd9}6w=G`RUrB!)obd|h}nbIHO zkgO6Bp^>^7G0G7VW`T3-z{AmTr`9+6J5v(z^X(fSG8fixR1$?)8!jzrt{k%FPk@D( zucWdb=3SV|rVLx^5dSRX*YTY>0>?lf?$Hm-eGS8IsvE!`{2sBW`IZ&l4@-0#@;cfT zb-GCvR}!)iBy{2c&+NA-P7fHQ9;zsUQrLF@7UWZlr_+j}c7QN7bU06KALW(9sKcT9 zfWRtXDj%G7KYFz5>EB*EtoBXVfm?ueVen#GyLC5DL&qS6_!C{4_g9QM|OW6oYDD*(*5o$I)y}i=R+$b{ZO3k%qJq4pf!WB=@7P zSd>O%q^m}Ph1si9M&oNWaLwVG*V{*)O@^yKxX;#@8ic7x96c7ugi125z9%<;EjsBn znrE;2sWZzpNXj~Mev_;wpymc|o7%FFR7@~ui3BjFtd{hILQFTLB&jC`5O0we)s=;e zqvZt!lkE|(Ap%!c|7cf~`AiFOne!0-XW78VY$q=(jAgsVsbtjzusV!{m%ZvJm@cvU z^>yU=tzQ}lrEtfexVy^lV$W;(pB@V~S?uCPFZ+LC3!qf`SljU#n2>)n)H3{fyhF|` zR0WtRW`Gn8*xDAT;^8NICGyZXr$IKrA@{`74j@|caafD+`+Qf~%H)xsR{S^LhWgsE z`gZaF60JUbGbxG})y@Y7qD&_XflAD@6A|L5SO>cAp3c5)XM@jiu00h#DCJApsr#Y{CLAXod7ey z!ZbcK8Gp*qyEjYjWJo-)XRf}`vXak3`ZJ@yvVg$}A9Mn7-or=09*3RCTw=czJ0@T8+h2=3i-2F(g}Ov% zWqkMIE`dxidYj#=dNg<$+&|y%=N5 z)|5EWUzm{0%&eRoDMs;7)PNR9Mf_D%hA4`@tX-zojt-I8U3+kNrqy1i)?WzMn7V-- z$O*9%)E9TlN5(t43{0U1D*(?6ons@TMEuD^6ZRo6hJk*VV7XtUr4*Wh5A>q7Yd zOX4L9m`i}^r64EoeGlUn97lnG)B(!^DHlwoAH6#hm&pYegfoVL>g)WTfGjAbhd=&3 z*|)d1*N5opY8EDFHwZ}61SK>~r58jygbi#Dss!m#fhq+h5CRkp$4LL0ZHDalE}IS1 z9N-I3MBasQ1knl1)XkCI%&Zht>|MA5p%=sGwXuj>$T3a&15+8$69K+!eDc1rWw6{s z@Hgw52})aA7(eT;yuKk7UI6tA%UpM5@|(jP+(V(freAW$X102#y? zG=PuZj@#cL_X=4`1sq4Sx8m8H)ca_q`QtO9%ve5#jLU!C?p> zQ>v>ALRU)xZwOis%A5Cn@C5^<^+WTg>ks-R4AZ`^^&bY1lbe|xDqsm>Lm(Ho=a{f@Mi)jz7gSNBxa4?eT_z4F9v|L9&{tU+s?oSeU7udUjD zU*#-^U8pF0=YjDf%m55%PE29y-ucZo6&T@zh5zc8D#U)0-+}tJW)IK79==Infpe{9 z5!X_o^{P1ggV;c$KfOfWKY(%-`ilVm>xDoNcmDtqhyw=piKL)=jow-wU2#ToLSFIq zKpi!Ma-en{0zO~@T7`Zo5l-o8=Ko=m+AvqB_}L@pQrct>7Ui;+yC# zRt*XO<`wu`3F|)_3Mq;E6Uo&|lX4-UE$@Lbt1GYfnXIh<+^m9qODdyss&)N#r4T^6 z2>!{C(P4~PUz$N#;1$1#tsB}wZr-xvp^H;yHgnxdBrC^v?T?Zx`O>UJjPP(VecQ#y z?tTTHY3wT4PMK)R$v)-3GW`ILtGSBZwOI{7HybWTOF^B^Kuwgpfxr}I0h%4Y>U{{q z4yjM_zjyAaX#6ysmrb;biK7{u&CPIwL}{4crT!z#d_w}E7n^mQ4cl}{>zmbjgN=is zqKE%dS_@Pj1?E4>q=OCoq<180!u_UDJn=YNvF9_mDk^1UuWIBW5NJ%$e6ULA!zDHV zNf$9&oD)FYPIs^#W=L=6O`b;Hu+}bY$(nLbn1N1i%)q3md;F{ z<)X+?k0aLbJPKcFLR;42uVw&%hs**nuA`|2eB^>mh$7&987-+!5%FxgtMOvfEcZ#+Hp1zmqdr5Z3d` zk5eSG$vee*DF~)3)fmspnv<@vW`a*rj*ODm2R1`0xCE2@>L+rQ+>9^rgVbbz{aT8+vfVs!?EH!yk_(&m3ZCVHpkB@xmJlHlpO9ol1UAVl|0-gTzvxnD1- z$mUs4R=Y8zFCqci!SmQ&?s69x?vZFXk!IkNz!wd&s05F8!u7{!Fcx>rBI2qQ>MGud zk|58g7u$&|sOwVYNH(U}?DP8pgb15SSg_Q8{zh%Zv8qn9>l%x>p=US;h=23eaK*(8*a<%@UCQfGFFr)>q$n1U$nIzG zp=hF>Wu_vL*Naz4m4oJ$C%8v8%3cq5+!eVH1*VL|EpK!m!7k19jBIq(k^!uZ=TcY#A%I3;7m7Q33v`d$K7fer27{Gy z8$}GHdesP9hKaBWADz|##<#>5j@iJra-X8#*^8om^6yEx$4uu-gnI@*SLhvxI)o#B z>AM^~C)E5;`h40pfZNdEy$lhubK1jd zmhQ{i_yN*D6`==#LcbEim|$D+is8rQr+Mriy`O0)tzVZQnLM04?WE%s;uy9x;qF9A zzV>~5Z2KO9iv)r?)!W^Eo~7NRXN0Y-E4WR@IFb!-@?kLGU@t}?i|qDKtaDg>89%pA zwXGF0-STD>Io21)Ggj$Ipi)>RI@)2Xk|wBoRtIH=-}U7d(+s9*AWLF{s85?ccvxWi zi@*Bbr7bZ3nY7nLEYO=F>WRb_eU35XRqiEfWuLu^Fkn981jmQ|Ld%JSZ+9cs z&M%9Wm*4mRlYzYhuVrx}t;W_T$+&K$jB*)#h+6Z^v*uQFACIEV8%Zr=h(;r7#mQW( z$4G6S#HSoNHWg2svCenH5X}-)mGRibEFRd4i+nb5O^&q189dk8w?MAXF@d?k04qe8 zcE>l_xV=g>Upu{BHGBOGZeNGX$Y+qyZF5R|YU>Mt`Gc#iNOk|LE`qQVXnx)gZ(0}I zA~k_KIeZ&+S~EiSriH&tGjYk{c7hytXRvg62637X*~}dCjCBbi4purPuAL(t?=$49 z_6^+#=Pja3ZE*eDzM{N8<8wdw8DkGPUi{s3uy^G+`B=Wf+`dbVH5gq-?6v*$crt!g zuy`AQ#eS{7#{K>pdOWHY4P}>3;hjmK;HK{qVm8j}T~(u#qiZMZvff@2I~x$U*|jvN zkpdO@vzpCEE5NJo96Oto!=7B|y;{m<=W+I2<>C(LX#p`iyD)Qg?c%k_9uZ=U6_{+KZZ0 z`@VVCYq%|h!+A`+uuA7*K|jdZ?)m(;z&P7xDR}S+F@J4acs!vv=cbxs6IHFnNa^$(U-yIs^+Ks_4*ZpIowK2XbsVh zuI~KbrJ_(#Uit1b{n(8zm)&6G9inseK02r|GuSW2Y&E zo?3z*7V9-pby2(5|Y{DTn5Pmw5 znTH+yz4!R=PPTz!_o@l#nyj=KYNGa@hg&$Fk5-$4cw5*QI8N}V@ zt`d}?!Sm;A(G?Rd(|L?Pv|P(F3dylwR7BYf$+D;Y%01Z~M<;Z_xWixcTeK!zQy;uCed$?ppR2+yBeqcU~TIXGh zGS{bw7@v&&NO`U&S5G^VlzqPGA-nWv;&x|j2v_9SQ|oMpMKLPL*bvauQD4W!0V}mo zf@~(5mmJG35HX)BksKe?nSlXrj07vNv)DytAVV`trM_*-wAY6O^sx$%AFBrxP!plJ zaioJU!=0L3%Z)^YbDFfUW4LwLMme$t(l<^^91qOKtD5fG6)S~g{Z3RQwO7;j(c70y z_?7G9?pxd?{O`y*y&XYX(?<=AIPmG_PX5Vt1P`sr6Y7h`KB%@WKsAD*5L$TqJ61_2 zpJCCP$;u?&w`~5xg7O6r5;8-5&L~wC9?I9*>1(ie@_`?E_S%E)c8U)vQ0`9#NRB~V zLHXx$-u~ER=1(wTTi*3z@2jrs&sWY#KhLj3<~xc=_kt1zL2}p3laUj^Fv5ZyUu@ai zLqkYU?|k@l%?dAy-2b={>0;9-f*djqVg)p#$E-MFcb;LejIz^dD-$cKR zC3VryJnbo4%dy;^mVP&sAegLY(k!zj3d-w~c;&FsWNzKGKh$Tq%n0FFpixIqCv=n5 z4Dtky^hJE=4TLh%lknacp*9>wRScrizVaCN<5bV(J1k&FpppgCsNH#ODLxKo75Tn; zMMq4!>x^^B(7gRue+%RL)2>Bppm&r^J_-XEzB1&q=Fxi~O@;^(psCHCZ?Bm05)f#- z-B&_yuM;XEP^8wFi0>c|I&R3hwl2#?4t<&5(@&4_Wmp7){dH)ECfbM*j=4h#+fV6_7F~4Wz;m;}s1}~6( zVUl^A|3?0Ob&+Tr;{Z$5i0VLoj|hafRk?Q|ysr-yXRYqfjYDszI`>P5hrpAm_{KZ# zHg#U5N4CsFmADJj)C{GAsNvPqtjuW;3>>AMf6Ra*>rT4tu+18xl_Yk!k+$jB0J8ZA6F7mZ zQ!0QX!p+W#RRfWZO0M7XW?#{%*7o|Vx5*f7HSVkPPDZ1`R}dsUn;E@(O)SPFiM*djF!*^7BZ zc2CfJUv!P)>RW7vAvJF7$c~`g4{9OQuZ|tvBm*ONSzXR{ffxd4&HEc}wetg6-HU)c zqLuRBCxbkb#L?A~E9JNiXH47gt1eh+PK_)fgT{2DgfR;2s`J$Gw~;G1I5In@vts`5 zX-Si(y4|BQWUwLIG3))k`QHK}xMjnl`JeTn4OX^sk-Cvwn8GE+(%j=2BZ5a>QrbI^ zT3zK)rD<@8ZcU*l#XuUq$;|PT4I#jt2cYdo;0^r)zMbx z2{xEoT8k6osbfjv*p_GorlYH3(c9|$9b*QUx!xB+Z8~Kg0c2Ea<#QeL3WKSSp08d= zYCb_Ie?Wts_U6ckAk(0N0C3V&;`{xJG!A4IZjdR&lYcPvSWR{TZm+7Ss|z5yjaQmW zcZ!bjwE7#PAw`lpG_wNgI zzU~^lb?779D}2(VqsXoE*BQ)ax}~c=j>%bZfFKy;hrO-E~12A{%Xo@6E!R!rYfgNWlH3ilWewI9CP@B*NrG9YL)yzTVH zVFtm{E=KoQ3o2!B%t{qz6-Fr;yPs3GCqqeyN*Ykeh^Gr?#_KX3y*=qnDr&<5rrZ zFDb-hvL|0W)BGKX>;xY@5=^Q%6c95KBZkW_SE3M!nK5grQzxn*t$Kg(z06yu{hfo#?)i(uk!!9Idwl@bOrrHo zDAT%{);9&t@c{4(hBoofCh3~--rI@oHc5fh0xpNV>>_wfC;n7-8?NdeNaOAYwd|@mv??Ru(d; zZT_Y+)s8D+##;spk`}26FUp;4J)LT--)lRjd4xA`P0pwZ95a10mq=`Obd#!q`{zE~>xD;7vtC7#4ubD-8B z(Qy&)I#n~3Z=n-#Q$XJ^su+WWu@!t{tMAI_6Yo$(i&c+0_#$}w)~IhZOMPODf;(3L zs3KqQss&t%%#is%>xn*Qslicj%%8kBP5GBWhJs_V2uU$AB2io3mL-D8hE#^|@1m#j%9uaAodXyTJY_aM&;9zA)SN4s%OuP_`sDh6 z>>}(Y=FeLywYUfl3QtTAl= zqpM$R4#Tfa7ShDBvem!G$r)(^p4U5!a>V*LQMh;3f1_BFmc-RILb+;K^qI*F#S zTPeHIekO%pPHWQX%*=V#l^QGkMRP=cS^`)<*a|gBn@a6v1w$yFHB@sf4;Q4IeJd~3 zAaR7md|#v;KZS12ec%C6x|L=q{hEsse=o4H)kMzfWmaNq)FszOHD_BM=WzJKG4HIJNZtiqF*^iqh6?CXBv6in#CFE70?wY|kqqei;hFX4b?lCu&9Y%{H+vlpmP6RfdV;>aAt(MxV0`^|XT2D0 zW56BqJ3k6I+DEso_Vrkkqo9TA!z*pG@J?X+BZVK=NTYqom~VPV_KctYi;ct~ND=+u zYdy3az9pen+S+wO^Opvi@BliB1vjT1IXDtLgIh(SpC2q@K8>oT#aA_Vkj0koc0Lf) z&Ms^`B*HeNyk2Kwo|=n8a{b*T>;C|Uv50}d|JHA^wsVzZj=N8VXdGA$ccy+XC3rZl zdFgiusVU_Ihba$Y{HCsJ3>oA&yrw(D^KvAA{Es2(ALw=UX>o0qs|CotPN@W^#b&)m zr>L(zc!F8zfYrg`1=f;s6y;GS*_J&WQM%J{W;J#4NiJ;DfDQ>!-umgzWVImvG{8I{ z74IZ!lNpl@?1F38Dq)2X8OiR{l36|*W}B5qSiB5{AHy2kuwiV3#|Q-sf;Z<&sU0S2 z6J2QO57i2z=YK6gtODe!m_aFg->i}^kvnc`4-WB}y8siUv0H(Evre6D>_LMOTj-^R5N zp%WvWCC52Wm{uhd#~pW9N3guFkCuMF9nQ23y?R{CV%LLHH86~`KVo1(~8sHLP?vs#<&|*`kjuh=Qfj< z_L1(O7?D9&DdJ;--DmqR|7d^8X>Q(W=wN2uPUWad&H#{s*=QsjE|;08;eWTG=04LD z(XB`=vm#HjwqA?!)-`*eT*rl|l!R!SMQ841!=5ibEW6~`B5jX2n9CLTDAed`C$Wl{X_(WsK3M&*(trl9f`rM+OwdI;%(_ zHTx@nivn=ow92;LzDOU}&-pd1h9f!h3wITa@1X5>RLY}DwLNy?teBm&1hcN4x%sa2fM{ zO~xvYGzfS_DW^sZy3hM)Twyh+1EK)bs5oKvjr0gT^czl;zM+|m$6;rYz2HzIg>^F{xQ@Fdv^|HB31GW3~vuYKhESy7eV<}2{aX&HFKEiCO>9www&c6bmjS? z&#ob&x8jsemKn{WE~|fS1sml7VRk#gl?*d0AZruWiahF9M8!#A^O2BPwq4i;>^8t`8)dk^>^Cea3@Uf@7ITs;=|$&C;&b? z{vP4!wa&X`ac+{wz3xu`d`KKa5eY-h`o6TH^{pw6qb@rzx2Uh*Yc;R7n|B6rk|}V2 znHw)Jj@uP_S+OsoVdcr=yVc8&O#4uP)7aS?a_*r)!=ssebHeHE{i8j$m>~Zz&c#5I zY)u8nI)l9Z2LY4SQ;A-qcdG@U3Xoc&klT{%%;Xi^DmH;a&b8QcUW05KPHA2{pbq1) z9C)D-P!7~^)o`EPsc(csE$#$5B6oTyMoz#9Yu(FUx#HLG?rimofGzhfzGo|M7f+6M80?5-r9kxLC zVEvjuj039@dQr$4IcW<&GCh(b(9nz`y9N772VX1@$8`15Mdz_uE%gfztDz1iI5=%x znveO5|K@T%dY$pPg|EtJsZ&-){1-}RFWA-fqWK}{ioF?EPo2A~2kWq#3w4?O!CzML zi%W0I2Js6vjBZL{rI6w89-v8qdUXe6b!ybuWLAv4mG$rg)t2;PO=7YuxII;cOX3-#eLrtWFL0RF>E}V=XOG8`0{6Z> zM>NlrhSkdw!F}ZaRv|Sf|Hh)s1}P7rw4{4Q93rL3=C)xvOg`7%!_YVcV~AubJG`nl zo#*l|B2=&owqo!cW5lyf=cc52+L?CV3Pv2AiwQeW7;CEsj&ozVCMGMo00d_bHG;Z{ z3m!J20_V!EOmzJ|k9A5?F=>48zAB^%;nO7?{)m6^EfUgEb^yYHFL#?}msiL_IyC}= zD78Wtv{n0Qz*MEEvYYac$Ipj2*(u-Vx{I+H*n*?^D{@nftfV3ekNgm0{>3CspSgk2 z2q%$~gr&YZ@ne~_v4bnS+eE%*ILOv#B)3(Pajz|Sja$DLrP1)<8dMfo=X3O=o|itI zm2?YzmTvc5JOL7MFg(>N#&zSEJ`OLjr|3hO8)+qZ@sdhB_B7r~AW-SfM^6(K{HaP^ z@e|6kg5zMfk+mmAc}_MW@pu?z2C!$w4b1!pjvZQW&(fb3KNVG%oWveytS#Nn`$Pkm zNDfs9YK9ski?CH|p_O%TT>Djc7}-^ZL+NWQp$QShFaRm`ZpsISd{;UB@W+?viqc+l zxEaGJXkDY-FKz#Wt#b$u98kD+?Wt`}IkjzTYMWErliIe;scqY~?M`j{bMIYzn}3n# zNzNj>ERwwEymEZ9V~4RUXlzzx-g~&|3AY6Da@xTumuez?viDIC?>h|2cCaxQT&Q(tOHGR0XT&Uifzv>mT6!)C?h13}I2ZieS-b=V9mr z0l4pm7;Q^&XqAiAB8pfc-QI3@FOmBpls&@dfLI?GXq zYO=_7d7kZeX>{i8;>fg;Zj9qo_3^2T;qI-?jI&%y1xy;^bAM#U7faFV_5V;|BCG|) zeGEz@p4jFFr^HFYs+Z)Y9jSIP?+;)+*$t7=Y~`D(*FN=Bn>5CSea z0xryvJ_3%Cai#qokwuhuFk$36Qxm#kqFx=pd30VMSP2aaewM|6@&L5=)>es+NF&fHr++j zGVS0r&RPfEqO4V1C#IrerLeh5GE&b9!WlhJ0a}y}HoaB&U5c%Yar3sFZ?ElmE zx6~rr>d1Sn$LWFIBlof#Hqh?}yMtJlrzth~FBdqIQUTYd*A^A+_6Gf8Y)h0Iz+XPk z_9gEpNx@wqXJ8=F3;$0y@Y%abF(xtuz(+`}SQqdY-J5}KUg1$03VXlUHo0}2{ zhpl~)e|44(q|0;qBRdZdH7fCzd}$AKSqnp`<4Z3^OYOhP3H|4c)@&S^wv{uSOZlL8 zt|p!HCwQy=R_~t-))BdOtNETeKwvoK&P%x7s`A5!RtesME3BtDT0|e;;`)R~G^4~J z(Ss-1X0_o#u?X<)p_AvpQtVY$=fnAUBlMo}1W|GE9Ib;i$2!e0?8mzx+wipYw=vUh zjV%SCq(D=ov(Ec^P z*mt7t2yO@ohY?gAzI-h763ijQ=yS3R`3kSuI~+ga6cHF>kqbD8XxrpjsKF$uV6`2R z!I0$ve&%nz@7qR7h*4nk7y54t)!uL#Mea+yNJKNZHt5c0x~LqCrdmyX4=ZuElN`90 zxz3;3^2zzOUyDkt@c=UhyiO3Re4A48>VIT=`Y2A0WruEw>DgftXmS5eAL+Rg>EFX+ zaFySgwbPPS<3*nd7ZVyp#OD0y%0L*#()&Gj-mR+Kbv;}%UP4TY$-Ax;8>6){$m7I< zQGSf+XO3$Fvp1_tJ`&|tK>v0n(ueXSJ2xBGl$YV*!SzO*2((@tuwYhm)ORvv@S4U6H##U0&u~#0)P#MtGiE0Hl0HDUtk3s*z>kruFMP z7^Q5Y_48Z$Jt5G5OpfJGTEykXEx6=O+|&7&KU145NhSGe?olO`m0qT@VEY>H2q!uW zQd2+N>#yd`PgG-H9>g8vHr|?Sc6hoK4n?nm79%o92o~H~`+Y7Og#wbyDpQDu@2rxF z%5J;q*8)1GfeXbBg#jZ;rxc9&BaJ1P(?;u7q4)!C**&%2pB^f9GpgtctE5$I2{n$l zj^!*&O8(KOE43kR46Gu#g?w^d+Ap+w-v@4-wjTDH?jO}aS;D8&S8+^INv9ss z|8PdUcjh4x^l;7a3 zF;8|%0t8cBIFV@#;18=@{$rk~y?wK#; zXNDdB@CPL8`!DMIed980(+hNjV90YE$Pje&u4`v!O;zstB$szS1qed95e)YXAUM(-$WhVBL0%%)(EulK|0#uVZM)f7`*%3E4Re z1V!FrgtDnJF@E>t%>RD4q3f4dGy82Jsv|lcmW@$5yIQH{yRKy^Zn&1SoqmA2J#e># zyku@T09PTv6qH=816UEju1wr_R0v`Sg>{- z$65{iZE7+WYAp3?xVl9!TtOyUCY2&%7gTg4cYtGT%)2Sti*^h>Z(_%RVy@!Quee;A zFS)xo-lk?;({L;x^KFYnf1`6WN4o@xe0S{~=Bd0+9zl2(OcBoY8$0rl{y940AsvL(+SnurhlW9>6I8+v$NF(~)zBj2z@SK_rz7Lw z*6BRH`j+{X)9S{Y;5~iYxMkjX7nCfm8Xz`-{|hcTc&%S^Mbqyc1eSt=%as@e=D-9J zjEm9Il8wY(`+D0^R?DUa4FRn2_$wtS5DgmwNCc$(cVdF26}^1ld9*=zD1Ea+az4HZ zAj9l{fW3UO2XfzllM78DJNcs03KIT83O1aR79`lnBu-Yl9_9rEz)mKyK`y+#?NYz& zp+i^wvWU%yy?jw&Y$2PxCM=-pAT0zZi49uTfuLNh2LFD7=hYP*9UToJS4Oqcgd#?8 z7N7?btQo8{z|erwyoPJXZ5+fjmf*R1NrTA}G+hmOB?s*{LY#ffm|<`a)hS+lyovAOcet%K)P8qqntJpyKA zXM5BB?amF+TI&&c*v)`zwA-3cEcI)IuGb!fBlhVVB?AY@1q?`gx@PT%eD3KH2=FN~ z`l!+`vwKz!5Bb0-hgfK0{reDZuqLm2i4r4hqUH)ppfr~kL_E-LLa*IclzYH zUsfX6AYTNsIIohrjJ}^l#dsjCer7TH7hlufU;T$)F-P46@A_$<-565c^^G4Q3ST9? zUlK@l0V{1ErZx)e2B2U(k6zYceDhh32CfoAVQR(KvP z+SqdiE{N~n93V5wg?G%en{C<3e4Sl3e?#AIL-T~f(<9^4&^=IJSl$!6 zw@3#dU@JObkVNNBZ{VNHi4$8~Tzt_tNS*@G*LLs2L6A`G{*#R6Kyj`a80eBJ=wu3p zQ6W8@KnUbd#CUoaE{)*3$eCR}CF3DlCfrlYdbpe-^S0HZsEYN(K>`LA%n}9OrE+Sk zjg~1*sv?db)N%#?TUtWlwf#eO|!$U zQlHQnn8IiT+|M%?>wq*D{CI!mvY+ihV}b39fFJPf09QiNGA5;B*B#nX`Uf*SV_9I_j=9O5nWmV3(Sz+vv>H{Iv^5*!^VJ@xr_yEeYzKmOcFlGIF%8e(K8pKpK1lKKb`p+~Z&@jA4=ke<3S%Vy8o=OxQn1QP5 zeb9O_;49*Gmg%A#6&d(JSSiaTDt8Xy{Y>i@>n0$PxVNM#>QLFCkv%yJ3_OyloIh=c zhauQuYV+GnAQ)p;lV`1Xt?7^-L`;o6ovN=*QI}9qk4mK@BNqKGXF3cx1z&i6$>U6O z-I{X&s<(-?HfyJ7$M}o}9@_2~rc7sUFjW+Ai%}<$ky9kR45l15PEf1fi=1`To@M9G zFOYrc`Pz2bX*sL21-2{~v$b+ONDLMHVg1eznmhJq~tMdl0B*#rSfm%U7B4 zUm%6?5e8hz*pOIN|zEL<4M7sx~%tV8HqItiQvIkN0*at&Ty zjk;Z>GzOklFFd2nmSTRNbhXKhJ;MncL5Hi@?}L6{XA+m3|IVd3EJNCy(FuhQRQpyS zT}Nqd;(ryZ+j&P5?1GMCF6&)F;1xI^Udf%DFdO zDv&kKWq_sd1yYi()JcyCNUaE^{00lLloYK=&yzSBMkWPohb+vB5Sr%7)wi z-MJVT4R^>C&u8__?S7ej%p!7mjga(R(<&7~Wk>liVXAT{Z+P^Jxih;#|GILb6#DHj zAgu}C0TLCdvo>L_d>q+5w>2i_tIzx}ix^Ra1;*kq^ZY>Q#P>Zz`4-C@Xl`|+B^yH2 z3`&z$2$(0b6Vb_DjHkLhaAam#at+~j@~ZtkbxiC2GwvlEdO;^32K;w>(jG?8z+G6s zOfON3U4^a?>pKO$jcsUUY{{kEd>sHk? zbtWH=#!t3on+BCzTOrySx!AYfT2!pD363lCv8al+sU2NHXIm%jxjN!l3aFGUWrOru zcmrcYAyhLQG*2d8fan@ubaC#Y#5r^Z7-{@v17uA3z9S}AFrHabgnCg<1)Vk<+2uo~ zWrv*g%jLQ>N9p+!_j%oi=<)_lR5Cd^c(WGN8y2Y?&RWLrTbBWJ^-|fyk3#m`9y%Y= zrx%ilLTMuZ0ZVwUAxIy1BN9;xFw!XpsI{X41ylFh=$!F0K;`7;8iK}sX_ihBQ(OJCXy8&50b*8_yWhT;Ltbr)bvDw-|1Wu~@0cL|21@=U&JU3{bt)I0>P<-3~~fkLPnFG9%}g7Pxg!*e9K z9B;bp)kZ}rLr7{`6}T7Hhh)|uy;!LYSruSdQL6%&fZoGPBw!z z290f)bDFQSl^k?taNL(`S^cHtlDo>RzaqHEdT-38u{`&d=QIC!xcC9@&BA3_p5GAO z&*yVX1#lA+?SdEQLLp^#E;V`e9z&xE1U@M-3**~gH(4^}zDV_-k^^(%f+d9xPQI-U6;Ey8Z6is{~%eer?n!dhIqXjt~bd*cCK z8z>7Jtx$3MG!Y)NDU%Hb0@>h0*x_1R6cNUcmhhlv^_!h;01L)*F^(VUNei+ea+$Xf zNfJ4p5yZ{oq3HTE{&q=JA^4joZ3$!}-f~OISMa0(ezD%RxJT#;EVphIq_lm3> ztrt6=3s&RH2FSNm46v{D_F#xH;}su_DfFa(A|S!EcwJgDD}@=ja#@}~j%1lrmP_Po z4nnWX>Z;Tq>ZjB1ub(%7SxWQ~t~;Zwb~M zSUfoTV63@gsy7T#W47s#_+o6R1p+VTNg6+T?};Nd^CxZ!Bj``$pH`!FY=8X~MXghg z=+)6J*7~nE18XbKP3K+Q>5!LiaA1oZWc3K!yP>$lJIz|*iqhv4T`t1KO z(oA%}X9<=*ADDk`!FCTkzWXWORaB5bbCc8rPqw`oI3gbrxzSqG=M%PR30!?ZO+I4! z1+e4hA;$eM3JFyUl4q>wNj5-3f3d55VmF+O5@A+SLvzj`o&Ep^ZJkaYmz3L^Ulx7o zD~->E$`7w4o)i^v;asAx^3r~>LLG{R|KUfBS@ttN$FFvVA-%X9-e%+c@WH0}2fjmP z-(WjjRdlBA%ea2jZqPwd8KC`hn@_@}*fPMcRYVw1X*T~*&Y=PtUwmAC@#Q`3{-Cap zVrNowF)!#0*6%ig$Y*{PRL_;$QkzbqQJU7M|B4YUwM<<_{CPqbt3oJoh7)sXA~-d* z_B+7_rf)ea*`j8qMLSWGg(!g;PKc+jj@*tqGfhokk5 z9I|^<&C~fHE%Cm811vwfWwDy!^79mKsN@(kr~BUImDlI{sld8yoyTMEFLZz@Y?~7; z^JezMjLB`R9UDpxsp++=ZIfsBAm{Il zt9@Tdb$&=nBzlt835_{IxGc3hrjVmC=2N&|b?Ht*yTVBl0H7#ZL=Bvfhs|Y6LK;Ul z-h__`sgfF)BmH8Vsm8kWQiyip%3u~hFP>RFOY=H7tI+Mnc2i8VZ_%(g`irOo8a%K= z3N}AGC85q^yo1`#oj_yiStiBl1v+vcC#FV(4>JQ`J4_eUGO`YB?@X5wxoD~L_`mvg zy!)|5D1FUjfp#p{`L4|ENj;ijFb5nq*H7q!gSShbyqVhkuGIZFEFsZK6W7S(3A7=SMm3R; zJM3r?bY-XOItL9wK`vBtV(QSN0}2dSkAEbyiNnfFfVmXN-YzYbxw+|WP6U~$(XPE* z5h)`Fm_a4Z>na!~U7o{2R8LBm{3eW)Q2A~$YAQt=-B~_Gpp%vC=`I{xTryj^*_lku z7>69~O-{3bVWA?$c(2KFEzFnA!tipka6p0#_NF4KW)e;~53;26XF{NHAFKATk|$z+ zP?L!%z-ZalXNF&s?yX*EKjFu>y9YB)MS6-Ia?RBh_NRFWyePJ>FCcVbF&ND97mY64 zyp&{B0_ZnrsHjh8&W!U!aC6$esWa`$3seapQA=DjfsV4L1fHp4PUFYcRu`cKdI9(a z^(sb_%4ch(TwYmeC+_@ctgokYi;uW0Ovt(Y#nwyrg>#} z>1+nMX;?Fhx$j4)S>qN#l<4h?!m>XuqW9ekL0p*UTjd5i=C7+FNQiBIlix@k4_CLm z0DYByUuqBF`e*cL!Bt&Q^EdI5P21*1indd|t+ue$`ENK`P>=q9rmH{FR8G{`3qSDf zLM49^m959<4s@^S)#mAq@J695jA7o)g0ipFzVLXAT=xyF&@WSo%c z9AA(5a>Q1yYipsn>qP1%!-UM^7fUcx2INn@2Y(=t1f_&V$c1gx=XXVMD~|+*6kVTwfFE2N)P90yH-G?D?9jyk22JO!$G1RQB3Q%Du z<9B}zBt(Z8lb4#19>KM$9dbP622L7%Q88|vpm>H3wi`>+5h?G!;gBD(-7~L)SPf3X zq-c&Ta&{DGz#qz(yeYBAz`a6S-K?_d#RI+c(>=GpwO*jR3T@oTQEJf<)sTqr- znZAMZk-N7s?u@6=?I*hm2@)$P0t^*FL8|<^^Sc=4-KkWY!5J3ggtKQYM0hm6h>g6! z5-)jC`KA32!!ZEIAy(0W=i)0t*MVmxx$Aq1b{5&el3D#bsfWx-a<}|e`U`HMN8_UI z)~M+IjGSoOGh=@rblNHn;SxH_sos%hX0kv%3_HB1&B=I5`EdyiY~oM}z;#Vo=AI&) z8rF(C*)S)2gXbH3t?FLWY^{uyrxXq$o0Ay>#;=+9S!%Wxv4gIpk%)fAPAw~{P=h{X zCy%Pw2#;AXV%jMA{*k0TP0Hp)hgagzF&ChW)j-y$6esiIgL||!^IcqY^4q*8*R_}U zmh{FOVg2ii*!v#jT3K-%7Jl1!e> z`oE>dNN16*bh~hm&TSwxUmjQc%BDW*??I8bo`mM&{G0&=psYU=CzdQ?M{r~Cnh+;B zf<`DmX3dTXPksWrwnU_|;2i+_i=&k|@xC@@P}fA6SxHs-j(W%>BPxBC%u33ThmT;d(o?nQ-#o|?_q9bFE=E;LGS$=MW(?*5mTx- z!%8&X3#zBu&8Y^uyzr9=mK2ZYz)01h$RWz%BM7*^QEK+%EH{~nR~HJNaVOG{JLG}n zb77dd5x~a{ynf09&RAIp$S7%4IGQ)*ElJzoUm={{b2aCF9S-law)<3I%M4fFu{X4H zapBgOy>L~l&FXOS9)RbAY0)&Dd^;}e?FwcVuyduKwURlg^rvwC5hoV>0PCW>Hk+Yo z;ohk*S8Pv(^r+>Q!y$qiERA4xQ^K!{b!BH=4W-U&6CZH`^bW9EGhiJ?Xq(pHCL2?~ zPH{7j_)42A{-O=}Ir(GlYClIoW1;>e=nf~OnrSH9^^L0~pfa5UDV_IFqjlRQwY2O= zPxgf=72ii%MftpZJ-=J5eJA|zH;%rI%1Jl&jznpL3gWl#635AgtSB$V^pz@sog0Tv z0$e7oj}Gg3z^0Ka>@49aVO~;u3c+#*c{P)#LY`NMpgzicWiLvBct>Y+vCKW2!!^h> zX42FY^O5~RYWi13iB{aY82$pK{h~_8AY4YF#%7e%?2GOLs=*em_H_U>;Yc`wd6ZXp5y-b-k^V zjQmtMZTY3urNyw!ST&-m1$)tcWmB9%88y*K$-2oIcVSUALs6B}|G3Wc^)&7SlMIW?6suxMBvr;k4VOQHa*zr>j_IC`1?Fk#F z0cfn!vSZ^tAv%Y8DmAc-E5z<6YWLB11T1|e;~Z3pnaewrP%;w~s&6m#iJ$evi~Q0n zCfCiyWtFpQHzj|hEkLV&ca*yP=!U}vY12yr+b%kns$N-<81Dgf8p~^nU}Qr_2a}rU zIJTJ#OC&I7n|gAbidINB?Do= zKH^{YLh&8Wh1-bjPMB_CP-OXbE%H4GNJI6E9ZRN)mxg}~@w;DVG*&`WT*Av={#;tc z32=t#Tgl3=%h5W5#Mq^JGRl~rH|aijZie~bcxQ44X-jym0toVn zo7vm-G_BaR!V3+I5Wk%E13SzWSQ^|+OfJ>U%s35} zIAuHeumd(|6xK@*w|6>vnc^cG<*?9Sfi93_atDDr)K z_@=4g_qXOgH*$x({i98n+iz4827c~xIP)T&l8xL0LaX7zNf@+)M)6|KfW!_f4nH+p z^l9PWcy&I7cq;D)YFn9~H~Y+|Ucd*UKFhIQ))Y=%PER0<$b!-7c!9@qZ#a=`s~-ad z>-|>og2$4v^#ri?-MFIlj~AP&-D6=((*Vkm2kZSqKi8VHnS`pu%84XONhQ0UC2vB{ z;BzQ3H+$qWCht}k(NEJYz!Kz{KZ{d=7^C?Pv;HLcah=01;ck=|XU`B9H&P)Hb!qUk zj(`6pf0f1S8}c#VnEI9O!+!Y>p_kok#nn^df%=jX~98KoY)rWK|kgI zs526$`@9;}YSFhehp0zEa}&0`3E~Y}fTsK{As7n=Ka^75 zOBby2z9_^at1_JHDlY3B2J9-mPxrrRr?Lh{OLp~Mx^B9UWcHirg}%mTyksLn$S6Gy zE?ydLfj9nlSmUoLDGVC$wV*)W+6Uh@!5Rn6{q$&NX48?4=lAlb;}&u#Dl+9xu4PuO zwd7Hy0e#c%49A}y05BLjHG)X}HsOh`EsY|DeLJAL{q(WZ7H|b#>z1lDw&h$A!YjVHD zw&7^e^H{Doi(X1psh8DJ&Ru?Py}C0-sC_}WH;Z}hMjz=3&~s9XLTY(YS;ti0^^9bF zMCtfwVdNZDM=xd!*jhvj{hcw~iI!l0&Ly=4j30e8aINo=OP+bbq{foO^}c!1}ypG{NXAAcl{jsnvlg`UO0`!VMldsaT*^@v*s4z|W=!DUE=d z!K~fnp0P6%u-5GnJCA50TZz9~tXQ{XYL?fkiH3G?*%@OC)2bs{w>(2kFQnf5 zvNdiy?}7wzuSXJi7sH;kpb9)z@+Uk+EgkXTi^ddhs#4`g5o+g_n|s`5284FmQ|btQ z9<#b(1;T1BO(MK9OuEG=9|^WjU4IR4^wplEOCUBCu<{7He>2e`OUBRxorvVC3;f`r z)1C$Z+vI`c*^gDvC#7!}D+%bq3x|977+ct#B~-OPOSEC}a`Bt!3PBNlW*#!f20eO6 z_tq)40B*{HV0$hb3gzMeTTD74tqI;^!(b`xnd!VmEbL8+#0!7)LrY&A*L)uT-}CxL zv~}(ySFTU4-~9w2V&iXL&rQ`P>oSuekbKgCgG%c*UOeT;VYsh)Id<;ktd-*0K6lmq zn(?1zkj0cp4JYnv0*>vtSW%`;$$w4bViIdRBZwWmW(jcb#lFNWY) z^`ykUW45$(Hn~P#?nIhXzEqX{Aaq(!SEf5%2x<7m&>#|{6YJFAnfj-b9wv6qBhB|O zAexz5i5lwX%3y91W2l5nDa7AUSTRmQ>_I?KuYluHKK$qIq6)nc=5*sy=<-)xs2*6hd=Z zL-OqQ>k6xNw6@>{T*7Gh+D8$08s-&5z{_ov_U}#XmGN#6upG|Z18TSz(c^S^HJ7Uw zskLD^BoQxkDk9<=x2JC){T8U_gC#RifS2Cp zhvxy0No`xG`rUa5k(B56@^ghKPvgD0;@22`=E_s6E{0gK-P$eNN0njUB;flY|#NCnk9x`MewnXT5B`iyiH zdPPlMH;*r3oHFJRns|ACWFU4(%qnza;jVPT^uLp$F~00CVoiYLnoss=IrCj%4Fy69 z{PfVXCX>+-zK-~YI3Jhwa~7CuZIFwHMj+isj2NiB1)Xl^e3}mKA&+j9t6@f|1dmTS za=sO#MNC>1w|!I=R$tafz}`@HFOt48*(@L_;`=46KDGhgs-sN6@4hO&;$F z^)u;q*4({_u1bG)?WHT$G;w{_#p2mkjiYu-+bab4Tzu$YNxf|>C<6wF^ITkh)!yGj zb%Ja~xE{8z!od#?{f;oGA(Cb~n`&k)K|tTMU1N#Ea@cmzgdeDXYeKaf-*H>k?dmCq0gvBq7UhfN9^!Ok#F9HH z)3)svv8gvsG$Si}=mEGU-)hCb`tTfw!~XE*7^taipa`CpI$|j3JR*XU+RJ6c&w-cq zkAv;Et=W)?0kmj7~Gkj!jsEKCU`h?GENE4Y0Yd<#83tEIKm zphd+{lMN+pQxsIx>QK`nb`)Mo?c$1SKzrN2G4;*kFY~Lc3D$FumyXx2Bx=a8e7Mfo zvQ8$&jjgGm?08993^=;+|L;NWa-ZsJ5*XW?9AO}tp1fLtZh1DGJDxhylI-z?ysRT%}hMlhe< zOcCtSK~&*oj=K+O9?j^l#N`D+Bn^x$!5W;PJX+S${>^IU9)Ms_Yx??w&s^CVuXcYs z0r|Rz>Y3ZY`^G5>3O{DH(+-@wB5rW#mQ4qlL+Ra8K$D1A?a77KIL zJb`}#W?*h)4WPMxaIJ9zN%ac{dDTh>73$vU z$cozV1;L_J&#ivS3I8JHG&A!pD++_}!M3oqgnr`TmPBBE!wU%hYjO%YV+eLh$l`mt zJl&mhFL-YH%g%iNW&OSSg*p6>)abuo>u>FQfJLC;N90K6oB9-JM zF>X(Hj2fcFuhU;|-Hl50;<+Nobu2*hOLJ86)Lt^7pt6_@7FtAi^PyeqS~YEbZ38&X z8jtp;X9dWKP{E5n_l#wlMgSqfRBqC{3?zi^6rdcwnzoYh$<2i6is&1ckPt_SZ!tnV zfOu^NwV6*nvtLaC#mKy}+U~0Zl6&h7YW65^-~D~87tsLymEsl80R%e;C;*;7#N)O@ z(Hac9QLub}Ug+Ov!}d^p;xU0^Q~V~w1SvSM8-O!Ne2rQGiJa?335}SrgV-Z|!}Dty z`4A_jzx_|t*By#BAa=t6=J4}1N>C>eAVbDV+J+erd-zZE_AgraPZaf^=hn z{7U-rvHzkG)Rg@cyT~>8{>Ywd`o=-I&Jb34Nz@^zJ!0{SSLdsdp#_%ZJC!T(Xh((| zw>=?LJ`D4&pl61OOg7Vm%ccFkm8Q=hU6qUXMF6D&)UDdVFf+b^*M0aW0|OAd+p$>> zSk7;g{x2yzZuXD*M~_wVfC)HG>KCZc4Nz|kF1OJ)-wS5(6ap#vDDnHdVwxHtsAyjC z<(qr-sfk;9*90U&ot^z1xgb{OneqF|gI=`#%0L)d{S^46$b{Ivm}`KT`_!}vB>P6a z5S;>mx-UlJIOqw4r@)ut*iFjB+~-s7s^FwUjl+lWQ1|QUvsZ#30>190u5a-W`7f~UmiIFU+m z{bbJD&;-JtTQ{wl@sQcUKG^5sHUG7fJzL-_1e_}PoCulqXR`otXxbamI?V0kGq``B z*C(jZP$J@)hjR0YbmsrNRzLdwDJLGn#v(avDEUIT z)I}SpxD)!uiY39mk7~2N^2_G4O*6Go2pJt^eLYZ?P%TUe(8`_aW%_qYQjU%xieynpNYTJ_SH zrJ_h@Wza7XWVd6=6b&dtTc-lBG{(H!r2|IDKOV@F{g4``QS?lG;JeuQsJY|4ZB#ib$e**(O1NH7*v1S>zxvQ=S9C0 zdGN8feO5p8y`VP_Qi08r49#z^T-DPsu0vBbN(X~6&nhpL_Q4i=Qrn=l&m;)pW2uT{hP!n9OQml^xeNbUX((2G=n)&6 zTiz>$|Jr?;UC+aQ?<@zIOF9{ zrpx&fqpiEh*o$*}L(k~47-Zi9q9LVg4ky34f`{Y*&FZ(r{SvVqTWvMv37;|2w4A^M zQ_%jZ#ZhUI&Mlhz5|S`x_}O<5wcG-KJ)#;Lu;{v zFIw`~o`*spvLYgYoLpNEa#dc>;#(mZoqND6K2_GVel;f_y2om&Ic)W_PS>`K{|^$f zTl>N$);I3a7G|QFPg!L&(11GG{y!QA=(Qutj98a|$qZwHU9UI zs6>pEzlgWW#b=4h2>sy*Q@hA=90W}RHIe@WbD>Q&qH56iD0#=sY)XL$sum0pVDk;0 zju;_{q+G3&(*eWbX0M_b-?Hd43vhN;6c0W?2ob6#PIxKcgI*~2e^tl0%@wqiIafXY zcEq6b;ncGNs$k2wh|D=H=GQDDkl(zJ$0_w&xrb$zAYX4y`3=)%b|SsI-}1a{voDpT zs1@5V?mGXzmJl?Y#d%5W+46WqmvvN0I(ha!$M;=sxb+Hu!$jKik(WT#oG<(H!~2b_ z+s4dgs8q<1zjy6?7!9>J(6&;?s~Y+?AtbEb4zJ~-4Hz!W~J$?i)`;O z)hM|*F(uU$emQqbW9L=fBGbKIFsX!3S9IN*fh}Z=GoUu;&DsTYAm_aYGQAz3%u1RO5play7mZxQWwi4lw$|O-QsxrWx#E z4dmcz@`a5kHOs~x9vo4Uzc%KmyAa#Z>S=d&Nre4#v*)iS@+BrbO5Ul1Zqt>*5C4Ke zd%h~0AL!KliKEjrP&Cd5uS3hPEZ+ZXh~~lqEr!sS?p$%gWaV$H4-3wY!&foPu%NjP zpv;GW;k9$7MO}jXQ}yOy=m*s?C?cn9%FcCNAlAWwUEK3Z=1I5<3KqWU^VAR?(h*{* zGY6JAu|@D?Z#ZRZmBj)3X;3+Pu%%)GU+|4wduCc_leHHyZY8(c)uwN)h#*zW)2~%3 zYl9+G+NG{G?rCA$H;olmR8p-ZumIsDAecgzhtm87|L2T@okgezdk+1-f<@W$P(kE# z=dY@p>BQ&iGe0Kw1@tF1YYy`RA(OY>Sx>3=A;)d$PFimI!(P6Po~mi(|>atK^~?;wzv8E0NEk= zmQ$R}s8*Naceo#^`(HKnyVnq3RNwGut7hy+X~El_3OUtsk;%vdSuQZsPrBJn6C_%f z?W-GE^c5O85sPbCYtckklZMf$z}bf4Ii_g<6*q)+>9TmdR5Mv=?ZmnjxZH07c z;e%$~sk^Jr09(IyX!`V}Vu;L7ns$xFm9gKX-*JVSdZlIJMf*2n?gBu5k&-}Ue7_%6 zHtWZbR}ZFLb}P}2D2!2m=RJ1*)Dw9~K(~LiSv=<=j17>bo+45_hH{u?1^x+2x^kv2 zD67FI^sV;`@OT&a>~ND6DOsc*MmpZ@<0A+*=-W|PU`6(o?XX^DurmW&`$m~A(;G!; z#8QdgbjD;MN3Wzgd%ss43QUQPE%+YG-W|9ROp&iagbOH%^fQQ#>WiWK-C4g=vHFyq zym)|RD~8=R$gdOjzEx5^0e5Nwo{XM2g-+@ySat!TborsFhM})Q*@NQ4ibRmbN_#Et zd7VZQ+ehTTVg)=jb)ZklX|3L}NSAq+Yh4>ymL=?I`YWYL2aHQ7qM&VG6oQ5Wrsg@o zbJ9a0^Cta~neSoQ)I_K>S2kv)Z94v$Da|!lr)JJK)J3`oXzg*jfup~%H0Y)UNHLc8 zbKwQ2mCTqpp2=d=i+#c0eCAYNHv$;swVRB}{yhPm)eJ~2Y-}SbEC@6(s(FeZq}w!i z{7tw=S&nBLwb9Ckuk*P}assbm$z(IlovwZ$d5pKVLgxj0`)nI6oGs}ZTQ`d`3oi|_ zipz9$Cg`fV-~ug8K>QA8-NIflyx8ZXm6CXhJnZBf|5&vG7v{ce8cbsgw3Z6XxHw#=ov>Hc@W<@XYZiY2x0!YrXo$nB_M7>Sy_(=g`6QgUvjw(rNy{H`%-t7B-onA$ zzuysu)e`hn(et|q$%aU+e6X<>o;#>Wge6YObR%JU*l&su|MUT2b^Y_Qk8G-TNPsx}<;85(X<20_4J+4h{0W(zsAdETnvQX%NK+;DHZypgW85@Uy-@jj|eyf zp~5l-&su*jog}w39Q@ga`eM3)9EPYBIHa`0XzL))F7qB-;~5y?L({vwux1n?nN+!< z-7;Ox<{GW$+u7f@*Lz(8jiCmZBpFH4`h0u{_CPWNE(J5AwNo;kjVnX3Z&D~$-F!^+ zGo4R(aa2kVKR^YV`3^vKHQkD8fi$Y46e?>`rbU0h^TAKW>BZBC9U!NuB~dy%!9&FL zCYnI+P3a`I78q7!h#-4F!s_}YH6Gl3Mm#@!W^Yi_kf@~9$2^oRXG|x|4<0K?aJvjW z{7Vya)z!J*t&^*kZ2EX8ld8#h_mUqiG$e89{+h1J*q ziR9O_NCnAPn4z&4VuOdqNS^(Vb>o-*@tc3r&2G6!b%A;-WN_q*Z_ALEsYEOU0@+h} zgH>>z3ENIaYPv536Dx5tVbhE%>I3L8Qic3YB0>*r>@~pa<)|-Xp`E~*ipiD@=O^s6 zysefES$%9E+z6`-TdAaMa-RJ+YDsqLpK@=e6?=w{y+7qai&dwb@r^qe?d!t_E?9pY zDOf`qW;UIsZ#mpQB2;p^dajT=T_&wweRhR2^EGkoeaGJI=|}RssyCbYh=1R8jHE_a z3I5e3hi2E3TkEc$8I;5$Og zt$9homAF8&??KPj&B_c%={L`AQ~?9F3}{qiDqB6GZKI1eo=?O$?-Q}|5VUUQM7~t^ z+a5INH&#oyvKvUGzrAl*WG?!N?7|o|1OKuSJ-T@gx4{-(%Uo#}X7E##-`9Tu`rWQs znK_pQIpJeTKe(RYncRcng8o#Cgv>1udpUH0`@m7#n{6p4$1|Z4Rx_aw>sqLmt`z}2 ztJ}So9g}CrQ)pUFCA7(~5rA@m!(JJ$|;=_r0J$gSqpv!Nz|P*o8j; zZs>Y12!ZGGrs9|wb8aPUzdr4o{t^fNDPxxBhXUF6vL^nqheY4 zJ??njP{&3@_z1aPyOx)-M#0!GPLq`rQ+bvOxQ~j1;w# zD?fz0!i0p0_x#5+-wLi$CGFy}AX-Z*h2E1Mc#DQA^D?ryJj;Le4_rSu?-T|6wC0l1 zyiPF7FU^>Q+I$PRc(3a5nZj$74cgQ*1>QwMlwNS&!mT{_NlV91!i%<6i(!st{ z-5%fdKKP4co2q|H$j;Hd+rx6ktdMvq*wANeE_JN$eKT|mZV}awT=DRrO;q6(EL!R! zh`@RWL`zH0-j`btX)vz$!mTM2=f1S@8diDDO+&EDem7euUZf(^>_MI#iaf%qk z{-!!8G`J6MhVu-OlArKN(U%7*f^@oF=60HFJYF@Y2}yta?ooFS@A*!DCJ~Yl4-%dy z6}HRgP#GYq+xG!c(shR8RjL1ZB84hbiFMGJ$AhdnhM4t=SY?#E%*q(`WSOGbMLPa1 zBmZ!&!s}@Up_BSrU=OZ`UQG*hJ3r0lS$21WUVu&J10=isR}`JS+zD|W6gqjz5-N1d z(g@vJd8L1DsY}V3vFi!qE&)~U6Tt=XGweP2SNOLon%%2RwY(KHZ5A6EbI^euY-TOu zqD>aD%M1%&R3T{eRiGy6i9+n$eV^&z!?fj_#V{vCz^|+C;I-hRm7#zQUOh9n=dA5h zx^RsPE+;Q9T0htgW3k2k4AMQiqk|yTvV0<`G!uUs20jK$2qLS^p>RmS>k_91@2%YO zKXDJK8~lVA;-4+zqLpIu#hPx=CG}pwggIkXTk`sLG32@}o?!+#x>yi$qqtr&Od}Q- z#)XZEVMe>NJtBRJU;>;Bq{RE! zEz*BVAEk+p5W8iMI{@uVFK3^S55D<*va|In=n;k70cK^hw(Z3|Lab-8bLT4XEPZ^q z@pC#~Oq&)GpQJN%SH55oRT5lcnH(?S1Ux+;`#_Jle6?bp{FW}X=kyYRS=BXJ^@2Mp z?bbslL+k?%ds!dy{j>U7TwDfR%2T>haJ_%d${Md)wF4gFVfnKJrqa_DKxkFdn@6%vLL?v#8 z=wrx}I92v)c@^>&hEifXCG!ZGN{QebGKK~E!qHm+z&X5hM{F*UN1r1yWZ3X%#{z$4 z=yB&)H40UB6eGu~gvunSg8H!fRvxjxOiYc6ov|!TNGiVX#;zth(c0D+qOTJlLKBmV zpVvzp6b|3aCPxF)yI`Tp zqY_V$WN$usUq~OX3VTlmzG_9JiHE{vWppm2jS*^BQkcwm znhB2BPvQYyj3=gyXfXASeDaidJ7(AQwyFL7gGImSf+x60lO(H>-R2J5Xuy9q-6)cU z4weWdB5Y%SHkf6$&&$gd@SxshJt&-vjXsjBD(nYQb=YRvoN%Xx8Ya{;}fjZ%~%r$#3sw9tTCR8tCs_7~dGj*f;R zmgJgK^r~-rB3eMMHnc9DCFW2OY8V!{0sMH^t0tKVu+n2$Rq?>L$zY5Ca*|F1g zX68_?c+CK0Mq9HH?`|KWOUJfXg@ql?oum1WTJ*Wd5W*2ZtA5Yy98GG#)OAgnn@7c% zH2T5Php(c>1JMNTJK!FiG+B&-m=>=`8++)*!d7Z>cRL`TE$fsLjT#C8c_m{p;C=?*56(sh>I@;&40V3Vv+<3&v(0 zt9tq?=A0Hppl`Nkd!7j>>j?GyzFB=uz>cVnzREqyxfQ$f#|B^r zn|JHp$wNp!p4Q}a4(fm5o1?ph;>Zrmd=A>VTyv35^;auSW}{i;9*N2!otk)@btP0a zK#=C+{H6vKzTcPk0`FI36xske0QaD=YL2iSS5s%3B6k(6#dm?ipW25dzq=aG&SkU} zTw)n+lxhbP7DUF5;U)q<5xXly)k0o|Y9c~DQWjO%_{FgvCf|P%Voc98N1GMlNz@jG zf3n^_m^LJ#>_b$)O3q_Z$CHWOVSUh8k;#0Zh$%7Ny(Clqsg}x;J;(gI*szrS(>Ex> z+G{?m1V&(#YoNZWqj2&&a{E^oi=?5L8n@0QT&9N$v3cJ+r7SqJ9I#m_cUziA9pp!O z(uSN71xdb3=dXWxdW;&={OGHg9?Dl}okUeum|%^P>_z;Wy=;URXsg8@wL6g)%oOCT zN4f-x?2=z^AGheI#q7ly4^29xkccOxA3o2o&!W#I^ZB+O?jP4>2wyw(r*n+#=AwiO z^Fe}q*3>Jpmts8Y9%y51F1h6+5(u7GWuUP3#Oa+)zqp=+%2=h5eDQ6mW{2c%>zno}J8o zaczIb0wz0s9}zq8(FpVCk-hn&oGGRj4CXsmdrm7o(e` zaBHlec1r%1d{8s^`!d+WLhf)u5wa2`rV;91;cCd&=Oq`_Uvlq8d3M~y*tO#8m58_& znuzg3Ns9|7vr!mkPN2B)76)r$-w%?zb83I+w-u_qAM>W+L^>uAGT2HP7=Kn)K&O=n zx#rXiFG}(hE){uf*r~G)F&|BWUS4HDAxP`}7%woK0PAU-VHaUIIHCphFcz;OOSYj&{HO4KlWL)ni$NCz5n=nkuBYy0apQiNLt zX+O9()H`U;Fo90`YiP!}$(l`d{kAkb6E%96z&B{VW4(RwAzSxv_|Afuj$Z;M!x8k_K_q&>dUzOfIAJ{>C z&vRY8aS>?D07oS3JQT9ON}0xc%0}s7E*?17zG5#aMTw>?t8I`jn^JjYjXl=-(sb&F zwLyto|3)3+j)R-+2L^maT4ze)Gw|*%Vs_Z%o#*D(zALmQe0+Fh%!|P7tM-2<-hIln zW*qWZgspk6Gu9v2JaR8$X`Q`2stE8PdVm-_x*+WjgT25a9ClI03gi}Rf zIMQYi8&{^fEu$%QJ$*_7Ms?|!-nlI+?H{c>t1dBaeuBwY+VxQ&28%`0IR~6JmZD>P zUo;cyz1ZKhjXG{tolnOKrB8p%=4R&5b-U&{Od3ki*2(eK*Ku~Cp=|XAqni?i(lY=v zzXUt9QLMwlWAJ&PBEd&!} z$=UoMs0W!W7W_8nn@+>Qc~0dIn>_YDW~2$?nO;WE9Qs0LJ@ZqSq7c#7>a;uZ@v z=N3jO<>%T=9fk6D=YqEfMIWD&RpmlRQK|q7wl~27CXMju{M9L`fe`b9No9ZuCSKQ z&`wV4NmDqJdF2ug>T(^>v+oYKZ3FNty%QqAD8aPn8>+|y?#jEK5k|Equhib@hK2HM-CA6hXbhnNj+A`PweNwG6k85)$KWSr{saa^B(XE@L9 zvZP3-rz^~ag=rzi#%CJo+1KO(c#O)G3e;=pEpA%x(v6}PU)2(31)|YgXzZ?)=sMhb z$Zn*Sce*39TkOPuihC*I{!N4x`p<8(DGM?;oq1B!6;FSLlbMN|8A4H~=3wrEQ)dGY z<%3njnvw$60^o9ol8+~~$q0DXX$Acfu-Yv*gfg6}Hfp{=2MdrP>XE>rnbi9^ot%?E zM1K%WO)%H^&|{B;ax-u%<7)neTA=&zE|m#ib?WJU%9iX z0?gyN%a(ue_nw44_Z@T8>2;5q*+9kCF(*Ov@tF^u;rBeem0(&)lp_8~Iayx2>ctDV zKR8EUg(z*fbLuXUVLO=MWhr|SsgLh-S_j~Rc?+D;fDi|h@RGZRaCuI>W9*QK0;PQE zIgZ4T5iV)?%xeSti?f-+6FQtK&n;rJICEM6c2*DCX3cs@vI5=e` z5J(Pcg^~6#RwK#DbUgx@#uRb4Sv&nAg9h=ygg+Vq)# zF5F(A`zhho8*CiB9#E!t7xJRG?0Z9^_7cUB^Wk`K z2P64vfIaV$!QGM+S!Re^`i*oR_Nj)o-<)RoHTViS>?6~iM5*WpVPprE(0!jaS+vto8*3(O|TPlk1= z-A83Dja?muxlw*|vB2j2a9f@WQ9=%=l{}0$Eoexs=G%xCu(AT7>v0k6QHCEi#@By% z8Nhw|29zlEO*gl(zctDsqgZFIj>N&tac7Bflhh^3kQMe1`o0a>8Br-!g2}>sNY$#A zQAnJR?d^Mzm#rj2sSazhj+XW8nbvGB`*Z>&p3tywSIHcRQ6d`0R+yu1W8(chmGYE> z*sr5TnftOESE%M+f52T|GG3CnzEOXQ{v?L-V3-N*u%m;U_lMBT=!Tq(#Oq?;G zUO7+ofmm-m>DBKW(-Rx>Bl;K4rPWowA#B}d!adRD9Qupqn&}yA{GKfVD0Md!bS;Rx z(ux>DbP@S3jm;}^=ptrL%O_gm75#sxJka;g z&BaN@LLH+ce%%?E+nm{Yfv42 z9vHo>OqYDFQZ?4HOdHO$M@bccHEJx|2q*ktUSQk!ljFBsi0J{Z;3K)m%wH|!jxuRxA?sep{nmAPBpun|$Gw2f_ z8zE8M!+Utxt6S_`DOi7-w^q)?SB@}0%UQK0wiLBntB}#2gJ5rZV>Zs@wZrE%;wO&y zJ8wJ|m2XB`C?vjbm}$#o?GmIISWBAQTLD4|X1pHlP1-zZI<1f?1WtEy-U&-!y(6>=M=V7+oGE|U`Ya&3CK+9B*Qnf` zRSGP`qJ>8Z;*B6$k@TQfWu^41x-2C`ktl&_U73+PTgtpVxo`c2KM-#`}Bq(Wp8$NL? z<;SMOSxMaa6}AmMi0-!d8KWFlmP{eZ#LhlT-*`fi9)>l?TLB7woeuXQT}*RfRJW6` zL*s1n4|G3v=quQ?xYh8abP3Ihs2;SbY~IM6lI4Q)2e*It8B%EEHF$>QNt`mi&-1MF zFGtvI5g$AFiAheRbzqA~#%jymZw%Oz6AM zdej#f==by(MiE^bBZ3k#(%{YnwmeSaQJ#41AKsP2Fi)WNhMhjSQMKmszkP^w*_(m} z?|Cz^AVYt;1wYd~rz*CsqiypkPzmc`I63HNH2jk^9ec*G8Ng>v!?@5?rCC?aOpIjT zg)+c+I^KPu$UN*#vLkt2<8nILQ}j&Mw?;fX_@>p^4Qh<-WXsQY^-wiFbzs%c*H+@T zq3;f3Us=TAanmlUs zzVCk;jbCS_BOuL+C~VGvg=DVv7qatsdw__ukLkl~m7P56I=Uf!rclnj{Md}Z|DC)e`$z%H~G{~?JV zUHeMI>H}np9o{x;VL%U9PeY}WdU$_Y{UY657ZwPZIislLFnlQMTO`d{4Y6ebq;;Ji ztDhs4TPWkU%Y}QccLzoby}$@6@ccq0H>I?Ea|+hmQa(Yt>LN$x&x?2c(DU%UD5!rr z!wh1<@vK-f42IoTC9}!g3-@|zMj30XvCy2u){D!7yMOTkAA&#gKFo}A%!HgZq)smP z-4$`sTlo@E=)|(%Y|bdj6upm1NjG?o@aHh2g2$ z5xj6yS<6Ed{Y%Mywtt*+dY*iT8CriiU-DBzD+@e9e~6{4%#<#GjykCm5wmik3QMEc za;*E?`$b$9>px>|8IRW{i%vZSn8?>jWWk;2gD8o{OSWj=o;pf%nM z-RtrmB@<@Osges~P7U79e7m}D)hHZJonnUY`Lqy9BVQ8lPQ#TW_eiQVOZu(SMNJxE zH#wmaq3%~A@LTjxb6~bd^Nv-kls8>PPNJuBwg~p9xw4+Hm@7Mw8&V2RIA2yy>`{`W z{PUuQ^r@r?;GTys-_6)1NFsk220vDL?T7zhBsZ#xg;{Kwo_39q{wY=aj$Oo{AIWB> zxYHqTcAaN|MdWCBJ@tVSUXgaqmE|}&K#n?kijlD_(2)P|I>TMddanfeYX^s4L+HnH zy6@RF3f2ogteo=$Bxyow*nuXo2#C_21IRP5zMbiThGU`A_~TboGk|}Pi25kM)M*n| z_qy?$yMwT=6RjbBI9GUzyoUj{ZKT$eY!a3~pVw=^ZD7yUU{ic#F%sQfawA80WT|Ob z-(E6k66j6&Q`+_Leja6Om^3JXW)>QO@6jShv6>m6dlvR;u~h*V)rr*4S1j=SZs(1J zT*BNb+Ha4ePMFUBx^#a#-r+iN$JKVWl;ENEiSjxKr_?cD*Dp|-OZ>!o%xUJ-Y*7O9vs`9xcQB8)E5)hA@P*Hw}bzq^D z>oFx^uil?dA`E5nvbYa#RFC8cgZCPqag*n)c+fQ)=T-2S_{x>&hoIW-9!02*XNtld zyPOe)Vd@s-jL;&q9q0PA+2QWMsM3=VTW5Ap@Om?P(B&+Gcyi|C2w9O(?HvmZn4hIUl6(Vag;5K<5+w9nE_rc_9=AGVpncIQ$!B8lyu$ zvC0CiUi;5-r+R1fLj~cO69J)y3Y$gW(02a30VeLtm8>qP;Che&$0 zd^BKif0NJC7{NO9WJljETKA!90FEJif{;R;d_hYil6CD?P<@0gvs)&~oE&(R z&nKR$w%>nvNI@J7T9nMLbVXc9+=3o&zo3QZ-FayihpK9xZ^g&-o-{Anuu?cpo$kpE zYEQn=Q*+8-p>QO<a2QDo{=Iy=H0*~4WN^q+Ni=87OoARX}5b*mPI;hu$qrf+$? z`ubfKJ_N|(JD1(3ccGHd!BtFE3MehZkG+SyQSfpmojAS`i7y(W-dm8~MGu4;B>HWM z*7>OqGzb~xsKnMQI)w5tx9lg_e1*~y+3T$mV>AT&kq1-L*>9ib4X>cNCPN$r-$v1UcKE|Q@6n$* z7C^>z!pDlnj>!;>(u{zK)>Dgbn~9^Hk`M(Kb9_OrgwB$?z=CtnLxQr zEUmA~x+bbQ(8d=mLk+YSQ&;c2vwT{ewn$mcP{A!BGBl5k-dUg1mSPwh#@)ZoH$TU~ zQ-HLJ0&=pk0KZh#iO@CM!G}qP1@-6kLY}A0Ey36)RD59Jg9!|}5F;B=9mf2?0{MT1 z`#RXbXE6$rRiQi&w%&S%M96I%gRpLfvozm0XaFrkWit%x2PH|L49Wmq&eZN*n)xCm zIT(i`nP%(va1>ld`?^CGm3>d%l=GshxNtmJ^neIev5+^uHJht}u56xMyThzoFWu8= zwKQZ2*~UtB^5w+o*X#H1*WC(UsXKoj?9KKD%i7wXJ%~2{W zDO(5{C)RsJW<=O^Ze7M~W$aO^!0#q3U#hk1P*<1%7ep^B^<0w!c<{xo-BmltQQQsz zkDONCq}b0W->s9NKN8w;k`2MU{al9K?X!qqQ;VKAHSj<+gU9}(uM9C!D$0M!iMbL5 zecNoBHAED5gH+CztRUR;&R2LNt&tqgJ7Q7(F`Ep-R5mDr)iqtZtV9|Xs5^H03hL(oZxdD} z9{*U^d!mXDe#@&x9$y(_-d2Cxp$$pBXBBP;ud-`;J{~H{@t12aFyEdH|Lj8Mc=lq! ztRL#J1A%)K`OzNJrOn8{erC}LwI*5jZsJ9IK&L}ejsu!SH=|`_z$c>mIMn$LTe;6m zQ|dUPKT7jmCzuFrZ+vOo(l&{7J&OOUks05m6W^Y(7x}{Bea3gf3Q2!eZpe$2#mLK3 zysx6tPLhW3)Vpi8vnI9QP7b&G9U7=F>Z<0dLw7p}w*mdIHcCBMKL zv59+VAO6_C&k?umNyMqM2$r?dcC@-&9MgvJX@lEpdugL$>t1R89`-|wYvGL?iul`{ z9mVRZ^Gll_f)ADqP-=fi9LW)@6%%-U$}g>(8GsNO#JQC%mu`u-o!e(<;&plx&&cFY z?J)Z_O9&hn)8nn}%)<^*I{~D9gq%E61?FS@%`FVtvOKh}p>B>R=)cx9(%Tu(^h-MRue`#1kn&@^U3MNiK2?8y$`PYEAb_k-QyyAZW$F85);lot-Lk4-i*fVId&pCHn!|$$TnU+X zB$gLNaX9OB=9%cbdP*2e05GaD{SGW6_!ttFSxN{eSKIL-i<9cq)2s6-dCRBTu6Mwy z4fu@UrMr-T(_#`Fy|S2Un76A|;?4Fg=wK`Qg^G(^7`%UsqxnJHfp-pPSVkq21#VB+ zkD->wnMM#4(8)S$JZjtymI|(op9*fFb#x92wzG_!R2hI+kup?6%V_s`X5L5RSkXX= zGAyKkVuTph3yWp)6ahSSH1{)+n?A4(`)5(?mlW-XO-~OK5w+3QlcvS~%LvGF8a)$l z2so&0Fol0FbqzPL_m5~7U8R($wSyFZ>+X)Ea3 z=d7lCa=ap!rC1_&taeqsWFktswf)Fr(*3R$eKCLav|X6o*wu0-k3Hg|cOfniMofVK zkRV*4MP>YPRRf4hiEv={N(Jhi$zUWU#2eR! z&(Mz2pWK|WhSR(~oJv2tisryYM)}n6-Fzaq1}`N9WT~z0C$_j}9Q}=FFA(1Eo0Fx) z*&l!07oaq51tf0?ntszYtYDSa=N(}*7u~~o&J!~# z=f`drlU_>1W71l!M<_AZ2q?LU z(y>HZ2fKPn8hVWkZaL-etO?li<`%U{dv?)Ujay2PeIiC#AVuG}B|g4)ub8g;>Ysmd zIJnrB8-|g--V}4{(;~|gS_g?f@3yGAF~zgLZ3$a*$DwpuCq7&~?ApI3R4_rLBeow% z%K{a}-r7<7DOumx;pKzN5pK2qlkATl3p?C>KE9psS;^+viV-P$+Ay+-X6>b2v!<)b zSDJ9c#o)KJ#E*_Ya|IQ-5u3}QNBmndM5-oYwQ)5 z{q_pijA*V>KO2#;b?mPnJIFc(MD(rf;w5m>zXgcrNc-3Y=hN9hm`7%X&aD&`7*Z)g z3G8L6P?_+3fgJab5jf3SFeQ@6E=9l_9ns?@rM_tvdP2xRQMp<^KEe+Qre8+-e4;|v9>2L#mU<7|NE6PNb@yT` zNYtoaXQ(+1B$CR2j8;{XZLxy15BWZ|#51jPz^f69w2FH289lOEBhW(HDYQQ#SQQ#S zYl_g16v3=gEQROL^;(g9aiD+4kKv=KcwKAlf*)eyP&IyG{*@NvR0q$2;kT0FX)J)y ztwPE>s=(ryRK&<{g72ril+&sbC zYCsAyEJX7k>vMpGz;e^~%XH-VUe}kO&n=F$p!CFv0Q~5dIB%AYY_xwC!+@)IA=xow z=OYPuL(};CyA?(FI-2}!qB*=3Gt%4BJd*>9G>=KtubZy zq7z9=Kcj4RR8(^~{p@(VS)BUeyaral-<&93i3(-@UQaVY<1>rZ9xp7r~`#sxTSxms#OLa`Ph2A-u>0H!|5|`6U_I?jk5S~m&iROdKF&9G!w)? zm$b$nDk1os6U70wf%qg_>f7)vy1t{mPwRlt({?pm39Ge9LFR;@q}H`D*&*k&aHk-? z`6wyeDD@D-r;hF-#@ zg7c2NzOKkKee8e2{WJlTl(0p;% zriO@3!TB`a&(|&86U7Y^RaH=A73nQCy?f|0l%na9O|K2s)YQ-k3>O$XKZ%Ivqnu#OB zT#3QXPX2$r$<>(9)F80bs4Eq-^7DIIKW9dsAQ<)OWt2N~GP{UNgMzbA5*j&He zP|tUpB%BcmdwUS*w#(mxCA8rN@=AIQjCP$OB&_nORDUdEYJ$`JF{;t?6a5RdwvtG^|^CgxMbFM0vt zD_hpj!qAyH+QgF#jb!$cqCpSSP z@OfR&o{rOQLfBL+nAyb7xD$yW zke+;?cL9SI{6@YFQ|&5T;`&jA@BakRAuire%=JnLhO^Y0(U?9KGW6xAo-Vd>u(^Mu zYYTyK0rs4*<%p9HULKB;V_i=mmq|9n7l8u?Z0(w8s9h1u>V%#~Ep*r9uAebxaCBd7$7dob`xSlDbcW940Fce9}NYfExmKM34gA_LnFM)W7NL~iR} zivFI$5#O4Q>!;r}LZYPGP*Lk*h}(bTMTu}wXDx@NeOvl4gH&~BUaMJ$*R0ZkG~AN| zUCX8NS^8DErnuxq$wRVsE}NdUXkCjeD9Ih=-pcsn%V6jP)9No?ej~@e#g5T8#OSau z-X?}%d0@ccFvnvQsm|7*UUHH0= zFbem_z=;oy73eU}8etj_E8o#{!@%U_&i`rbR+{^4&SA2~nvZ-@a0Hs4vk^Z9l)l3c ztiURWgHt*gD_ciPn>2i2)S!g7$Re?BMuE#Rbz#xKcR-m*%x&M|6!F~d%WGM|oWQAf zCju3u+;&G+oQ=SYvrOXCHw=H%F_;k&m+F5K+SkuQ1DV>u7ZJYk zFVb5WB9`&8r=1DrKzJi1)nW{4C(nVu(4pzOhso|@5=bIgr_lE$qsB8ij_QxbwvN2W z{Z%U?XL~>`v2dn|==_0bQH8XmZv-R*fd$AnFH!6tS;^8838@}GDHDI~+6F+Ae5u0? z;!Z&-5>t-0YU6I&HqIx^zYfzGPn*G%eMV-e!4r|74&5Dh_T)#G9XySn#tMx%XV|Ev zR+3RUJ}GEz-3MdU7uyU^$QMT}^xgF1vw{^9px*A)*LNPSqWZ!xVpLAmdMTLq(@TuFR(`3fONmT6^3@Zo)#5EgD3dGY@X~x;;^h z(O_Xlx8+T~F~KmHrWp!3$8tCNiaYqOB?sM2a{?vq_^aXPi?P3_jhW_53cn*g_h3_; zNva%?^jmOkh-|Xs7LfH0AykfD>x-ygcvx!s6zhR~0G?f}iwzB@lE0sb#!*{bU=;_WhFX2q1ybN@8e1HDp%@u+fyapmgnr>&s z2Z)i!Cx{NjHrQP*jGZm2uT&Q9xxtf{sOEkqcL8IwdxWj|aXt*b@B~&KF3iOFQ;~xO z4Tl&H^~F(NrR0CkUPw(fe>52qPMTfx2vMXT&B|Inl6?t(;SnKdHLLYw%v>|Q{+MS> zz1&>519vp_m7MsX@9w9%47B;@rj`Qh_|)A;Sb(Lj5Qn2fUO3&mt8&DDK;EuRinj4C zL>Pyid*F~}PRQU!v+!0YPld3NbzO=o-}?)3*3-U1*| zK~>1>y9?_}8|KA>U9e+i#q%UHG9tER`m7ag85s)0#?EnK3%k%1;R6x#j2o>F8 z+hbQwNBhYPu04)S#N1%YTkB+mvAe(dQ;lioJ=;@=Ce6da#I&MvFVv2$AR=+i z(r15vhn8V9^o^yZZBX!iK4E%f6s`*0ewhqN=9>__OX`76VmL?%p3;cmGcw>r(->vD z=l_H;3#=bY!yss@b*#c$g3g7vZ8oRNi8>9_lH@;m!i5jy3g02eL_xIiFqBBWCRwVK zmELuFzIjYlfeZbiWWs^&=&Q>n4Dp1qDgC7pY#q=*D$E5AV9;gvXm$M*af)WEhc zacMf64hw-pwe4vIj@@tHI_VOA?6bf{I7tfQ#&b53Hy`!T3POJ$5Am^+;m2$QRb zLvs5Vmi`8wulXX&eBEDySr0+3Su zA^fbVF5ejYQj&OvBRTT7zPf;J{5(I5^e(S6nia*ccC!zWc# zYtx)V6TTuEzw+b*@9Qo**NlNIpzj{n?+Th7n^|YfZKg-aFk>eB(3KhDzOGHG!BEQu z0nR{=-1cDk#3ptuJug?b*=K)lD~tfa{&Xj634{N*B@jIAcG<`vz5;IVWA%2{hKp)1 zi#_;75T=2yLYy-%kx?xmJ?&l?o&25?jC1%4*HX54jc&lVr#O{~sqvd!e$8w5wFEp| z_!V2 zsR2kcX@XliRzkBInILLoDh!AB)O|J*LvtPEi1TXC`0@!ElD0iArsFUha+pJp3`fAL z@&YrVw6E;G;aFpY5}<#0`a8|VCT#}eiRl|e2-3P97j`1!cwoQl+O)x58l7FeP)n{z z_U3?!>v=T+kb2NUERE{r%YAlr=chZ7aI_E%DKDw!L(G*sjomV$K>bOsd=7&Bn&%-N z`J}gXfy?w;3Q$*P2ku96kx$}W#Pv9}|1&X}Gw|QKpCWLzL41FoVSj84%{f1m2ieW; z1@xI$aBw?n){?}NT%&$}td81{36|cDdk*!xS}mT&_oZ8Tiq;m0mpYX$*|q z-O(-m%B3vi`;LDgCVs^m)k7MBv{@-j7oTLEe@ZBh7oq8fOtdvTNu@`J3WU{kajwH< zcDEBcycntnzR1}by)nsnG`hP%uchf?k-p0K91w9HWMQ(6_1ch#A51(x-v7conB^6g ztI`5P0pYYYK62yE2+A+rSHoq$o%Ure5PIx-2mi(^yRm=3rd{UY)(g0)HE?_#;fk;)$gM_ z#Ls{Xt~n;xw<=KQI{o1ISJ>~*>r(H`Q&L#;Q8~DD@%HPC2b$FzGrKSnwgZ%HX8%Mg zEmY**=)HgaQTG3@&dIP?B7>Y#@_TXBkA@6rVa88}xVHS&{|sVav4GCJf7*(IbapW# zn;_4|S@JMt4n%@V!6BPN<=YopBq2O{G%@|{^^q#Sq7^_zms>5@V7wrL@0@^AgIix% zggeDZ@-$YFfbp)awamv?RRuaJ%hGcRhAG`)$Tojro7P#1GjV74RO1y+RmenQBx{4q zG>aqY+%9L-%;<~W#|Y5@n4(VtnQ(zpZZiR9rg2EsCKTn#u`rHPW|xF18W=j2>j zI#Cd<(WgNo$88bla*db!wy@ba$09_)59_*-P*O zs^twh&?dUV)??U;9W=s8_b&G_-|wVv_8Nb6>5f45ri4=;$iCi>V1wZosq!cTXiK{o z7bSGvKG5-1b@Jt;>emb-RS3#MFcLbRuz_3rvADj_Q44*kVLrI|BIbXW zK7cwPJdOiQJi*?9>rIfx2}|uGI^1+ytxj;XbuMODW81RR^o-DJ#>WIYOW(H`BmRHJ*>^&h8lSh=K5#x*({5`9JP^m8 zFnmZ>VbhS1ALO#5_Bm{njTx>$hBP?4Kmc2>MJlePOz==BD-S6B%+s!GNn?K?!PDCJ zxZ~@C7)Ht#ofl?IS)xlPr+pB`n2BIW@4SH6$AGu5yj?P)!l2n4rTz?Gj8^sQV+D); zkJvBaeTyO`-Gw*0o24QocR+t`jmJ{ltuTB|vLVF@PZ|eC{f{E2SCuk`5<9L9(sijh z14`XrxbE0owHsqlsy>}XAmD$6V_L9t1KoqO=PM~12}b~jKzP5YYBz^t#<{CfSUw$J zn&eMk>-+iB2DJOaDebG`I)pep&nIBom(w7Sk=QxX-|H736n;{3pza5Jq88E}6WNb-!nQf-ijwdeE?kd|cyEx#y#NmG;Zzt?#VGX@0NxA$sb2%_^` z#a<)#<~`w5X0aPP8OMEEl>6g`Erxi1I=#ywahc^* z=2AFqTtKVu>X#AawKim{;C8uAk#K42df!H(=MIUyOz*^ZWO8iEJTB|(Zx!eB^ov0S zM(L%>Tu?5zJXN&lv=(**xSvhE``RE6?;}HJY;>|c28TyT=2Q~jEa1Uazz)7`^^e(s zYtd53n<`G`1=G1n%f0)5w!Scr12Y8&xJ`slgmZF&QocCwNy;F)ntvuU(LTJVT9?r| zGk4;Z9=^0>K;N0>+)3niY(}cRX_xwfUaiX;%A;GNTYv0BpZ*ED1eVo?J=yMr@mnO>cgWp5SG)a@Q(JV+xU@UWTRyq?NAxS7%|;2% zKETyEGF5KKs1=+!3&KF!1{^Vr#SyVeU{5WhGuMT!&917jb`Y8r3P_nd_$Uc3-tJ26 zN%h6WE?-Jil<%^)l;Epo~nC+ADH z$t)DI5^=xV^#;hOy6Nw|g1YFY6q9S&7o=J-W>2pk71Dv4YgV@*?xYG1G?2_~280@C zboo6gz1B*yCr~I}cKdFMk#OY#E$57Nnm-x_wt@wNG!vhH{j|Ed&*y&OO5A^LeetA! zO6e22EHlDaigP!YJIrdJc>H_TQ#$hIYNQTv_&s#*$hNHH^}ep^#hi_i|6pQ)1XbNHBvfEv9B=%K`3XK1cKE09H+I~&rOfm^bQRw^-eR$e#bBo&wCF(1-{?^@S z4`(N*vI8?!{s$Er9s&v|qcNHtCeG9oj?n!{i3>k}aqAoDt99uS=Ou)dpn8P1f%RzI zk|)_&2Y@gL?doQ?7S$%*rYKdu@6wB)a4^bUOA9Ce3-0xjR9@4Pa$Xb&pjU0nu!%;v z1m=Pc$IlVD?Roq&4;TnkFGV6zb;uz_WMukH6PlT*Y{x@51+?psX6~8U%lY@OJlfpi z{%!kz_yjs*2NA~DF8KJ;%d_=g!`|{Lh%tbA;UpA1cQ%1u5}L=7I-+%H#KC-XxdsvL zxS^VCyw+@8d1oe~?C>PYwy$BV+ZR_0V?2rCkW0a5uROYx4wfH0xQ<{j%CU<-XbUjG zT#On7{qorxHvHoA<14hmVRic1DvtI{jIgt`E-M2rAl3;#zyVm`cb2ou8!wb zRzmI0$%M8Lir(+9M-sQS|M9GyHn1Xpe8o2|TNmtlTY8+8Jil&3v3SKqPm5%y);HP% zb-}!*u(!-*^->7y#lylJLqydh|C;t)Yf(%8J2KmuYDNCw*)lYa2IeWgJ{y3mA*nh1 zPS7iCgv-Y%mK_t=62mkp<*(UMokRqg_Rv_!uS8UooZPck_VE--#wZ5a%n96o7i0Tb z4n|~xNMGs$ffgV#3tytd#-S~^R(pY2u4w_nT7~1+XA5E&An+)w9+Q`5!0p$UDxz%@ zC@Jn$Mg}an0=)}Xi9d3f^f={{@VPKU^X+*o{8E5;9z{i4od&r)w@-&Drs(>eIDMua zcP_nQN@<^ZStQMQD-j`YQ&jtZiJckoh6#Oos2Y0*&JO+4z1p0v4b(c`tcf4UPcqRk zHl~LB1KQ#czKKB}cJrDFah^AK3)2y??Otis8DrJvbXBT5f{r)m>JDVR?= zcSZY1)j;5XHni*O?@NDyK%8gK@{AA}$r!a1 z+LIaP(p!NfhTO~?pE#hV$oTi5H>IziDVhClbpTn#+4NYR2UjX=sCvvkc4}o0714f79l^*|7xr&5azn&!FhE2e zvfvx2M8`cx4(vRCE{u*L$W3RBto(R_F9_W7q_r;0r* zFcGv#|0@~wF;gLQXC<4h@L|96c7HhUGEdoNDr}U1N(=l%I;GG~i0WvUa7OBoi^D8y z==|wCls=BtAsPh^>ao_mgLThmSBJ_ROZ@ss&rbX3)A) zWfV1dCwOFcO(R0{eq0`b7;32&^bB!DDe00MXa^L9TG(pM`Htf-%=9@Ehj!=Nt8b|n zNlACESeMC;0mz>T*m&Dg@kZY>+7z%&`w%<9vdj5u0VYQLpmn1U%ho^e3={=(k0m0Z zN^V~E+0~tYL=G!wd4VF#5P)-;MRY@UN2PAMOegA4B~OxF&RJs|+jhN#Qwm6R7@E???nb3}PlvM>tRx_)f;*5oq2~e7hEPho<<<2%uu(z+-{Xs%` z$-gcdlK}epgtOIPjM019>TtU@p zE@tI_2J4PX^FNN^mbMShT0j+6F4{k`heYiXO%->A%Fj&Ej?<;GMJolV;H1>7g>a^z zec8)e#@=-}B=_`LsOEj5&!98}QO(_)Lk7e+*#(cO%|FXpX&fmVezohmWlFg{eo5(nf3za1*euPNH<3Mb8*3JacROEe8v3XK z07vHbB0vAUe|kzRQD~AW)S1SvWC+Q3ISRB^WjiFGd>NoWHF~`I2KFr9MOxesZli2% zscg!%i(W(8ywh0h91^tx_DNkP$SPUjF;>qUg65vrP+OXUF8To$4AUiN(Nbc>uV=e| z0(EdGU&&Ua@%8oEgaZ{8+7Mkdv#1Bv-y9$0O(cB7nCP>+YhQ`)?+XyTHs=Wh-(02f zmYfCHW9m(`xCLL~wWr$QlZ6~_@Q7b94yzvW&GgxE*CX82>Sv$cN3tc{iOA4ib;U19 zq_Hal#^cUOr39D&jWIhP;qmq^6|EY79G?Jp&S*8bQ6^Epuk~SQ9WGU8jdgk|QdykJ zeTSro4ky$aX3E)OG^HP1H*_I@ z#;-Nv$4wMI2*rflvla1Q@?0%(TedjAfl&=e`)Fq{T5PVD^?%Se@;q_WyFa)yb9DV9L49@5G3WC!dd#f(qxmYyvX}L$~X}L@k z9kda#c{CEeD7BP=;qrsM_4g>mTZ8pnBrUzRTZ9rr41JJCQOzDt@kpNr)28dB)38;y0-Bn55=Co5g2W~vk7Z6T zSB?nJ=I8+L$Ch9q;|c`qWV8ZLK9{bjYusk|PaAXxQdqCHD^>SxcwOP@7gNV|`4*sGeD*!GgAM^@sg5x396>`5o z@(dgLDWQJhI%l+2P3~5aejUTAT6I8;F#JT5fqOBiETmNWDOS^}aoB96l(3fnlm`%% z>MWsZ*QJ2=7g5uBct5RH@lo6J$k_2=i|(Tl#9dGaTXlknY0z?i+M~fYVZyHL%a_=< zDL-BROw3TN$swW%Z73~-x>vss$_URpys(mB7&nJwGKxX+hxk;*`24tU0-nYj|BhKj z)CUeB3OVwJmCJ_{BW$Go?k+(Ld6rLQPYfO<$X^M~{D;a=N6oQ6&FCgi?a2eD-*S{R z@MkT9XQk{R?4K@w-vI`k&`hc|P`KRL0{y(P6oIGSr?k0*;$%tW-{v*_N4ikVuVoGQt%rE57!Q)2us{Nc`7<`NBOpSwL4OxG9%S6K|tTgvQ;au z9tdU8kCMgUi2yYv`q74Gq6nYgRD>^=w=H_G5^=RaBcxS--QoOXwz#l^j|}3`Wo;P? z9Dd=}CUEzsRcNwfoL)NDT-Hs$@FmJGXjf)tg1ZQMgV>+Om|*L*D=DZ=#PVJbI}dKdIu@pn9zI4s649#mkr zrckxGtn#>j>w3^1{xSHIn1MP-N(i7u#heu`ypAK^=Du(};=N>3h~&fQVn5HQ}`ACIC5iI|_=RS3%H+8)yQeJf-L+6eHhy#TB!N zAldw2m zhlkucNmCZpp4bYO0E;PC#pQja5k|=pe3J)|LC0;!Kn0n{>)R6VW@=Cb{cc)ILDce% zw-Oe0MDJd(BF<;q}=xWPMVHa|dku0NgdqdgBrY&mjCWt~#s0%FKx4j^(hr5=wFc>b zar-7hB)PcXbONT;yW=BEeTd{-Uu$mbS+P)1BM<4$c2FkAyde+KzEsy@xL()!xN>KT zGu9XV1x;|fUTAJLcJ^2}r-fFn$SLV8Vi-co@Ge?UeuXzJ(&?ZNU)h%ZN8=&BLMkMR zv!r18lK^my1J9Bm-{v_=P}GONwOij8EQ|_eZe(+Ga%Ev{3T19&Z(?c+I5anxk?IE( zw*dwS-WvlrG&z@%>IW3J_8|!D90NEtF_)3*2NbuCd#@$_m zyF0<%9fE{FaCdiiw;-?2%-oYR_q`u)u@>}K`c>_!y&DP=MHL1i6FVcIxSg#t0}CTF zFF-_2MOIang&Dxi%*M#f%nDCIp=#l54g8~lr%(qvI$7A+^8Uj{#1Uxd43dc&I)i^q z$t^t`;T$IYxlAoh{G_o3==~sEug=EId41^#5`P2-yG~EsPCq0dj`U=0F=zMq@*3fQp^51<={!zd}%d zGIw^i=VfAYb8}-fv~gmzb2Jm6p$C7sSvZ>mlz~n_M^~T;;IFa)@`g6Rzl&pprvRv$ zTR8n&rDA95>}Kc)1b_tA7REqZCyPw`roX~~8ZBmPB4TG_1GIH^ zg8z%3sD&fY7}R+WroWHa%GS=!*6R;8wXiiY{i_rc7kef(TMGvlpp@u;tUx08KQc3* zGk}AcnVB1Ous{a@(B0Ua>92nzsvh>he<@l15`*gS@v^tG2bh9N0s2^&0zrS^y_^hP zfdFSm7od;VKNbIt;8|DzCKkrd03)E8g)RJ_=pZrB^xqgX1&$W(03Bw~5V8Q6|N8m+ zlOAZ!Ozdo}J^q;g>!6u5MAU?oCFuSx`fs(cu$?==i-DCHz`)AJ2H<~W;R0~-Z~}b( zJCLHG#eazW!&S=G)DFP&Z@!>L{SRl?{|o~4fA)k1@ZYiI?LY$z1W^B(cU@)T=}gk79L>mX+bS_<3$ zZK?tM_ZG_mO)Ok&{n+q{@OWK4o-lf zqobh*JTqvhSUETVUM!&HGy%H*Yn}m2jJ9^pAQu3rLLY#sog;tzUq{Qy0bmmPOZ0EV z1z-~SgSY`qqJIz%fJy9c#Ki(&lK6wz08CPU5Gab=9|Vdb{|AAhDEy7MnE_0Se-J39 z(jUYMU{d~rKp9m2Aa($g>K_EELG2IX1Td-pK_D&~eivHu{4=tc-sg4UMgUpl8F>`H!6KZ~4E6{ci=RMB~2^JBTXiX0-W7TF@~wnfwje zKqO4;tgQ_l|6u{j`G*haV3~mbm4fpxy@LyAZT`4`+GYC31w_u&!u209{!-hy{4)p0 z#Oxms#MAr_YfwwgJ?zavkHH@ckj&yA5X8#z9}v`9tABq$P!Ipf4%8EyKM-_JG5v|m z0dlnkP5VFep!)3okO4W{{g)b4jQt-qD40FyRkQuC<6&p{kMzF}2eg$;_MrR6?#}_R zgBoFP?c(&0_#lbHp92DQ#=*tT8E9hkkBV48(tmW51=O5B5ja3fC(u>?cM4F2PS%D_ z=KpX25&3^Z3v@#=Ih#8I|8ZcTs+`^I{^07alayuE@IsUUXpuW5Q1ARtVpXVQ-wRms9Qj-G!JPKH7g8e1{R@r#y~T89sa79d~lJsN#sZ zK(!8}93*&19KKjImzx!ja3>=S8hS7qgk(70DPzMYd0xazSnA0u=3cz`D4f)9MhYkwR_(T;2Rd5Q2^E5nI&w=L(lGeUTywI~lU82mE zfuVn?2-4lnQ%_XV+6c|359n2$Xa;lheZoi_ezPGG@Xqy@M)+D~T)bUl-K@{%oid-h zz<1L!&P=NvrkUS~!P?1HA zhzYYJ$zY+jmQ&pOckB2)anATJ`i;17S}$`+1J0Ys@Pq!U(&`llsAoWWgb$983ZgwT z1~P^^1q5f_bd%H>BZo3u^BBHj;;K5_9V~5z?nygTO9gYh#(mSSdfJygmHX8mAq{^M za#JjH@nxExr&Eo7M6MHTdVxY!xNqOd`4dot^d=X<4{&GmiJq@!q-lUv-6Y5;m==d& zP20VyEGK6dh`(ke!F}4ad%)hyg+{48I-mrq5sXtS;pqhTpQfQqR;VbA7j-an@l4iU z!;erjgHpK-#wVCH@ynK(=TPsX)0cnT95zlE$*{bs;?GvVi^OXnFY+d^Dq)e3n;3CE z!0b<>BY7OFdXv;vy6dL)HxhRSG0+x~G*J<(gn2jaq41#eAR`Yc^1|ay_JzBR0iZ{s zI5E&~g{4IZ-QY)zHKxl&JmPI9=W&e$6$tz}HdL3;E_U{%^6Hmogx!32B6WX<;kQR3 zFcfSW23opSCqJm)W+@^+RZHm|+wmu(82?baW73dIh8 zW`4WiA|a|6p$WSy4q-ss+w?R;cVXkW=pZ#?$Ft{4P)opf?DaRkufOrzORa`3@q*;Q zfvavgeF0;V_?677txcu+tfYS@BFBoH>jj;aS_Gw#nZ;#qkr!9)Gy+TAjtp(pDuhQw zBgcrW_t?{*to3Dv(sIMS8tp+Z<@c<-$OES9PDihhj<8cnU*JM-3jlTJiwj(Q{@Obo zymKmMn(B#vGG}j~1;91=V2aa|Q7C9GUb57r4m>(s1EK2d-<{X)|gbiR`E|He<53VEgiM zGz|TwzM)Z-Ir=LPJ?(#4FfqQgqM`kTKY!|SJ}eG|6NM%{wCPcz|H-}5*XyZ!@M1i> zmcVYts{^YoWxO01&)6Q;^kWv^GvP6<+uizzMtP-=NlYZDl`k<@55h05|9R2XSo1LUNfAw5EoOp|$W>M6$Noim^rnGpq`xwJU@X?>$Ql!FiANvhHvYPP&^3vcGaZ+D_8qPO2od!FpjuMa$O z!ubwg&;&jWL2r(Yc#5RFZqe!8vvXv-bL&u@)%>hjzxVqMRmbdvc=Pkz+uFX1AS?mL zLq2dN#kVKJ6E(ThRTTMc#Mn}tn3idz{%JKGR{TXrCy;+iH{j6WOLH7G3mSz{@#~nr zXiD9&2&wC5CA7N?k3s$7`_(k6VJQ-3I|JT*m@CRA7-du?50OgAJ-6fyQYP%Ya#fup zW3UbslXz9*DAT|g|4^cDo{SYa`BI;ypuSAicj&Uk-TQgwc4Tqx8-&NKB*Io>rQeod^i#{_mz5iKC71M6EsO9!&~4Cg1pVPC^- zW`{9L{b1ZCfu|RUU*1^uLaY%Kc)GUfN+IG@2O8{g9N!gd_IlsC$7|Eo7{3{mzT2UnF=K?Uwl@*+qOLCYQw()q0ad(ReY} zFKoTfnA$T(*TUWfHAabk`QGW;&iZ5S6MPfF3YpQ!i-zIlyp=*&>%>tDlI)N?wvV1z zs&{|Vap9&^vP5G{TY;ekq&zb%vz(u3Bxy>esZb?SNB)tCpM%S zdW;X;`f4aGySAU}nEk@$UR-R&5J^I=>piCIH{E-rI7k_ajr5j1dialM3+?q*jbavM zAeT`+B*;SGR;FXsu_|Y&mF$SsA>sJ`2Bj= zIJRR8)&cX%z(v1!+rrCz-ZHtPKCA6Zhkd20&Y#4QspnM-zwR+D3J_2|ZVkT4Ni9R1 zS1%&BTMbvZE@o|lH3KJkYDT5>ri*lc3J|*)ndqsCfd7HnAYGDw#}iN)&RjUR+|+;P zc%`Jbui4}|<(ZWSzNj#XlScfDHfO9WD?8>!fHY}CQk1`%wfoSBM~!|_EP!O|he8Fk z^3@jtp4Gzy^p=h}n!#efYZ9k+iCk|T=8YG+6s8a-{<`p`W3jE~k9jDXsHOfRMEQxP zYvBzLSrJEMi#dW(8l0JIGha%h^Uic*bFc7eXrh_`RO zI3gHGLy{`JS4z8Tq_%{6EcS1PjdBS#q*re9&8;#yw>9`8)vle|F$^5yuIzv6sW{V{ zbS=iMy6%X?Ws(lMN}HJt-)@+1`s8TrNaq|87V@b~us!uQ>r%tC_?skz`pi`73h!cm zoxO}b@Io=L<=&m3Fmb;QVD+YlB$7Doq1kT7l3W&WFFQ~}_`%NDcJyOT~i@vS{l~F}ucS@Uh((M|4-0si>dmO4Xj06eNvSZn9g<`rFXM}&$Tu8xqGkjJq zHZFJ#ws?p%X0yHFgOXFE$LnyD8GbSSigaQ2@T8e>&H(LS046X;d@XaZAmvS*MHREyof+K3ye90N>@Y4EU@V^+`9uCzEm<4oC-CmwMM?m=bv%l+Rm@~P1dPC%7S`Bcx>j#E z=5Z9F4&jKOWVI$eHQ#4eazONZg(OjE;t5_PM)D5*SH4mo=K+7Fja5(7^JphYS5i1x zMEg8ydo4GPAmb%3!*$wylMb${O|}L)H~+UnOGsT9uNu#|{94GG`mC$fGY z9eQofbhp3Yr@?e>-OCdrrhOl|tXVJ6o2&z^-hct`~v7fx-4@E6PdabDu z-c>uw`~;{Cn&&0^y1yolouOU`BaM`Up790)^ z$fcCpS>2V|$n`pphz^f`9dd$*K(Nco`=BdqF;>pvMJ+yD z+V{w^&B5%NFPZrLYv3i)u|X^8$MSO?WRH+{|(JQkx>c>NsrjLSAl1SQ{ZFhUv7XQA-JV!+vcqw!kq zq#C|LX<5df?G+RPgbj#t)m;L_7ft7@Q<%b0Ove0%>$d3cd(1*u*+c@^`bB>I=)g(C zocDhrR$q~vs||i=UfTtPo!~`Vzp!p=_~he{``}o;a*&!Qf)c+~};B zU}FOkeCX>8A*x6q*fiHE8pDL^W#wFJT~8X9oxg(6GJc9;ysco+Wi{El zn%ORKOWUMjhlh5FUk+k;!PQaLe{1H#bf16Exv4+UX^+48owE&uW|^B`9m0|FbLX33 zNZ62%VbnO$_b=$UOakVTpT32~3K%1iAiGWve^PzR#hz`nej67lI}{22!A#o)Aaku6 zIRn^A{ZQ{K;$NhWWt&=SP}VT8)!4&bc6t3M65(B6{A*4x2h*qZ=|{mD00d0UvB1 zWhd85msazN$KfVak?t3!+cBAg&;p@u)}qH4rcKQ|?L4Vo-bs%-lkf*RV7mZ5hoL8} zjA#Qlxd1FK7Y0}(uH$<{L!PO_oWg$r8m3sB?gtM}1q`RIgwxHXVJ}^W%h~Y+M-c|? zI@cWTL*%POI()f%2v?ctM;pN#A#0cP@{NQFqPnJNM+a&-r7p0+Bfu_y4lmTU^Yj8- zC@q6jjJ#GUd_uYvRHeTG)E?QI{`Ej_+j|Wy4#C_)P#S~i;?%~Bg2X6)x*C5O)Z#c) z*cqA$PisT$r(T24&e~f!ND(dFT2V*1_4{vJDNp%egPARCooB&QOvJ(wSx^nwXo|Z% z&V2q~CwqMr?VUJsCP@f;osB6XpN7BwYEmw#3y#QNIz56n^RCrRY4DgYajjn4qJdEj z-N3Y1J|Xj>e+*TD>L%e?T7-Z709~1VB;e*Hg4wc!dy(;>^ibDf%*5rdZ4hi&OQ*fq z#BRPxOOC_CRV@-8MtY3sygPkxCc_+JVIi2Qr0gBny~R(*X4%ku9s$Bj*$c4rnbTkH zT+Yf1z7DdyWT?>!^-7b=%a{0f-XeIv;vFcj+Al0OLK~Jh%-P&=G2?&9bC=*knj#gm;PeX#>XGLw;j_opy2J zTu6$)W+b0aPV$mMhlhVb!1_v>&~-+#>051)pwI>rF4#Vl;(*4S4l|5*jv?A>b1gDA zs|zGm#8pcXPjbqC2W|pZxUfJGvWMtWWYD4;xWp1 zTWq$MZTjlAE{^6i`L;Z1sOd$L*7B3Yjbl99+NtSFvAy|6aLj+BdZEq}d~r&kXzW#M zO5*yeQv&QNY>lL0=e4}n%-1d8R~zJkMmx4V@!*gb_#9%aoA)~WtviFt9yjtFZwWtO)6?_iy72tPGsS4Oi z5c_NQ9L6h7Cj()#Ksjb^V|8Si0o@y6Cq37slPT=h898{}oklu=8$5(Vf7-5iq_2K~3!7)(r?-c!MT|gfO$fJ599 zU9YBlf|PK@$&XLse6o5HepM`rdc@Dy^P@(6r8LSDVo^Y-^bWEQrK2v5~tEaZ@WBdb?k+m z4;5NbW1J2*m^tLEcDfGc?ZZUCgF8F#=HJF}tml*~9SPrcjG52MY?Y*67YDVebmu(A zC}6j?siXL(%{i$kyp`&CI={ac`P#k?$Dx0+LIDPPa+Kc#sjyDC!m0`*9SZq_#+knM zoIvJGnOlxZOm9Yr;(3O333GK3=OT&=RaHUfkGR}IJQR`aE%VvBZE7;VuDKEeugTy-oxEPJXTG(4HC%Kz;8~q96|LoUG4cT#=O| zK#63j&D!y*bpX@rkU$I!)W4>SVcg;!EB7aPg@?H|g)PM3;DCc45b_)<`T_UyQDANj z=|*K2o;x8N;)0H?K`DWK4A;!4&I^AcwnAn8ie3obsekfW9S-q#naD%s^7SMnr2Ny0 z=sxeDq1wQM_$qIM^t7U`-w-=GOx`fA9;+g3wXzF z5MrU-Wc~O_7Yi0&sq$xw4hbeTmpaZI4azA!U3{v(>%M;lN&V9I zkV!V8fAUYCI5M%^-rfkO86Io0IOwkI+0y47-CX_yXt zB|JTexP04KODYvvqd6>l3g$j=X;odHdu~KR6b$1AspOM2QhQ>9cGP5pn(6u`X{Elb zM#d>xdS-hqga@{Ax;V^Q`2c?g%f?&+I|Ofxlc6?i9@gijm*mU? z92(5SR@yJRhsslGS3BmaLXKg&(%w)%(GWe~W24X7VPP6>RHRjk{d^-G7d;`-#Ivab zx>Tm9evucDQ1kF*i}OGAjgT=LFgq)J>U`<%U<)(BA9Lc8^_IV@3G{z^2rpX=xT~uQ zk@>L7al=`EU%-HN%7+P!s8+8bMc$A=*Y$P^+06i!`0TCxfXB)it0iouJM5p7y8{l8 zcpXW~YR$mK6o$b(&K*;bB1Rd zGJxXqW2VJ*Z`UGs3kq$8&l9ZDh)WljV;{NmGqXUZZ}ab09i+-1lGxMQ2{vB|xYs}9 z971a4PTnw+;zX658Z%*O)5*&Gz97P5EO1Ofbfhw;F57!Dq=0{a(o5F#{7fvah5MQE z<4)?Yk5=-^vFiZ|;PXW7*4)QD)zsI*mkjr@8%JZ6jXd=EhzzH7w3P7L|N+9;?Z?@`PK2U(KK10 zo9668nr6U`%ltwpOR@N1EEn$j8nL1d3qInKTk{$OyW@7Qb?V>Xd^4`lh?ZCJ7mW>dwx4;8C$Xq=y1DuqA zYVzp$`O#e%npVTR7bbRx?q8Qolk9#}hR@0WS5ur7$qiOqj*v&Mr{p5)p26q-(riA5Y$5I7z3 zZDRfAoE60p(FFGE*WG)|wW~qMr&tjd($5KtD@`K^u5@YcTybt(IA25ueG?t{l8o&3 zdSjP$2qVATw59qa@7<|HmSmYzy6`*V1!6zuMKWUc&k{RdQ6xLLNy4_6b3!Y1;N{OP zzD$3y8hAuzkuUOWa0x_QrB%ki@enWUTo|(B2M>hp1mugF6t|iX_6B~wIHxAqVgE=K z&G=5@B=lsieI^jlGgIIHUHousw#0!)0ae?FG}m#A?u^AR1&a{<7=jjt)nXWYN)1;! z@Jm3-etPR*QC4yhYLwIEePI!Vh|=`I_o9ChIvM?sZz8XA<}1IKEW>bRAgm1(;5UmI zb~fYd&ZV**D|(v7PKP}Nv9zWcmz!N`{nVm%Ym7WToD33!IaE%nJ48eR51@c`oW{46 zoBLHrW9pQlnivdl1-QKUH^#xLin{5Ix@L1O^%(RID@5{O%sIi*@|tzoxMAjg9yNcK zc%hF)<2TUk11Nq^?zXCN>r@?^u=&>M9&4V&O!I?3WZ!iV5KJZ~Vyd6Z2E8`!f4pe1 zhZ<#!waTSkEBuR#7E1Lux%|lCnl6rseWX9c&1wFA*{L z*poGOr05hx)y2MSty;;4eyOa9RZD-HBHh^(9>$Fp>wN*FG-2CwBm}MYhvgux+GcY7 zFx-C?qivRc`KS{E8S)t>7TYv?e=N2uEV#DScfx8!uUIbOXONnlfUUnXk3`sN@rbDE-d^tlxj*sac?#2b20aeRC3E?0(6+_TmYPGvurpy#sEwEBM!&fO)W z2Y2Ci=hNUP3ST}U#SvxK*jDie#_AKuJI778chO{33E`H0#S;=hCzcweu1^P+G#<6| zS#=k$WKrt4#mW(9c+-3^Zu<8II5``?R$597T9v=f+f7p! zijkvP3FWDVC_JzpOqFV~tJx1RDq-o``buj$O|D!Wjql!`vfYLu8N4ehTZ*4MW;8ee z)2tp#nSD_Ol5H0VFmfh-{0?5bYS?)@yI;P3JA`59GeUa5K+5Ly#Ht;c35$dU{P-TMT{=qFZp^!0otDuJ*`-@n7(ZFIg_j9>K^(e6Z>I$q>l$P zO=@vFbH<`^ids%coi=}7^2dJ1BdxU=-NWB2dS=R1H11}7?sR(~EPs+{aLdRj@{itF zK?X!wMAM>+codVHT!wG^Wo|92cE!8P)b73}1ETLHI0*Z|@`_hss!C2$##nY&0tktI zUJA6?eg$;I!60^Vr}mfiQIG`SV+i_sIQoPscyaMv8dGQ@3@d*COPt72-fz-mZptR1 znEDZCS(tDXR;&22$n}!PjThdXg$r>%ZU63RuEcke+b0MchAg+Bz_^8amC#Y8<^v*$ zlytVVt}!7*2f8rOe-|XmPbyODET)ren!tZqlTHXZJQiHKOi&eA{n~1pLs6$`01uSi zJ{1%DBv~UOJEeaCu+mw;kzts4XE{^CSGYK?zKES)_(-Cu0E@&ax8Q7HVl?+~;^KE# z1IWUOk?Z-`E?@_f9VM}@3bU)h#muj{Jc$;9v)cUI;CL<}G7kln*Y08P2t_K>Nr-|s z!CtiPX(2*7)GQ`N>%Qgf3t~hs;FA2^mtU7}bdaOcw}yYwqQOpy6@tI4=+zTFFsRR( zQZ-#Yz1GG&nN$c4ab&d*=Yl9UvxhSo=iX}(ufTu9or+y1EH<_{idYVn-pM6%@b&FC zB5uLC`w9zjn7}f}3|EZMC)n*E{K?Po^F9ksCotTozbVq58{?R~;+3c#ItPg_(OeOq zQkuROj}U(nR=LuE&9Kl@hPP)Gp*T`wS_G|0Lw*xm^mX_7ae6Agh%62FY}VskmxHq1 za{le;@rfc`HKDFE4sDRMd-SEmJqt@RGGMmKb7HyicgUv!mfAvdpQzIFqDNX0nmS*Kicw%Hcb&?pq>???Tu(u+l)1M z@Y^2yzX~j^4|%d($JwW}b2kqy1*2?RD6&>W-`3E0^B@rK&+QD(2mSu=zb z)7!_{CfuVBL_|OSeoZmjt?!|WEd~pAwm^T{?TpPopYf>>x=pMibFbO$v^?Ilu-xk_F612CCGRSVcDY&{2X z3LHN#o2XC{{7m!W*P+p8VIn0tY2gCx`FRO&f>bTkU(T|BNBVs47{jCVF8wk7qX7J_ z8&{cpJ=@`Fk$)Ro&P6b&k>0_8^fG_lnn~1bbW2q3)2DVc#E%>oqJe$T#$1_K0aE#C zRlSJq+q&qPny@J!b0-7O@@Z}%jwa}+g#(WFv&*Mu4M$AgOFkYTU|jR;v#iX#ql z9K2U3+ZBuhBXX&w=NR`hI2fvhZ$4Q^OgIyKf?5fV!S=A~%}?|mDT8Zu&NzSlE*W-= z($y2NBH&x6WzVCXfydc592suAsH{%`+U)2tRH}obucVpDH8J)FiCG+%1 z-=G|#kSZMg%_jl&A#6W${40N}#z(h;TvviH1#kDUQ*r@JyoF3JyP>g%vs7}JioK1# zpzeq%CAI*31`z}%a(M~Uz`HP5pzIUb1;7=ZP85pYAFpf zlSJBFbuu@w3qn8W1hMocTwR3hha1H6I#5U}ms{F4HH{@uJ2|($@3X2IJm?qwPbjp2yZC0{vIahF5;{!;p4MI3w`5P(V5@ za`QhK#gdg@8EBnYRpt_Ml#mvuZ`V9&T2O9NvnX+vQCy{yhuHx8`bh(n46os&J{6`% zS?5*6W%Yc7`C5NNbWTL5K~XzI4zw=)sx!Sx?x|bM1&xd@)}BQ(eNg*aw4eJ8%bZm` zx(2C^h?wq;-*rirniU4qftNAYx4MK+k5^o+F8Icpd1gWKc56Kfl_OCT|2&QLt&2pC zu2s$5xuTkce`I+MdY4DTk)(n`F00bdylw7y`TFs=9D;uiQsq@s3Mq)g!$NTe`OuTVc{>S~KM(;kKg(90wn%38|MxY^$@I ztxP78>LG_;MaW+0g>!h!GqJ0Evi8^P7@AM|MTAJ{y8(pZ7X+oAfv2K-U)ak-YZPni zt)eYT_T_(k&#U&|p_nQ-7?X8^5!%Db80wIfiC34(N)ms2^y-eUUs*@Xp?TxKvZF&W zag^catuM>1xL5Qx7_=}X3kRJwrsm-hdzF9kw+s-9`So#3t}s=oi9PO!cA-g-z`FDW zKUQ+E0#P7{&Ge0gXzyK`u(CK~%*W?D-Dw#yEy;h$6ry@8N$_{LtnSd`@=%e*)9%U8 zn)dcfB?3LlEF5Ql&P(Yvv%YS^u^0g$=@r>-OC>_ZnmrM z|2>n4xc>v7;TIs-R<%Ku&-1fUfNUx}KW)J97x=^GT(yvQrFG?Sc;aeTjOm}gAPvEy ziV4yql+N$&9hZ|-4)w9dT*urX6qseS?unCZ%iQ!oKJSXJ89$M$rJT=)u^biBMLmG% z5dETrOYWU1;GF`x%dUqS8ClkpqX}7m&2wue^J)N^KSb-MOP<0Kvin|4z<=d3Cv+MH z@e;#dGnxg+Del^?pjCalMTkrPQa0Ya#%P|4FDj|OqUzD>+)P*wQ*it=T3x+O-_1&G zNJ~EGm+FK(eos{Nm%&@~HXE96HYD`3ngPQoMN8Ft%z-3+kzVs-U6anj+|B+f#`A4z zRBH&J$Mnv5z}R(v8Nk>0+E}b_DJxsObcRq8GYPVTXVAcHK zRFETgl;`kOrjxqwH32?{?r?#Uv`WXI+=GIB{RxB6E~6Qet>IgLjtWE!X*^QJ?{KU) z0krQoj;;p!-uZa6?%yhp)?da+7lqd?n3-IG^e`8RtT-X*dM)OpFk!1qajYKGaE(r% ztO>8p&=f~vB_E@?FhP2r&s5J8&@&(6mM2era@%cW%!3Uw6Obf*`7`)0m2C*mu;^q=ukgo@Qlot_}*tof6tcM0r zs{GfzDfb0XFYZTr?-gimfoBix?cyiUkot<+S?p-b@*?mDH8C7Hb5Jzham zQrLI1C8IZXulG87r_3jOPsR&@Z9bGC#5f?HcUIU}W>hVKMfH)jm8YB1^Zh^?Myhfs zi$f#xErWzT@xJAenVF%}#LrMVu*Whw$5XY+wG(B^VHxLji)C z*RKrrnTqJBs+x)J0U1gu7h>xx+OYwO6s>(+*D!?YZ2YFyD;;)Fq(vgfXssDAR23nQ z`O5sqP5I9JSbo}$$V?OfPwz7(@?Jk+7t89$^T86fa_=*~JmHB|j~^`w?S;3zJ%yv< z)=2+|45<0FHUDb)%w0!17!EpG8n<~kX&q}dd4XM7on=^OzLc|B-3_Bud-;?P#+l3oZmz7OSRPD=3$z zTZm2dtxo-j_m9!y04K;1n|VQVbbHA*^$_wofJ1FZsD z&f`4Tx;_>y-b}(3mu-ethqIlWPr0QN7 zVD34`4ij4MDw9-fR@_0y{YjED3%{;FIG#Zb_k)ojvT=0%?x;!o@$d$STvMp0>?!b( z{#UoYuvWMkqh=aH`-6MeSNy2j1A>c`F1<6 zl$smX6i9q&z8#Ig;GWf>5Jb<%`i0}JbqG7jBDyH%dt2NuhOKOPUS1L(UTsa5H#!}5 zWPIU^2~WDb_ETP9=+|@ayw~shCT<(<*0c(}5;WnikTeR_OXT`PwrM(>L_fVk5S|Ci zPuw5}g+kJWmOo3g_~vb%!c>HLN5t_>#i2PeqW0}#Vpn-+ZC_ZW-H18yw#l7$z4)S7 zH(&~c7|VD1A8nVH2r)UX-0x3*5)hQh6dvg29|vt?VcL;wF*3imOYuh#g{K3KHLpre*|3( z`!uk)Bgl#Yp7@xq^pXL$$mUbiryugosn|m*=J@I5h_$rpwaN5Sy!H{&c^bxUkF-YuM1>{rWc?*i%S>PP|zW%WS-zEbR)5=GK%9q*mWQ z*WB3$fd}Oj8>QGyqp`H5~xa|7L{MX`ZOJDG|rcAyQJP{D-IV<7J8^owApi{-1kQ^N$o%k_0QJlDS70-idb zpz$j?6xG!Fs|U_Or65PTTc0om&fa&dhag@csNSIEvj9t0773pp4ykZ^z6SmGvqn^DzB>O`|AwMQ{hr0Egzd^o$Ue{&iB5ho}XAZKmRIb8n8%=a@m8Pm6%tZ znP^Hy$XAPuEc~0!GAu{``BTBGoM1GMgUoB2%8u^6=&f@%Ax!|4ZpprfctcaEr9eSo z0A^m#X}QPA04Hh{lzhy#-v7jtI=)n_(e6g6V7Nt%rb$d0E%~AY%Ka*AK9N2aYgX)m z0#lR=(up>D2TKYtr=DG>UwW!(nO_iECsNxDC!~5|->K@OFSc{2DwmAMW4#6>(@m9OKthU}(W+(nhMKFU-kn3l5ZfWq!t5vM za;Gcu_qL(5?vYKZ<>}6Nz0uV%D}7zsELG{-kbB4cw0(M_Mf0YoXUpR6@w1ut_eWUU zhghrACkUL}or)`74*ihpmK*SJquu+N*rDL*(1p`3$_|Sk8PeS ztUi)YJ*zj2c}-N+y{z*upv#Q>9c_SwD`#AJ*R_wmYj&FZq)3qIs?mkb>Xu>SU^sCj z9DOxsGxRW&(@WE~&)Q*`CcEWV*iHbx1dTBbm(t%c++Mq0rsW*zA(c0n6F6f`N!We& zN^eNsdvMKyz;@oqm=n?N%7O>LlecMn!zDbBz&vI<#UJsR&LKjF+oBq`hZNQlShpcx z!wqZY7xfWKqzyu*lhBHc>QZTwyj1v68P01S-z%kseK)sE=Q|nfQNff`7T~R1?sHN8 zqoS=QZ~A)6h})Uafs`t7rCqB!^=)N*Xp?54Lvl@{GeGuX&8au2JV_CV5#4?*^=?r& zGq8@SOu+wX#QGS2&Lr53Tj ziw=LNV)5RjC*E;5HZL9En4+B(6gaoEU)=Hh5sKozZ_koEjGWx;=@8d0z-T)k=XUB3K+fceVxWO9-SQ9e96wuEpCTma z9;6;0Jrn=rCT}S_Bk~7-hSP=go`764c4RQgNIij6*(2fh3(?NK9Rp79_RX0Y<{uKX ztVOrNFP0=3LQBFxMXD7A*)(*3CAU83&D>admu|1Bu;XQG4hRdLh097FK?CI$`czMr zG7dxlg~83PeH}RrAzhtB4Jwlg_%fyLe@|A|^(W`hTbr72c>=>KV+|A|U;cO+S7hg_ zxoKQ3w%|~4xT*hJQR(kL;yS9|zM@4FAueA#nD&G?xUN3X&Er#!5(cvHpYVPT(Bz|$ zv@pq0Gyo z%nyr2I@|Bh)-uFN7aPjA*b<+tVM)_HRE4jB`9J6;-&^*@&Xn zam3$IT>c3J3a+8|2J2<8|H7~UG7TWbO-XM@bf+4fYFuqCM`ym-zqPR=wLQs1;v1qA zlA*N5NV0i8aRlL?y)Nf%rpqWR4w>4Q;_ihgzTx~*M6x5k#wCszo7-YcRfHVld{!U$ za`iw)(6>ohLFe;*t`9A7lpW@G{|aV**Z>6Yj+LiYh>nuQerV?i=O2A${`flCcFK&- z+Q*(KXX_OTU%Kb2K8y`+@FsKdlqdq2TAhNHm9{728}}-p;v#9$YieH4P7%b~Xwodn z2)KrVKE?gVA_C+F@DXZ1oJcs93EzG~eG3K_$@U5X``dnDcPqOeI0En6nv?p!LG#=J za*T%BoGKR21q&zo!;!(+YXwnsMCYiLCYph>MZQYM5~}zbA(o$#xAZ^6koZN->116B z$6Fhsbe8PK2Yp}ukSX84teB|hTVw5T7u=Y$B)NN6U*Ldx`8W+%Vi$g6eaDR%zGeju zU2SCqSQingD*ZulOgED?$l3;?hqXOjLxD|C$rTB9M*g8mWQ6J8mH2(y1)V=d|4#1N_hj~N2psBacB#_(swtC6PnlMR{%ci z8zORRnsoQv%Jlr}gP(d=!HdudiwMUo1CONzF6Vs7WRpBwL-XBG-T&<(TwkYeSv}$RF6tt*bX^|PVjAb8G!RCqHG@yY`ap%pEf;#-q#m9Dy98f~CdiH^{gXyG~%Yh6(28ce}1= zbpKg|vA0c~k4t!?(x_5Qul~?HGsE(z&0y9KJBfxtVFLGGQ&ic6IhmL-Dl1iFuBAR- z!MP$PhYFDZvq4v$qle_(fJ=o6GZh@XjYwF;g7TvxXrv!HB}TS4#LgLxB2-O*lPm&2 zNp_Hl@%8P-`0=@JG#McG-O3a;Y73P){R|-*8bP;$5_LRzkpgG<^-W8=!Q!+R?&%sM zQ~P-8%;_hH!rsdzPM+T?G(c@Xz1439ZjD=SQ;8Y;CVUlX3%%$`8P!7rDmIp5$3AO3 zzJ%t;p@wUWV04i7q4bDG8`+xT_Bj@Xjb+ahOkOfw?{h!*d=yAn8@HErW7f}j-BGk* z>{so^7H}AFoDu5ubliM8@(-a*YfM>!iB%-IAhEh+58p{Y%UQ#Yg^|ci%PYEUzSDpV z!Qw5ep<>G9QI_4T;)Nwo?O}F$JZY-(I*^#%1zmi&=)Y9ZJ35d2*Ksa|oeF>%*>mKJ@Ul z-ppG2nXsskqLi$}Adx1*j;0i!xK2DmM6xcEKL+9*IDo|KUN5Cw^j2x!W)2#NGgMN6 z3i0QL$z0GhQZ@&=rsUjP@$6Vlm>;E>{QY-#_L0HbM@?Kk&@o^7@Wihe!C7+~7($l> z@+lt1ms;Jj@1C!N?e`+z|H8-WKXhmeiRR267(!*};gAV5f*+zeyI>r~CSbh&b}y_E z#}(av$_9F7O8G-klMikF9vZ{YMdpPhl%kekrYY_7CYw2QDTjpL5&r43&;>Ji7P_2Pk9q011)Y*6ibFG`5*6AiY#3^!19fX>RmI7vqo%_Nwfr{A#UTGrKs zXbsf=&Mca6=v|d8ae}S2IuMwpCHLWFAZy41%R_#nsehmQh%L#NlHsLRY?0oAKZ!Pu zD>lRW;D44=gnOm_3BE3OlPw+DUn!xes!j7W%)_LSMR1|jg*|;a2S=CUguKkSK9n7l z@d9uwu zVTaq!PGrEmSe6BH4d#}?Kqq6rG!O}&V+8;V6e=)9Z?ia=?0Vf*mmXxP#6YjL zoV$yLd1@#gPLGcmU(H*3ZMVUH+7cv)oA<6U*yrS6(r;st*DK8o9jEO( z_N7e0x~|xkx6Z3AT6ug+AnBg2UdP&^pGx6I8`6WRk9gxVXgmaoPHY%`k)$-{$$$m4 zewv05!)YVWMLtytJK3!{$w3y^KIb!o2qZJ(lC!(*M9r%%ADVNIGAV&90dZo#{5nprIgIU?^!W)6@#WdA7Uu5)I6}uOH(`NWkl0w$kPZ%Wr3B$JqF*k zeh5!%TZ%F@-z8b1v!xu*)XYxlF8~W+-@2jL2?#HlQdSf40%z4$=-4rlFEUo?Ur!y) zX)@l9c&wPM>WC*Edw%eS952gWZ({QhsdF%H1wT^8lf50T_!wFs&CNt1ZBeJ`^+rzx zb*9^w2B%-cLn6XCaPG`e$VzjAN06;+`NVf6&Y$<2&0ak{_L&i(mE@p^AOp{-^DqfK zko6y)0!CufCzz32OCCSksS~jtC)OsgUOLQ*1sg(JA?=vonL`L&-mdS0Rlm`AXQoNW z8fg?Y=3(hul6R+3Jb=AOLg};_quQHU9Yng7NQ_k~I^Se1?e{AafyIlaW#< z(9e_Ftk2toG9~V#zSKmGy6NZ4C1FeDu4BCyFvv_k;|+rK7x4?saccSo?drWiju&^G zbtc+k;t5s-<6#f|+_YvB>C(l3K=~T7EFV@&7B?{|8`##>&L?zrYf9CKk>#G-%kb5KJ6t zJ8#h601p^e-SgGvBT;O0-*A2N@J*gs9yHSl#875(Gnay;Bl>pKPAH`UnRyZ~STxZn zqCgV3Yq^^PzQc~g&!v|xHtX}1X7Hj7^{7l$yB1*_iNSNs4Q0UNOV5;}B zm4bp`s5DfFP?3pFjv0N-u*+L+$_&owFnF=Tz}Op%@B$?CkfE)WA>!B4Fky%bFETI^ zW-v)%Xo_!;P!OU4g2&wfG$^?Kc*_XLUyk4)FACzBSihywj%_SKUi^*jU3`B1+JmqI z>0ekx`2`4rOx`dga)8Bwn}IaW{%N-yiOK|F7#P^|6K4x3+)i;GGF&$m747Kk4C%GA z0DdNiB|+qo^x(%jLgVz$FrhHwU|w09`13C#J(#opRA%rmVxbG8zlQei0n>zp zKn@_CT}D-)z&rr&QHM#lIE6r5|2fj#(++?`fdlL?VCc~A-#7A(3?jBiwt5X=q4p1u z#b5doOa$O=JrI5^$j}xv_~1RuaGw)l zp&mUO%oaac39hZh6AUbC*D7O!%L~@m4J8pX?=%1_2NWDA&}A1HfV_kiaU=fn+36Ha z{wub7$lt3yDxGu(h4<&G6~Ls4aolYy&x9yQKtoSTiVV?11ZK#OYMx;5Mb>u%_}QJ) zJ-Lhv@xGyS0VZ*VQ!Cg?j`%8e5(n+w3xF7IKmRS9zX@y;lOhZxWU4f6%_fe^lt5@xO1>FA*bOeW1Y- zZ|Ak$DQO<`fpvURz4+#Hq6T*Qe@wf465f0oKiANGexLmWks~QU?Y87SGWLA#!GL(*w!EQS?_=FA!37Z3udki6mV=XQJ)O>D@RDGmrG#rWG1&5aMVRUD=T7y z^Ygn0sAw6$!Mr{EeWmKmWAq{*--I;Pgn#mPk0PR?i5D;}qk_2&3HpG(627-}Y zEcXiv)d>&9#%?HAzP3mJ5T`;O1smIv3U~H<0w=<}e7v=2-Ee`nCztp3gZ+m5Ot8mq zbns;O7dN~edmZu_qdFGG$qt@H->F8+1+)qJ-!CG%8yJ+D59T4Uye%fH5{5&au@aIR zz-8a^riXnly2p^T#%5@Hwo4yt$_U`GkbQal_MwbJ3UG3UD2a!GN+n_n{5r)|ojsk~ zLhT8CQOSpJTmDy7pOui!4(d(>2gu}H?hI ztE?D@t@DW}jQV2Kf;@eO;d%o|#Y0vr@^D7WD$E#MGV8Exm`ggk&M;_P^^O z#u?6a+TD_JtYayF0FlxzVO_@KAyG!4I0||-31*NtEBj7Oq#B>sXYOppA3ZN+`cjMA zs99&4`JGDGuY!_tNUQdOxddK=9`n$J^EO`{QE4VhVH%#q=19BsPnLK}N}ks7ZkK&5 zxfc$V*{kL5Vwb?$-S-ecm)k!zA!<<(Um+pOtb3WQT6fD*qz)`5zIsLukB^=6ysB88(lenu~&CX?8_#f7buzTxi1wr z+gVm2G95*zKB}Bx#O{(!eOFD9yC{RU3Lf zbZPHSVm!|PU!Psd47dA3BKobZ#JwX_@2!_&yt>eqskl4{BE!+wPB2|IoSPZv4%?#I zlA!E2<=U5(>oVcFlC$+_6`w7?nn>t;T=@kUzr_XDWKqk(d`TTcnteoNw%kw=vUXG0 zs&(1Kmo%L6;i_9&8%^pu{=IPb@1rq(@u#MWSv+|M0Hw5$O|x0t&MAm(Y{mz+c+P1_ z`f%jAx9oEhgJ|MQK0t0z-VD~=`zt_fVR@N=4*+quiHFE(^TD<_r`MXy>_5p%87 z0QYP@> zYn?VP(5nqnfH|y~8SJMr>+m7o8p+;?0(**G!l|nLj_D3}3bcp{+}Y~VUao?`XPxJ1 z1@pZE&qQO0jrwC2V@6P}bVti~<~RE>i;|iwzR`OW&DWO&8dSRGUrMn(GMC(`H@!80 z85wPEU!jE-P1@spS5@=I08v)^LC}NNb#JMUO+>CBtsKOpN^wgT>cf$m-LfG?2dL{F{pCiGIOPs0Di?or^}WZS&WG{rj8> zujR6uu9WLx9#Yy}frq<_UoUDfvhG}=tQzg20$K|nrc503hDE%SO%xQ<3tkE!9@2z& z6~d9I0ZCaRAm^wg8Xs((hy^GxKX^-f`<%6&6SMu4sXBPu5w+_a5%_McM(Jx}&VKbe z9`NkbSV|suQg`{sDraN6;n-Z79aX(--izY*U&vghP=zIz0>+D0ypgT-Ppu+WYiP-D zpZ=)2tfFNC+U?JySdX?}1wtr*Ye?SH$FpE>^DTquNn5!siKejCu05Y#!WKs#6A}6l zNd;XXz-}KI1)xl9Y_}Ruo3v}6#YUtO7}6{I;1!sy6(;^;*{4L4)tchBV;NceC zmrVJKOrcZpZ;0;54?XhgQRzzJ%kA)V-irJtG*SXAc=)@xTml?r+Vm*^)ngzVsif6S z3miPIjXdn@B3_BW)RRWysYdWF0jKk?ZZ$jN14sw$t*RsJxv*+`)vRs1)?CQ&oJg7HCP=pUW`&@*@OzVQZ?42b@oq z<+M%?A<#WvWja<`PFTb#cO|5wkypByzMHW##A;yvy&9_tFjN; zm=eKd#69Nm~`sIGKB!sVeV@LanwPPypF)AGFk&+Jsa(kutx2+3xTQu*#sPix2ArSUh#}qcHl<)S(^f@`pWUnG0BVamvZYxo$<<$Xz@eYSBVja6z&~< zyNwXW%KWObaB8}hErtCh`$!G&QAa$^3oRD)UTrnZ)GbBm%kg#RtF<#SjW!+AuYP9h z>MpTS@Zl9cj%|m)*UeZA4)L}aSO)|Vx$e!1)_%jrS#b>v{i7@i{P@hL&8V0I!R?>Q zj^Ne?nO+zaJLk8a9hjW~8}4LmC3JV7^fuTI1bR)kza77<_L|c^4&{*LA_VLI@ZC&{ zetBk>%NO3yl49KZ0|mi;>anCVY2HOQY?Z1(P9fq-kU?h$Jkfh%E}xHo&iuVPEbCxF zw=9F9*+1`E$kHeF?AfNi`Sr_l(_XseQQt$*I(n6DF@Q%nc}v~@~*F8WcntA*e@ zO0}pOh6G24{?Bx4C4B{J--QSvmxVY>y|3DP>7h^g!*`LIWQfL3_<22sSP@0BMS7f@ z)tA5R>T<0EfG3KGaf}^;Ym82qU)h4GZF!_IYMlp>S9W-rINf`4WdR@l7wBJ(Qle@s z-FzhNliJp{ONPHAqKs{FmLaoH5K|LRblQw}QJqW7bp6DuYR9Db6Xv;`x}JwCr5I|o zvjxx*a%?m6|9&q;-B-45u5p$=-BVtsk|mSm^kUcs1In{}KSC-scwl}NdmFl-DA4__ zTv3FL;6CO&`J0ecd!kDf40j$NI`wTAoeM((^{!It;5H_9Lcq_ zQC|1~5NXw)LFEwP`@Qg!+yrPeX@qvANLjAgv;EtMbD2AiQeRXAi96`GJk7QdeWTkyA6twKE*}g zce?22wzs|#`wibsZgZ^(&3W?B*@rI=u6$TSc|aH8oPCZLD4kOFy!mBq2Gjjf8|kC` zrsiHILm!3+dTxffkmAOR`5rU(w+jwHKbKoHu~aP=(i>}o`!P$Y_f=wj)}vkj7}L%n zGO7|RN(7#Xbl5Q$D;%9_1p?lO=BJt#glKjPCMTGCogXxQ1gQ-K|jR z5ip~qKZ>PSS>NE}Ty0hhN;A+-BO1%!zu|8Tf2al8>P)`g^O3+%@3HT2INo0A4NEmF z=fH%L2AE^3fZu31t}K7D@a;l_hjs?*9N8Q*npVOEoZ^?g71BY-UyStzx(3$#h14w< z9V}A|6Q*?Tnu;$XFqvoP%QY^BN8kK93(%Mw;aL%ae>F9XcGg%matz^0@BC>p6^*!o z8QQC0c-mlhA(3|}{tGYV0& zuLPxPyZ`KsF~6!58I_Y_;p3KhVb_{elGxOg+fbsbFV^tqdHeZk2bOZ}X1r*W2apr3 zt5h14!pIYvMkKWFq%7nAn$3317|*G7N@D1@tE6j|)A2Mm{z0xKq4FO0ltsNE#BZLN zt?@;6hmvQnFCe7KcxkJ!lcxLmXNQb2>vtq&jl3KJTA<<}h5U8W+UAWsC#5f#ENe5^ zmw8CpwnO^%Hm?4Cgp@u4xP4s$Wo~H!Siv8P* zi)jjToQ)A;3^m1sboLIzVT+t0uIxdZKh2I}^=A9FA2TLN3-6-A+f!wfE0E~e9eZG9 z-la@}B&>w7PN(rSlQwLNLZ4>ZOh|SW;NamuYw<`-+W*9$LH4DrPdcs`QB z&JxG_TbEwcw;?Pr=-Kpc==pvdZk>ZH>6Gg63By5H>4C+0Q1sM7=`z}QfZ=u;QS|}#)g~4 z$VuT{mrGBN%l6`I@y`yiFo6zy_mXl`#jcJC4PgwyC+awl@7Yu+Q@L-V)!n!6&|5!? zvgHTfpS0($G0f)XVf65ll%L!L%#+Jh+7m#H;`R~Mk?St#Z*A`;`+!;dEu6@D>|*6k zqW#b@zPodt-_A}+w~ryNjTe$zOgXo%% z&ye;7Z$4UeV>TX^6@kjvAnDLX?cPs3i-*_`Y3tk0v@TDx%?Vl83-0FPB^RaK(*514 zWENiEC^4dkBgYbQUKxXZ`0*a=>81#WM;4yf3-#P@p4Fk#DFEq~7!G!SK@~HT&r7%F zdf{w6{O_}2Bl_ae1^j~2Vt>&yii}^s5`*MC=)q)8-WC+PIhJ4IT3$C`+t;9ie zl$o zYADk&hFg_fk64?Qyi!P-_x1^B*n3IIUh5h8gls1KQ&Zp~K(FA>>)S}7sb)Z>R@6i7 z(As>XU^Y!{jOYvfzFHso<*ayc7HsE~apun3qr{B;bh=OZFEza}px-TvX77WGe}+~W zrgW836Z6*q!BKJW)XPF`Jt3i^olj_F|F8uzyJg|v1>35RN@DM@YOBS$LJ$^ z6?b#J0IQeL6V44!fERa~_O&l5&KM;0@CuW37fn<%mCzqmQ;vd<3AzkziYl~lkW7bL z0>@&9b5wxhKf;jYp6j<hn1DwDDBu*ykO~%OLX-O zX31aPzcWF8d$?f9S|Zmfp{3J4%v_36r0Li@_xL7^u^&bnn z;JXOZieG(De~yDib&OTS>tK@D(?l(d05c|N+y4(u(gFE_QhM#k4PXJn+orG+*i~*JPs&eo!&Ujr(5o)uw<3)=D^4eUuA=gRgTa{lX?#C<&IGkOHfh#UNvzk!$r>lU zO4138mDu5DBQ0-#qr>u^DSy<5xexkpI#h?_rmr)Cq?X*|30;XuozR*NUAyZQy}xqG z+Ka6#RlTQ9J=p7tki#}8($>y-05vzB0;iq=kHrnZVgysg zIP{&QHcBr&D6eQ1P8%Efvk4YkKcqmNmt-b|03B9&To-21RN;tCRV$Q z3#iyTe?hY>S|3KMe85HE6k00qeUNFq*=KbrU0VEo04x5OGhW@YnUvm&6|hIu%6e== zX5tt}l>lLXc0r$u?P}hCvwuq5anEss*Ap(CfivX{d2QN%?{KLPiokI-qytSQYEBHm zhMf=nC{b)^!xnTU}?XbBGveZ6ziz zEp%ztGV7|7Gi|Da7}wR~&LAPx`=85y=?__c6IUjcw^NG^kJ8`qXKYdsgQT^{ShZ7~ zU|UCtoq}sBN@%0Sp;^aQ_f#H|Jm0su&Akn*tvXRMC>yz8Lw1Io01JM0W0wenYBI#f z7`_3I=ZO&y;h4l@zRTy;4G%l`eaR}H6O!8JF_J%K$V#WT2@-v;7IJ+zZM14FnaCC3 zX#Ev5!vy-&l=k?;%D0sXM5j6VVIW7682h!#=HxLW{RN7k3TQ5Um8OZwf~AfT$@eb% z^kzC=V(=-(O@~%rAk%r0dwANN*EAG^7`j~FAqNdWWK-X4I6RM2yirJ`zG_lD;*ZwX zc*@lU9ar3yeU>lS?DJe-YQT4D*9W0~MvHJ6V8M*SL-DjI7_E^8D2vUn1-TfJT9AGS zmmI!K`DW)Ttt${@2$}Qd=7m8XEn|Hh@B_o&vwv{oZOZWkQr)8LsboezLxha1b9&F* zm1y^Fa>b?dmUKV^HnZaI1Js+e@))BPglSZ1=g(XOm$YbH$YCXZ^;P#7t6eOQ3tMO% zcql61y@R6xGdyZ@o8KIny%S<3rTh1G1s2bCup#Rqh?xLgYK9SO7Iuz} z5)On|w|CB%(2A|e>;uKb_RXou4WS=oH?eJBe5&z3U!!&k*5lw!SX-Op>Od#qvz}BY zYlV(hOvzyliaS)wU_$@uI9}Z`Sz9Ewy-|buh;5I>j=1l~{`4+aeRH9vqK<5QKDT{79sR=JT?Xlv8b8jzG#}Dq zDm^IzE8FGf6O-%DoHZ56w~gNCdS$263h`Z={9K56{VzNNpx0fQl8{Y#GK=dIpIbjO zKWZM1U!;bukG4FfqLSmRLMH;f{GC2=wlZ?sH&oO~q=wlcWWaHeJq+ zYvxh+x%$;fBno3dj^UozDPgiE)Nj|P&NC{QQ^9Kpux55Q^Vd`w9=VmzW-;)iJv|j# zfJWI5O@GiCDp&y|BES3suu(a)YHH*~MQ$Gj@`l0AP0)NfG}?^2h#P{Qy?W zxCtD|f3m&*27&)<1V5OV|2N!Y{uco<{l`Xt1;xe1^y7bY1dm@7b=MaCqjoRs0){p= zNLxYBVMD(h-4W19S~1q?TR|X@Of6JPep+XoN#18?xpx$;=GGPS5>5lYw6ZCixDrb| zoFcl(W#7oyKpzaMielPNQ}Ei_T9&%nT2e*ja-Ftih);5f{COyDE}ek0=Z~MH2T-hT zo@o*IpbmV1*8dE=zPcGq-3{zJHt}~3Vgv;Ln2?Z1flv}^e*f0GtQ1_q05}qadKhkU z#mE2x-keH0I=#)?w?i~RV`&7#{r!Ejr!_(nD;SshA8gGqG&mqli1SUBOuRLSdBI|J zLG0;II3aSw{lO;p(gfKf_$QemTDL;Kftt{8+4 zHBSNg?tn3e@pm2LB=?|h84hU4$a1jgEuJ@bV5(SN9hqTDc1pb&xF31bNPjd=t>wUg zTv-7>*ULuxC|6h?gR^+?yU3mR^lM{(AaVD;{ETl)>CAXGuuX5{X0l$^;cAC49r(1p zWe|QMZvblt7P%0t^uGq(pur_T`;msKsReGPUd#Z7ksp%6*Ozy6X_Wum{q+9MqvZ+4U)S+( zjL?6-ChwE~xLY9|cVu(%Htu<=CS6NKUPwea@MDYqQ{(xXd<1fPa%2k5&}7d99H_0Y zcLZzQBKG;@h^?<%=stZ4tBLO9g6(;&aoUMw@Tu{A&*-fSso>%3CC22)SnojJHDg;!LtzTcjO@zT z;QWn&$}MBtkin~&wJC}!G?&i1Yk!A#UF{q3#%p56YR4AY)aCEz7Wk(wkl- zzt$b&Yp<#@hz*$8vA_VbVd}F8 z0hrl}z@2{7ImNj#;v&U9+H$&^^YKM&6$BcJK@h9b!d(O9!hxe3EiN4T)wZqL2iw=yu)-onmxj7*^?XqqJG`C=|Z19kJ13)NAiu`>D7mji9G_V`!l)e zd&`{x#|SWZDb|@+k99;`()Y5NF(+`Jd1rn5Gs~#!Z;!QfQ;V zd^mbLz(OQj??{58FSglZc`maJA}oiBfP3pQn~u}|dY=1>O4|q&zvTH=C07{>D-j5u1 zrlFxWo>uK4FuSHCaSx}!GZObCPqJjM`EfB4b5^AE-|LqtJ>D)&OC!QEGb3+SbJNLl}tPW1YP%|q=9 zXxV$VVHW38@EU#o==eJ^9j{$%f8@k}vn64+i!`5;RDbXX9(U*m6Mo7++N9Kcdm5ec&J75rNWa;EWljam0&nxrXhlDGkrTi17%^qwl(ph;&U{1wE6Ih3d z&7TqM--`>3-OYqHX{}C-0vV6Zr}>vW&%uF)aFtw1a6=304$njZ0De8@HG%S+Q%f-83|78DfV+L_H$qBwlYvXT=G(h60Th#kCe9OGJ1b_=Fv6-G zl#8}O_HTuhU-}wn72EVp25e7o@EOq2Qmsv4W?ZBl?}k`^I*^NC&_MI>grg0qpjOd1-@WYJr_TfT#0^!({k|Vmv8F7sor6N0Z>=8XJ>N6 zL=hRk#mYV6GvIxsG&gQk8Y`mDJ-g#GZb?&OR-v$esnKde8Er&We@pz3B{0EtGWm3$b+OLHoW*ou7>Q0gQB2j@}`Y%*f=@1f65#Ur#~{y&+E z*!YMm6QZJ$7{xW|!6qZJtB$`S@?sM4g??y%Y8>{I;>(L^o~KpK?SJHI$O-vUAYtkf zbzhIZ=UCA7tS&$foBuchn-cjzS>n{Ae_I^)uJ5h6Lm8*f2k8Ek_SYL}PkT^7g3GpM zB-z{FLVt~H&h$y6CjAqB8prxYGme5~5mVLx82%YqwG#_F$)C=3{@eZZuh)G&wR|`R zPJF@)uWhpDP#S!iS`8m}T2KgARM>W#sw1j1dhZw-AMb^U!X$oJoC zVn!9a%cV7|TL5Gsel(1xFaB#u+}ZoD+D)#$b*{<94B^<3!G_))Vp11@6?#}|ne}q8 zKflAMV#ibS_L2BnkrT+cUr>SXllBHR@kfAk`3UhZ$9-Yb?-X`!^@L8hP1QALIX)Ml ze|H*L`=v}j(Qq9%ARf(ulSJ0*?}U^gQvF(FYUsQHn+Iy+=1^9p71EKE4<=Xz$a;N$ z5D78H-M7+pR&q|dCv6c;D z^t}4^mQACtv!0f2w)XZ_R9Ybf3#3SsqE}A`rE6Tv-PbDZia!^BD38~75!-kc^vf8a zBk7oNwE*MRRToELZ-pos>7p^9%VFXSNWzd|H9EC3RpceRUj(#1)gn%17la&l4@S3s zc6#nn3?#K@?Ic~8d)&|_(Y_Y`EE1Dc;4&J&oA_@8L zrA*PBla$KqUHOVFpV!fPqF_&v8r5Wn&YZl^LI61W*cn?t&KG=^6-Fl2N^9rOU3=X& zisZ)fG7-#NdDONY@H_8nh*6eEYbyZj9hEQRy16h_7^Cal zCP2jIS7YrzP00{1IF0$%FR?RlH5Z6V2v##}HqwpDUyDkS3-oLvkLcOH&`2XK&`0ho zq~YrsKL9^Kz`qUZ=RBXk)9_d1vVC9wb#gP#xcuP9U5PuOk9He4apF{2 zOr?^v#PX>#V@6a#Kihmk`-gC&hdyzXRm(%#j|+PZf1$_oz`S+J(dHMLnQnFAIQ-zW z2_`ILV^(=6rh50raVz7n&`s&e^4H^gaziMT8}etSzK&A3TeR2w3LohUl7G&?hnBEw`=+R?tura4xR){zEkTD zR`8kPe+Q;SpLuMEm>$k-e^vLQ`q=uQqsJkY)-jhqtf%u+r_rz}>kGi|*x%26pMOBG z)$12~UJlsi%TL<*l1c5uIOiPX8H&Bg2@c#F6E99A0ir)iL`+T8vahC~6c*(XB|Y@B z4jxI;+XbCCC9*H#)Rxj=xwvS{5xS~E z%9G?kFQT0Wjg=G*e4OV1KY)rESGGgT>bHyl+U~!0eX2C!9c|wHt*Bb+6w(Gq$s$Lv ze|e+w~VOA`~SVLg!!xE5=0sax40cL6DpVy3CqaOUbvD zee>Ttwa$k)U|rOuX_8j%QdDfCCWZLo0@|^tx6{&-_g#=~^K(I0h(g*a9AQ^@gah+b zJ7g~UZk-zJS0j$OExliXHru-KCd6#NfBi37?4+$DSbZH|+R6=;*>HR|jCp<*j`)R- zjoZ{J@7!^Keis>@9)3pqNwg~{cz})_4S$_h=$0`b_xU;+?YewqYi)w(L4)&Ci{Tu~ z31ZI-?ePr|M$m3_=W#TM@9m|Tbv_!+qzM99yomP&*LO_P=riPES0&56>0#5pe|?_n zdT_VQUARNNrz;Hkx@%*;b`!B~7My(BHk3~i2`juWbMI4}ZH5W)m0UN| z=7(#E8;v}3j~e=MK1tbXtWaXwf0yHb%A9vbbZKHj>@P=;osq%jxLIy)j;<0E6N1)b z1Pw)oV1I-qkGce6jtrxAk3K|%t1zI`!puuFt}Q9d437XcUOq+cd_YM`M72F*!>6sM z0%AU52h%Uthu(Kuyyj}O|-%>%( zK;$q8kb}L@?&_p6c%}Xx@QRx5iiU>$QB+OIN}5yvHk8v?Qx~R1=HNQ@3`n%QF;;bA zE!o<0dVa3K6cPJOgJ_V|gzVC|^QLORQWO?7A?Jc|?!C~3YVon|iELr3`2Jkn{DE|P zOVB4OzEoui0dkM;`7(i~e+&LoWS?J&6neC?8?m*2+Gv7JxiX>!(H`>!UY#J!BI@xr{9-K~b3*f+EfZ~mB^yPaE2#V#QWV=9j0 zRlio!W~FRQC+~L7ybH46eTr3Ee?z>;gQFoH;STC!xrvEd9#IUUf6{Q~dh!bKoQm@B zfKE_Z#~un$YRZNqB68J6m4=Z--GqgsQWA5v+7=-8TH-WqoCR&&Og5+T!6A4?d}@dL z^g0YFy)R}<>V&AMyoh9Au`_RLP_!O3JJcSnG(@A%HS_5=z~Lgc#M6*Gw$;Mb9A}(I(p}}Tyd`)Hs{L$ z=#tmQ6At>GWB<<1hyy+$QG0)fM|I}Km1ZjKx0v~n!j0TD^%w>b_25u`_j1J7`_g&* zZ)Y`SUc>Ek#oIVjCtkpXdpO7qn|M<}uc3=u#Mhiw1f4Iqa9R28zs0dS0+ZfU- zQU;mK)GlyZZ(1)iqtAzvu0))4CIgVLc)ZVwyCaa%^;GlKzN<~-S#)sg2#eg0VS4tVj6^Oi|N1VkrijehD=D~18!<>qzQh9#^ch*6 zNSL?yo&;m4Shc7E(TsV!&PR}2l2r_ynXsm(A^O}E18pWCVu(9hzfpC8u%j8muaL7j z$?4_t3ZYGzO?*Z|J2>euIH#veylk6e{RRLNTF#a$3`e#TBZ^JS?o+? zSpD7b=|txOjJW`7UHXDjdJ4`wB0=BZ5Ce%UbO(Hq^D8~C#Plo%n|eR&c=PZNvHFV` zrIt@&bqknPa}UT@vtSm#1wRX1h-}xf$#V=EbU}756x%}>MDuK8v4?hFl&9|A6$yAY zd7PRZe>|%f*JU2Kfc1AwLbw>ROGo&&oNipu08@B*1&T0rwdWz;ukv~r2Bq8`HNB> z|7_-P`NCm}-Xp9n5wug|2Kt%SXG=GiGdUQ2f2s;PiXi6s9bNq_gJ$)ct-g2la;r9U zE`rNX(l}oIkTefQw8v3)1gd&yF7~ZfpA3#R+8pN)kW(G|qHM1`yLV4n23{NR+?8}uqby(1n3>;tu)T{&I*w;ItOgLFZ?craj6 zf4=Nq6n#^xaZ%sXCDb1`6}{f@(hMesg`suAEhvQ6onX;|L{A?+hI*+rfm@xdr-XXm z219Eo!Rz+tNC{`zLC#dj51qXoE?vUdD%({DqBp{vua!B7*fV(g_Y!d|;!HRfgOBnJ zcqnExWZ@nM%V`_4kv}59#X!2cxYIN~e=h^lxIYhI*4FQh7alF`q6oDOu7+VgFi39N zUwi7COB=oz@GGKtN^8VJGjEWS(GoAbxirI^GnI}7ly;L&WmvgKcw{HyfAID{LGl#4 z*+3FA@rj!5#H&O^Yidj5|Gt;RjcL=9kXW#v^(J?!#(LF}x-V%{h#Q!0L}wo}E@+mPXOu)nCazx*==5GN?;kOuYD*Ip$FO zSmp|M)xGEGf;#P`?@-Y)T%1T&e;AX@B-}e5KT|!=yl%Vx$##P(^AL*wBSO_LR?Gb( zY5q<8{gb`^AP>C^r@M~6v>WDnZ3n*w*-$FUP4M;?MSQ%Tb!d#13y!Odu&E8Y4VeC0 zL2uoxpBh}TAbNSqi4DP*ng>8ORa>3gNHDsqZ@Hs$O;C3UgUW7=WVNtrf2Wk4aVB)v z!@AWr!9ljLCz9v$sY$TMw7CYZ{MT45bd~!85!5h^bw;i#;M(y@ODnw$PTgRV%LS>H zIllYgS(BZp;&^+2{Lal@3*?LLg3gTmn$l%a!M&$!g-QY~QMWQUBpVoofc8+kCAAp5 z5^)XGG(Jk^8Od86j;Q@7f0&stzV5hKZoqkgO_Bws{SSLb;?or9@Nxvpk6h<<2G%Xd zt3)1O=7daygKA|e;*~-2T;Z$@RZVyg};-vVf+@CK3xtA>bT31EI3GJYB_9~dj_;sA?c@mWXw%MymKul zktA}%gY-FrKiM;6`0(fxSBvE^;I@BU0M{HLqf-nj)SYh{e;P@#NUYwZGpZMUUi%80 zRj3>mDTXK((w2R7O)I}3A-;;%yFPnPpr$qn1|o1#GKqjM(@Td{7i3;}x_Jt49-bMv z9q1;Bx@&YP7Zxk3Q#O~xExt=1gI{Md{*_E4M!LKOE^g9u3r94b0V7fBL9k*?`Kx(G z=hr~*xt!KZe?03|M)YR#VJ@?djf+TU4im3G=hbg_fj$)g7$T#o8l=C^Yg&*m8xmNx ztA=6hxM@j$UnM>{B`vIU&ozGq-^sa{opc(0V*|T*TSXpWuU*0$$`s8L^t`xt|LS-A zKsEX*bRiV!xf97w)BaUOacr@R%|!?+l4IwDf0>KDkj5mu+Hi=eYF!K-QLe;9 zIb8)y!el#%dnpI)?D?F#y&OaSHiCWE54g-OhE!dM?X^{UdqF8VSwfj_op>1uujXa$ zgeLkzK1OfPMCs!rlFo-x_l@GK^I6{eUeh@^?E((53xs?Z{+zfoAX5VZhc6u@>rr>R z4cOL-e@M4RCc6$F-k=7pBZDv5wSIJ~;zB(M-MmWVpqPZ3qm1a$dKURe#2C*l!01H6Nc5qP}WESa0$x#X&AdfAmmuh1;aDc{G+zIeH3tJS75${!L8}El`dM)DysZM{ z6=^r8F-d;!{Tq2X%;^`wUav_WjKYtAS^hLee^+tO39b};S~ncg^j|p=w8un{_35hM zf7|E_;&%fOq~EN(QLhR%(w9VzCe`?$X}U4DNQrZ<)8?b zoTzZ>-N#>*WCtHdZ^O`MUvS82ou!3Xe;HPLY;(-ZWeOgIbbfjf+%O%7-gZiD>bx8l zvvF$8^m|q znv5^JvWh;cCXLb!n`nx4OVUQiT7@ZY)|n?byl$15TQD;8KRrgAim=B{Bkz_!f66Q5 zN+tZx@{kc{+g}ViC#lJk70>wa0a0pt_@eG%s)N|a=4fmK^Ku~vLcm)TBU)PLuzEDt z^tb(a!**h_OP`eG1$dA@#$~H}RcKBn)&fK1##K<}P?!;Zw)v;^%O58Un8TtQM~L_! zbKKVBYrcW)t$;|FgatbWp5`v+e*8SfzV;-JzKe#|)G~=vN0H~$2V5O8rh|!v-Wb&aXA&tHIot7Ox5g3R zn+u{Gu~ANP4xv>d)dvz&Tc8K$sxWNesEeHvj&Z-~Hl2iKnV5Gaq6>bwBi)D(%bSDLoH6&OOvobYVx?7B%tM*T4-$(}xK5ij8nc zBV+%(C~k`z=wE*rY$DWT3}6aAC(FPmP?QWm3#~xmUar6!ywZtb-LT zFEra*sV6?dGATq@vd(9xZif@rpC`?zM%fu#ue~FR31w9Suf7c$mG-+cl#F0>~ zg0y2vn4Pw!@TlrYxOuurjt<6*aZOWe=D{3kx0z^lsxYZfb0#3aoINJxPpunmEa2~r z@9MX*IZzBWuiB7J+HHP#%Fah7mr>ZNJK@{C(mW;AG4L$cu)Op_5fpJvc6Ws3TSKp7 z+99lo8#YbJw*2D7e`IPgVV&!*XHuD2N^U2_*}dzTeslBd#y%t^JUa(1z{C66hQusa zsm%Ae`4I-{l3-{T>sc}ZRIil26PM+Tz&*qKtF_|bcg(&R)U@%b`T(ksJwxf&9Y}XL zx9p_f!sfKAm^aqkn^Ps4v^FxFYr>J*1(f2vKLsUh0-{q6e?Q;0;m@{cd)tVh(3Ts_ zIOf($z=aFagbFHM&Xfm|f*@ws+!{>)dpFuK%mkEkf)r9M?f5A8#PQgq1)`JG4S})B zLD7lzG$gKTWD7YZ?ZUZKFF$(=>flb>ND>a`Eb<&VZA$ZKXNyU43T~{ zY+%e0*Q3dJ35(8G;a0G=Q^s2i7T^X)P;uIiVexh`}{EM%+B8~!4w{i3kEoJ9yd zbK;2d``D3yzq<=JmD7a^XFWvC*#vl<& z#+lc`hKOG#6&>}CCSN8+?!>gGwzn6LW~+OfvLlW_f2`IA&|C-JmNZLc2fPx^!yVYc zl=kCusq3yx@Ig2efeelWrRz}L)uD_pz1iJv&Dul+Q)ii;c|lh|($UY(XgH_*3;DCd zzIeD>f5Vx+nw%7iCNZXD7>Hqiqn8OiD@ghr60@@qxNakrhg1Ed`pG`vqxL<~4t zEE~mo$)R7RIwFgeMuXceg_Q(dxCyfkqr`Umu}f9n`X^5KwV0RX7P>XP&@_&MTir6gP)nkEBn?3>PL*b`mlSa0yx4CCGG(e&B=xc^N=^F-mmYMehS&F zv2~=}(-o<_#WK;i4|GG@s3+w3A@dDA&P%UT1nEb1n+b(Zd@;)qM?M}Uj32JQ^TY<+qh-_ z(MF@`4xA|V!)A6bJD11)fYay+6YpjIsa+L8yA@nG0iu>Hy@<>L&}KH>LCW)2=@fik z>wUECOJzTz%vSAr|MV!N)^jo`e>7IUWV1fEV^p-qvAu4VpmsuHk=dj)U1|8}Ny7aC zq9n;;y*9tKO->zfJw$Awqo!I>aAzp$Nh>99m4Y9`z5X&K0rJPXzyEy}Lw*1f*z`Hg zC6;laS-{4*5ArW{m7D zNW`cn7T!I)N#Ckwe`i7ji@Wc43fJH}lL&&I%bqSB8`&o82=#BV z=vw)Z(x~&g0zKrk59$}wNau6hprs<$@~WM6pX3zfb3B2nj`XXinn`|-P>8)@Uu98` zjTI*p{jwjcP|vVb`+^R=f3sTDL`AG+@rGw1ZU?q^zt~am_=L+4z*G%2EY5$dKuldI zNpl_kyjgN>e?;3u=Ta~fZmIIoa{#Nq?L4mwh8X5v`yQWm!-*O#<~e+#HM*O#7e+Qd z6@p2fXLae|dxpXAA<^wB1*%QEX74j+qM@RNaf5o5=S)avv^3?A) zeT>mZjc(>6xq8toJM;>u}3~u`pes}L|E;SRVq!iPZ_Pb z)%OZe`{FYzHwY&t1m98(H2j8HY#A87iLMhjp8iXsgJ{tuknw}Fn?KRRRfJC|$>QD# z0X_k`6KBRD4t+;{e+7F!pAseM7!amQZH8uPGH(kp28qX+|a$krAow1-vlm{^{1)I(dA6CggOrQi(G(bc zlHKDPI1uK1>=B^oaPLBwKlnoiCnKQlXI1O#E~Ap|*1ewb-=! zONaA>)bJwQcq`4~1q|&-FJ#0h$@;!*{}po2py8d$O=A z;cG=oBrICeo+2A3My}K6q)H4`1BAQhJ0f~aS%RCzy*jz0mKqU%*-peX!IgQ?VHR$M z1>GfZ$Qqh~>49@vH{p(n$J`k^V{gL`ot9(FDIMrX#9Cx^te;9no{VBd9J4lB;bwih ze>oI3Es+M{YbK8kxYTc4;1{G{a#Jg4eio7E5mVGMZ$oA#Y3X7br$`b>?4nVpE?f8l zr9?la;GW#7W+?@8PzI6ui)&EHpP%jZ%hrzo9$fg<;?_!^&^YK9ZuP)Tuaqlr>jbG+ zEE9}x2s}VqJZ zk&MM` z_%lk@prcV1ozD#s_h=&TCO{jUR+CYkGKnECn2_hHR{H*#*39uE9z+jOAX3ev7JQ{=dMzQx_8%KnYRrLt=ho@nqZp%NT#9Eq--CYE5gb7tn~DBkGvV}t9uvCf8+DRp*5_|TCMisa%bBdD^V#2|yNYu@jk1%^NVN-}9USDMG zot?ktRAZ8`&*(}LYa4%OYpJoO?+8+s!e|^cXmxd8+%`3L_G#9NxHVke zYziqlCoToyN___tU*xgi7i;M?2va>YJE+YLkZ+*)bVhH(M1<$_| z-h_>Bu^5oGVK(e=qPt{$UR@LuZwPjey87~1z?U3aXvXvN&SaxLQ9o9+a*_u878PR| zW^vcHbHIiNh6<(=63K$Ai59C^L#-J-efHe^WJkYari0o`w8MYSN*4M784T z^n`fl&hD*~Ke}E4Nhia!R@W!t^c-8dqV{dmqD>UASrl1bdQE`M{;banW7D94|IbE7s$-{mCR-nhV9C# zO!`R0f5Y3NdH&Tp8~Q!om01OD~yYnMeufbGU$*Z|0UkbF{dhLpo^RcPPT3Ys{bAd~MV9u-@)FHk zxM#~ofb0=HR$(e>f9lrJg$?B-#G)<2%FV7pnBe{p8V&l$%5rzVk&i2H-Vp+QB%U%# ze}?uWax9sITV|cl1ZVSO$+=!)b>}>4sxcdMX{7NZHhG+m17_?lOPt=FF(ai+sdPPW z1vDv(=$n3E50xEKwCZM(HNrF@6tx-ehRB@6R6{ z>hHj-QS>$tJ)Vu}O&n7ZW5DzmebdFae;z(742HC@YAC&I{QAe+#JSDElux0vdgR2;c+x=oDjGq&W!Ck@ zR77%^k8>bv{PNrTmy6Lb-?2*nJMv9^7dD6u;h`B=wRXm@Qe;j#Aq{0fZh|*?f)zy< zrqECNKC}t6V=dISn#(}64dOkTe^{Mut_OM=&M4AAj^&DFcPBXrqvvU=!0{qU&(L@- z=CE&1xh;e&qC&Q*%=Xq8+QDvgxklK}TcR_l!f)|k7ovPf6lGFZ;516xe$tmD6lWxOA5#t$1{%S52#$76GO&FGpF#t z9463XWXm?KjkTR4{AzW^sSh_bxf5VB^bnhr#T++Q9z${@W622cv=^7NnJCaa#AlI8 zMefrbvG%SK2%u?feJg0*zd*`q^p>}SP}l{mtB@g%z0ZHQ2t1Z(!1Npn=0Q` zXDf?h=HIoRpnt}xyX2Ukz(F0hHuiFyzmaR^hANa#j{-{Fmfos=Cj7YRtw zlV!)_V%wKQKbSoe-15a7=6dBuh+EqM*z^k?0e-I_CoJ;kf1}FLSA*G==H`)S>GHV~ zyl1$W1(94Q4O4PM_FVkCDhOA|5v+u?NEKI{m){5@J`Qy_!p7|SA|W2@EwqbX@(SZx-_;w`q_JjmA)Qy5oKle>3n;9e~S2&}A$hA>WRErn<7de9#Xk#RC0r@p<0!KQu%2cW@ zuqQx1B|BZQhxzM5c5(p9+c$E@d#8}voR0OZTxj%2e-?bGgUE*;7=bHuSINHjRU^nw z2z=MuuzZ`8(OZ7sEFplEQNW-mXQa%v!0vg{F8L-B+^N$J(uORo`sx{@->J1?zUHyz zFyL5l=67aqj=wN%AurKguAk!x4TnGAW+AIz8B^UAzhsg^+(JGmOr@9!q_s0`36o<> zp5^Cwe_x>ud%*u_w+dEo;o46oEzABW*C4xD^kH+duF~j=p%V3qe#APxlfYWKdPcIN!L2TE^3p^J+l!)c$$q;1W0cLOi{K7VoO+{=4? zgS|z`-LV=$x(9%?75;^5lO(;u;UPHEACp{Ve>5a`^hHXtDpCCsEGlQd&bp6ju-e^Z zhe>TE79ZmWwYMli_HEsA7d?1&8O^4^k3_`|7}+tgcEJ^7TcwmQ;JS#khiSLx27CDF zv=dCNheUzZVDpEazTY_&+X~Ui%HY-=qN8VXG?(exa|>w`u?CwmBEK&4EPcX>MbN@7 zf0?o=kG^_w)F0_@-Fdk z9nZ~{nBv|>WIz__sifxQ(IKRENTZ_J&NrOs^4yWaY_5X1h&8;=Um;>>imf{4DuGSf zM7r>vBlz8NPC`^EXc7aPZ}^4bYAbyUe_65xN;B4Xb$~WA16m})o;w_sLFvH>L-WC9 z%VIDg(a*16mNnA)_W0Mc_Y9A%Wr*{rcX_--X?~cq%Uq}6|ng<;OFqCX!j(uV_AJaL!XmNESbw$2XzD`fAf7G~Y z*v$Xb$pJlRVbH%Vh>_9eP*ymm=v90vy1MT(*G0NCKut@%T;`pKCcr~@rz#qLs@GZ^ zoG-l{>!bVXwl%RfG@9(iHHBnYwmC|CX4_ciiL|uoo;=&ysmCfc49USXZvG=GRY*EG zOIZX0X$PEmN-31krxX#Jox(fXf2*^N7!Z8phBRvmM${HL4ePP4J z3}9~;nSjz7S%ovMiy~5P)!mA_F=Tb{&9q$!>{PGmH zR*8exg}OxKzJircN^UI}{DORxq_bbCdpf+}LaN5lRg1Px{aq195f5K_8V@DmQIQ%y$ zEB;q(I=Z4=IjUo%8K~=HRCOY=?i4&3uVh9dN|4;$T_+Ue#(WVgcv0WKpR2G@5&5^k zS!b6As@}6bx%CY^@41+`eZph-$rK-jmR{*5Dr=NWca;7e_=CC}+X>IvaXk=^RLqTde@kXqe5H@y*d@m$NPg%+c-DO!uNj{qhzCi_}87+c)q$ zuJc!J1%q)2=7je|Zh$xs&Iw{%sNtHQ3f- zR#_#AB6_op-{CZ@&lm0ICkb?|8|UXle_{llVz507#%g@b-qod_b5I>N#A-osyZYzm z(x)ud?g*{CD{1>UU0CYFHhD7Z>UFRlvne?yhr zhBqDa3k1&So^|X4b6+B--l;9F1R501-S4kxbx-&Xv)^<=Q|Xqb>w%Jt1WeXsWN)vQ zcZ05JAr%R{j6IjW{Q9ka+gQ5D`KGQ9D8nimJ{6Zb9K{6SWxpfBty~3HV;e|X$PXbH$Gf@0T#7g72clKQP5n=WZ@XuT}hGLS=L_m@%Yq#`G8#i z)w-R`>B`M)!)x}E^&3x6PWBTF!?(F4;icv)R(S@dWA=Q+lqSyarF-hab&O9{#ZnDT zqcBVj>JTf0_Lae|ik%}He*=yK%dp)eQ%0QqHthMGtjDBBn~vN1Q_$W(A(H0S3PgHU zyKj8e(o!6aRXA=>dXA&qvJE!TZv_#RjWC=Cpcz;E(IY5MQzqQa_NoQN{0Dbku+MNm zQGPVp{lK-105_B#F;~q`=*pMyv(m&J1MjVWlf4%U*Jkl|Zko+8j zc_IPS2ypq-^s#Ju?q*Yy6BIT>F?xY|>&-42H++E<_=UK3hcScwn0Zj*v%?7t`1V4m zGvf~Ak-XjJ0s?VUG;zf!UeMljt$|fHCN85OT65#Zp z=%qgr#VBK-Ci6w=C9cl)&UedOU8oa@2?7ros|_;3kfT1K^!MJDSvaiy>i)I8?@ zgseSs1#E>*Lzqz&=zg($fnvWHjDXi3V9=)0c(oR;y1n>V2gjXbugf3c_!B;{u&OGqK998E{@Qf)rr z^*kz)R207+*`;uF;dN+e-(`NogTae1&f-__T1w2c`3N?$3O&hngey?4Zz|IJY)2es zU$j%?pz#dTG(FVCz5=K-pz}4+Ct|gXUZ^UXz7a`d1jtSX-vxx^V5%)@!^sz&b_AD$o3(hnl({G; zf1DqP^*s@=%nV+AjgIv&0I^uPKlZy$w|;mC6Jj6_@QeM#Sq)5OWkh5fss|{8HVQFr z$bDp=a~ia=qfROa{)?`d+f`OD*^vfSUVi6K%_~TJE*&wp-wUB2KKsllAAB;PbYm;tU~1%ofrDDL3aTkta#%pmD+b=w7qjO9k@r1-jq z&XG>e!mHt=^5dh2dbav19#{mi`rGMND{=tn8eKKOj*V}_ewa4eFuBOthC5=xf8glc zd$%AicxR+2F*tKWFn1x;Pp?feb!EmKVd_CSRp-lB)-O02JV_Ze@+wsTVOZFunO%vZ z)smvmk7;{w*@w@kbH%><@Nfj_TjeZ)qh|jwmL6c{6lUw5Z!u!KO6H-{$+CBEATxq{ z4C8@k@HwR7l}y8E#1#!y(S{X*e^Y9i2y$O^eqSsl>~}1?Vhh0vN*?4a{XA0e-Kmhf@OdKL2aL=+Kq~r>*~k~fJ|uQ0@s4|aAiy~DUWb_ra z6)i>lhD-T|TkGEL*b^t5e;QU*ri9rH8UXZzidwd9b6g5i6HX$B>1W6rRCoZFzeF#2 z{zg~=V`o;DhdvOZ@Rv$Vz^4_~*{`PLivTf2O&U3;O5#T>Z;Rg}Kc;_aj&v$hV2dF- zjox_=4zVPvlH&z)_U~@tIz%-yClFpYsLhd{zAQWA_n~MI9~71Pe@;Z>Y)O4tDob>w zzoeoSK<)?IoyF!9ZOs+aau{k=yUJ;ahWKf!M+Br_$#8;GZZyxKU>oUk$@(d(lT@h4 zK$wK^QqkK&_*IAtbuB)y^pavK6!(L;NNsWR1Am{K)gO*8hD<$HkM7Mu^XloIEw`?| zD4pYaR!PZ#J|p}-{PDI$M)6wO6B`P}dj`-B*4k@;E%G5}Pgh1SU3mGv{f0O9-W;l` z>1K-r3EMaNRiCjcEWHV@yCJ}x*te+p>Xs-k@VI4pYEte9e|QwT0sp^PLt_p}wx~B6 zkv|IDE$ND3cGR}rL!T08Y=mATdDOHmaR4VI1~STGD_DGt9is2J4PrVa&XFheFM&0# z%4xlVbMYNo>HrFg?rMG4EyoYNMqK|LZt6kZ4^y4*!Z;YM$dxc_{O*&-I(!ttS})@` zF(Ni08((+Xf35tcwZ$Z!Y(EJhUdo;02PV`OtH};Q=}Ts{&K^P&&I1o$@U32tYX}|- znmnGdr7oq8E=(z!g1^Anp%w^1ni5Q`pDIoCNx-OM?wGRww*=p&`QW#SHnrNNfyIT$ z%gg}jT|bpj{&dJU+U{3xwI%y`);nJT(s6{tt!9x^e^Fh?3|?d^zw#!4iuTLV^n{kW zFu?^0!;F}?4}$?HP{JSUJk0;XsawQ7h276_Om+9h1tUK0o~^fG#c6F890PhFhLFDM$sbOn zc;wAQfB?%u@#~WFvLqi(_DDXmnN-Tf?m?t8cz`*XhOk;gp7(tW4qPtpMjK!Ei7m@f6lF5N+79@HH3K#1DlbMCe_lBBFT^DT1KTRN`@ z1nHRLS1~iBwvbqzt3OAIOR%5|$+iO22Zn%zq(k{v=s@q~-2J7Z6~@_lPnHI-Ze_&s ze`zQAEkcojHH$4JAm^@ zU(44gNnc#D!1TDPTq-?xIq?a8z@WEY$r|v@=dym%aG&#V(lHP=BZupFKd1}s#9^$X znI9V-_GwJt?i;n{bJS`UA-ef(=((7be?$P?97<57&gxvwdf3|Ieyl?J!%hnHy^~A{ z-)7++?f)?{_*3!0qV)tcG1lE2k=5nYPeQL%dEnV8)NOoQcm$=w`T5(}W_mEEx|Exb zrf+F}C851d4N+B8(&*oRp|VmalUaT7yD$IZb`AxCNC4=Nb^ADevsX9S(@YK!e>u!u zG3F|L!<5_tsC*(kt==M>Z%e1E-Iq0SApWqSq)HAyxf3}&kTuvRa{7*3n{$AQEn;J9 zg$HalDyjBPt!@0`7&L+w8&-AdN&AA=0qvML&@P*nm%C(XLCP|CIcwl!s6$-#vACKu zDR{>t1whUmLA(0G)+|S5FfyI&e`b$|-~1K*MRrFXKZp9S9E<_ec^RtP;sjxT@OUl| zQI(Cxf2}Xv%;PiDxEM$z77JW#A>V14{5YmnKFueyDC&e%MvRvN55WoCOXI8b2Z%R0 zpLCjFli^Nl=AyE1V}i+}Jn4Z9de?AUEwE-1b z9$;g2`MQFEU{9Qs-t_s%h%3O54LsUX)jU6JE_e=pXIrr9!PM_tEu+&5&7A(?s+GW0 zl)4K)$1CZdRx9d=YFkW^+;`__`t%xiUOr{Wonnip zginCLQ7_Go%=uU65)tfYf=5m`aKOxTwJ&&K!&B>Ing`EEOTGS)U+NXw%jme60uxV8 zw!a0M8+9M0VN7j#^Ju@=FfM>HEJ0b#hHtVP!=xNwzGF;X?Xn;TfA>o(z;eD@W&9R= znIH3KrfJ;OeGDJNcd*v*KUQmL8r>3dK_2R85CM9w5E6!Ak{kwI zY*!7t#cWG#$E3U>h4136dz)(gO}l`WRIBDL-fmGt`o0ccBqwu_$7252_Xv z^Z<-Lhl_s11`=;ze@tvX<1u19bA+*~#GJR}GsF@3m-K?f471A+Ez;2B1T1lZio`M& ziBaSCCI$18P@}0%TQL2ij4g)k<{UNgFMe(^-I!+Y3q-_gsOzBfp)ArVZL4n=$V*;J zG~8G`oq!C^9;GRbfaqd55VU05pRDdB5fB3%5Y6tA=Kz>MXTRyV(tjM;Y!Tyru?+vzew^?*Z0!6rfX`n27CuL5J<>piss^OFIageGZ8ewB*ufSXpTZOS}h&FE+wSnkTN$5Krs=91JexM>Tz z{q;Y}w|{}-#19R9D_dX%xP3e@;VcSeZe(+Ga%Ev{3T19&Z(?c+I5szzk?IE%6EZb3 zFbXeBWo~D5Xfhx%H#s;tm%*_J6$CjnH8GPR4kv$%_61O!+tLOKSZ1P|^`aCe8` z4DRl3L4v!x1$TFXy95sqTml5wJA0pfc5=S|ue!IUYG$6+r+am;_oXCLQe_Y}wKoPz z*xR`y={l{4DKL+%G zf0_eeVPyF?+&|ub1+ujJ+u6v(#NO7y$j;N!&KzK7X#)f(O2{#~c(~95jO@ydO0HBA71=D|D3Di6tfPX7l{t|;~@bz}EcL11y$^iOWngKyS z2;R;{Za{#GlPl2I`%lIHA_Nu|fT^X43&0p?ZfS?`J32@VH2V(*oxhW%2SAq@G=3}q z=D$Aw{?h|Zn5n&;jpuLkzmJ$nM_pN4N&NjkCI71w5wZ6Gcr&mA7+BfZ033g8TmViU zPJr+KMo}`d{D;LKp3-(^_5hy$kOg(=zl7cXnE;x9)`J%Czp)hTL30ZP(EJ|rkIWp* zCZI2t|Ifqzx6A*(8UHKF|Ld6lUq%wHHa36jY5rsI|D!juwY2g4#{x9Bt}dV*khceI zg5Cc$)dc=$api%gmaew{S1W(*Vg%X+VLNl1e;aA(EMe&ZG*z;6F|qi2HvS`4|7*`| zEbV|w_Rf}n-7EkG7G~!EqXR9Mi8bi0e96%E1XRa&j{A zL|_I@5-SG>z?%iMjix}4zlRvW#As*l0&)R>s`CYy**hWpb)=je049IozeN8*oB$?~ z--rvqB>Ee11DM2qBOU;g_`isY8Nej*8?gYGB!44z0F%sb1j-=y8-X&&|3;t;3cnF3 zgW|u48`;9==s{ckHE*ifPh>PZL1j?@cFX92| zjQ&L|tRP1tXA?`%x!Hf10)O`el*`EFw-cxwr@NU&H|lW@2vx8k7G@*xCP**xLTV3p77Wrhh{pA4LT=^b4`OZVFa z)Jn78E+B4ZmTrH<_)Bf?>hy;L$i(~)2qI0kK)ZiGEI=~LKOl&e)gKVl zXzM>9sOmr2fm&kw8-gx5rr(h{K(2P6o%ur#s?Yv68IZI6|ENL5IQ&+Ff;oU5hIapR zI_xa}NdNb2*g>ovK)09u?{2e$8sT8$>ikE1ki_wKv;RV_KPoE|JV{xP>(+#D68im5Y&CI z-<&~_yns&sVemf}oQbOwsQoT~KZ8KG-+%GnuM|L_2he{6VP(nQgfGOZKBV)$R+!M8 zVQ)%co^nq!jh4ZC#i`5n5d}7xwk9iR%js4)d9)vGeOH|3=EExKi}y)eBW!qkqH@QR z&$B_i%IsbT!nYZm;pw6i;f5h%cp?V15BokZjy@WJ){t%BJu;Mej;`D&N(Gqr?t_vZ z4MP=c6JdW#d&>JYoU(||6>|(R>d_wqOBX2fjk6c=$YERa;M)& zW#Zo>_@2hF6?^M^XZ`%J=(X`hjn%mqp8{V8j~Ma>V`7%dTjVfB26x%JtR#9%U^$1G zmJw`gVu;jdyocfhm7bYXAv8YfshCfv(=+1Z5fTy_vha zgk*tYYssMg{zy3dU4+wnQ+c;oSOrh-{p}T~8|{-D*^{&SXwWHCpCYTwa*8_d0jud#hk1{E57n;WO7U}76a?6!v3?_K)y2JRQbh@E`kVbw}}6CDhj#9WVD^+0XJPbkewGpS;}NiJClZlBP?fIgBw{ zNUcob&vdZAK%k^Y(kF7|7A=id$ky@RzxyDwW?KDWwnO3+d{RFA;I7Z0BLB%0{hj&y z`o|N2mP5glNT>y!>~JIh)|OOFqDX}stVn+;4sE8@U4EVV;iAK${P3Yfa2!JM$zt8{ zDXP2}X~cf^!^OK>;&!b~kH4^3 z=aqt5d*4uFzZ47HRP5+Z5Dg@*=+(aSOPkP{@6-9_Ip#4^6I~x(_cc(HN=!F?{s@08 zlM170uYHk^oQJ3*RLS?jw>y5TXJFRoX32`t;AJW%Z*^RXv}wa`KSuxfe%L*BbY8w? zp3cLzbrEM@yE+wXLhy7QnbBzv^GlqAnK!douMg*k%7fgaY(BkXrkZXGpXnLK`%)EQ zsLjz@*#$`3ESp>Xbt=vE0NkjJuik$oEd>`Zwi1H`8$T>eWc{RZ z!iu>*;HYs4uggCvDxpL&@_pWji5>{?0MS;+CV z!(lB>M7pzNJ3mqtfb(3Q$K{~h6~oSK0xp%lu*iPls(Xs;uH7||l998Q&Dwvd>oPMp zI+n6p7~Pm9WO1Dy*=n4%{Ytk&3_%fqU^-lE7)b$83PlRd{KOlGMjn)RSx)=8+(&)( z=WzK+v>7hULnWqIQ<2L(+Jb3ne`%h;m&`1V$zPrR?+KfW&aJrmC{HXuyrtsoV2+^E z;)BmdjH@w$%{WF7sISjR9Myl_#kH{{xrk2(&9!V4d_bNp2*72*A1JJs&{q*kVaQI- zzUnV=#0i1b&mwz@hM;3HcfUJe?agOrD#^=jW$XD+x%07>%WKWZ>v)!h=!o`H3*-0n zi(i3J;vPZT)a@e`cgGwN5+?ovc2?c2J6($~OD0%{aqcWgG0*oChircV#<|jZ;aQP= zo~`w~pC^7Gn=R#+uOCMx?j2v2Je8T5pHlWXA=NeM#;lsscn7%k1Z4+T+hK^`V0MQqOfta$yipR^F(S66$$NX*8+# z=^?|xUPXo^E0FLRoDwF~^23rBakMQ1pio0Z=&iFM%MPf}{HK2^#mO^WFFRJWzdx26 z{uD%?)i)#oYotl8nx}@bO@Aatl%+a8i?jjgJ8i<;t0$Wvqlz;3i4}!vb%)7P!>X$5oeX)N zI+z~@STg2w#_oS)JE=$$oz8=SV`_6rU{zwhkj6oyyqb<=dj6Kf`R*NwhHr;OUiP_Vnx z8I-(&hIMqEBx(6twp7&g0bD+jXKr?#B#VFZ$e-> zulKb_4Um5YJEcka=`Bn!<1D{RY$yakN6Kj$_r*u^!MMu80K+KLpIa@Zd0J)Zam$ql z*LwaYY!yeQ0re}`nnU1tm?~2{|J0IX#r;>O z19ikWC|FuX7il63$N1&f2ogGG$UMjMfcCn88o`4Q1?;ct@I!3Fq$9E4xljiKcobcl zb=iDkE;yW+A%C?w41arl`o;~rm8_XYQ-+ha%+D3RwC#`KcoAn4<5WV~2InCL%tA7s2s!ZF$$oIc%a2+-cBKVq$C` zT92)>1&5rbtJxPli`x0J3%2OIIgQkMcR1npsq4J)h*#t;El5^xyT-HgtO|ZNw8wYs zo@5Ft+V(p_@z_a;RSosYGUFj#{vyCTT4G%;>Jk_Y|6LPu~lVTPz}H zZ7tj64a;K&XKwyX4{mehQ%TiH7%+#2B0pk8SGYiN`-!;6?cI8_;3=2-smYQ5oa#zBeZ=5XaJY}v}2#yV1Ynu06Ld!DydRCX@(C{IaLn%RH; z*R3t&vOwhi-bnH)udy8zd&Pc(B%d!PI=7CZ z#T^KePhls6zm39b_lEvF&sxI$bfi{dF4AZNg}D;Lqy8dsEK?&yj-l(I*N?agJ?1cF z5&f+%Z>l$;>@rLpf;(V(G0`jkV6%T)E=;H);1MDb_nDVOHj1zM<}qLMjdDUODdYx< z;#)W`uJI_3!)~L5zG;TY6v>i|3Jo}?E`u7~89P~4%&ez2m?5=%(+dqGVijD?*_b?L zrIPTwVYB!EHF7heXmTha;9csgRw-42O821n*Aqm&GFG{OX7qk|6{-G#HEoN}9E|x6{P73T6-={!L{w)@(J80ZMwf7+rs(O_yyrD&Re@$uZwJIhz|lKk~IpAtBfeZ^M|dMr_r1kZMiAVCAup3Gsf%-bQar*+au|c}Ftv z9Y=_=ZT2qNJS0-J2dL|nkqevz3-twzZCTK!K8l?pL0U09hQfsmo+^KaaJt=ixfaEJ zFb;-~8iKdhDM>aV()9mHGag^t8I?@312L%dhP(3_Evu2!@5xX83JH_nktG=*cN;YH}%`B|Ao z#27sTkxE(Q;Mh}fEr@^P|3y&$$jgI%)muMd(nQZGbBEKux+Scs?IwR@^NWjAqT0IX z%k(lJS&?tXaVJoZ`+)+>v$aJe!d}h^Ilx=Mthf0K1vNn8e#n>ZCYbVpE(!(XvEtFi z)VTmBQxsgm%cQv@qx30=L(*^3|66}+4iIq68b!QdT1kWf9=U(3XYA}n-7~_wH*KUT z+*pw=$d30O;RT|E!*xqQhD7b^u_{q20vnvWOnbWp$F3D~X4?C?jv)}W^ZpSo$K`F# zSLyjQ$z>3`pj5`AA(?x5WL|}6ZJQDIqxYDh>&BZdb_5}5Q|UG3Q5#!dMC@}YSO3RF zC%d391sT$&8tdYJc%mnFZGBG&{y4xRz0a*1l3!s`|lfJHGu?+GtTAmiyEm> zv=#ShDZl!@!&b;#YHM7MXL2{5_OkT7z?i<1kz>IT!jx|u5yB)N$}}#WQ0_w^`q|xP zX3{b^8|WNqZk2YI+P5rMKF+@H3~GyyNttygV69 zohI=-kj)75yCz5zR(-)Hjm7Ox>9g2Vzo$druiKoGE_cf${ zeU^uJSHH_TWl#OKI!K@O-LfQH+i~RBYbg2HRFFK9rVREzYKH(wn&YF#&i5Um_|Z& zl_T<4Vw|G-nWp25Z4?yba{Dlu)X$oIR6W>jKAJi-4yCtyg@Jdc)ltaucpKSHMsbHrFIZlE%JOpJW^ zAxD2=ynx+DUt_LlN6n2r`aDumRZWIVAifYr%}4wi%7Oz3-wn$e^QXrDWu!97fd63HU6IaV{01~kg>)mUR5J^JgAo?$9nWLpa!K}j6`QoVMs~M$DMFd6UR!&2hVP`Iw=azIVyk2OP7`%5$V##Nv6<$K| z-YUEQxoZ-kPS>~lgr#xPss=BPQXG%wUV54w``z_WO(`q!l91Fs|%hi8is1y>9 zh2S+m`SQBK)$Dqw0O68a8eRdapkqg96w}eF;xe5UVeIos#(e$z zoaTICjIXoAGfo9N9p#)?qVUYq$hX)4Vn_4@}K z8~kcKl#XEZ17vqL7w~cWzH5I!YKMxACz9W{CQ)DsnYD0{q^@Rc`J%TaXpR4{Ka4nF zd(NWXd9M}<^x?Yk%k{YDk?2lXf*DH$BxAWEO#DnsvfPEU07KwWa6!i>P6*`*>TtC* z*>|4&Ars(^V4~yG+=m5CAv&TvOJW(h47>y1 z`%RTdYAj7=yENyB*6$6AGON^Bw+CHOXy9^)4?9=IJ$$R}t9TVhp%8hF=AKUg-(s~i z=Xnf7k`@H@N7Kkd*@tUqU)E@RZ)IUxcLJGX7g(+DpRq03T%%Pm*dWw09XaBSeMtOH z9uxSFSf6^i@QsCiLj!+v(+P^@v=$UO51yFNs38n_Nv9VsyC52S8sp1I+QG3exy_@c z!6|Me>670ktM}n}wA(_T$ zT6B$YLy&C=Q5BkG*5R)p?!1!M&#@PW*k1aXgv_rMVQK`FMb>}!`}(~l;5|GUO80ku zTgf)J7*BI%9>LN^auSxKlCzIhz^O0N%}A=w=%XUiYAw-=3vO2!E5O61d{TIo4r@0X z3ZE*=yHjRIhl?%*We>1fC;MHgN9BlybT_ZdgUq4e3t+^`C3~!4+1CkRx$#|EyHKSD zb&Zk71vUv%Kiz-!Bo6%7eDDsxDEU26(d&Xw)k!zZbL37LgwB~$s-o5V5CZ*ko?}Q) zx(iPW`ut=JB`g%R9CAsU^62Q&41%vXi%HVwU@YOL@3Jstc+RQPrDGvO-_-%8hs-~3 zHJi|Nh@gYNFRQ{{m(a8`4U_ksR?&7%H&%jd3953-I52+^GBB67=FqVpwBNw`K~$F8 z5Ikp=7Y6BvyLBenFSzjht?6Vkf<6~5?@QXQx%H?Xgg7^5ix6a5?l(iD2n8lX2bfaS zgQ=Y(R#r>MvPYM3cp4_%!4JDWe4-^=V0uPQoV$x!H^?EHt@IF`qEV4xV2U`mpOjBO zX&aUu7|MSbh}3*1u_#?|ygb6>Mk3=IWxSgHO4vg;*dr%@_vP!)gG}R|q!M{q{8S@v zV|&?5>3zFd^$`13Lia^;;w!dAY11vewO=h zCmNM_;P6drBYie(utmD7n;HhjT>bo6PHyb{-O^Ezevm-NZj5-KER$`&KN>hQK~r33 zKK#x8sRX+dAV6`d^3d{@+w{8r1fo9&f}O>+w;3yd$#tmdaOd-_A9=#2>GieJxi*cjEhd#sY&- z&o$mZD*5>{-A?b;Teol~yq2!R`5zG{!7fuqW(qgJ zcqW`+Hu0JQ$+L`uT$^^&0aJq;1@) zJ(aFko=v0Ybx9zDT+rLSX5D}A&02qenH2M+)EsFcsdArUcxT6>EM{wekzu#Sc5b_% zrf_5}mhzRyJ)Lc3*46(y%Rp}c^+E60^cTh(BG8_*IMw)`u8GQOFdF`~Eg%?uEM zCZ731w}XIMiz`2N8^p^+RC8lRV4Il+iBv(>L|p>agT>b7fi+0Tuj>Rj|51Mlb^a+2 z|HLkCLqFZQeUvv>KULm)fJZ!Ws0C}52x5YJc0**>g@0A^-ARL{^BO!`f^R=s;J%;U zw1V)M(wSM(1?B5p`rz05TKE1HBIJP6h?9Zw2KsM3%1gV^jhX{rqAI}0bD`rZ z#jr@KQE&-eVo+Fr4nQH3t(F7K%uXJ_n&2hTj#xs10#Qxtv5Wx74eFv zDj!_v9Z}LwwqgDB6*((7TIEf{jc4GFb8IuP)mczP+T5zG@Nq6vkYKW=m07rHCg>eu zn&&xO-plM(bBcT%Rrru+Y$~_@=fG-TveF0P=OTD;toOJiL?3pp_5rVVfOaXwFE#=z zyezML&EkPO@>>J1+sc2YkLn-19Ww@}7X@HoINprx0L^vpm3@+F<>XiI3O_&d1XIn> z`a9EGe30gSxMxl73UP8*Pt`GpCzBpwwj2<`Tpf$nf(&a=jo-5kt10fXsfG@zPje({ z!F4>dnTW?Ywwo++_OLRrXPkpg&-)&UjJ$vZQ=^%Kp~;ZKXrq5o)Rcuk#g$__p&VvhQpK>Q`3T4v0h7K=wp{Y{m{2QI^aD6@Ii=g08kXpGmLMeQ|B z%?Tt{B-2CsCx**R;q1><+&|h4t z+zGYB!Q`eo6U~%XBJ84j6g|yZ1B=>Y(ug(ne>OJKtF^Dv+0gJaD!tR z@pm|JjvD9b1O)e1ExSu$(*$2XJUM*#IMEP+D;MHQ;+2|mp z6BIR2FIciUdN~`<0FU3v7QxMbRFAbcMhu{CN_n?t9`Das&s`K_+d@5F4D*C=E96)57sH z9I}67ZN7zNr1?x;kM+1;blj#TcBFteGG%m}g@p(n#0jiwRWGA#5SaG26L@EMI~MMJ z&&!&8~aslX(BhF7B97v$N5m5Q>gdX-*vy3GW++f{b?KosUgz zPnz%W;S?#qu9gDh91J$UnX*%8ZaGJ{dZT|N7aX4z?1uo+@kh-Q>`kmf+n3=S2s}N{ zfM+@SoTf=VRC265IZS$cpl_NT7rb?mJZWqf34+cSod)?V?0;4RO&H|87vmK=>Ss!&Wo( zdSF^v;41?;Z^1$pi5c(BqrIHvqt$=&Nj{8NHNIk*p`sNYx*YiqKwH=z7-4?+qd?M` z4{xoOM3(=q*Y*4HqqBTGX#-wcVn|qb4Tiz^POS>A2)NCFB^~!X-NwQtH%Ba0Wec2! z+wof5^Q~72*|)jkvYD?hKebZcN;*=3Rxa-TA)v>w(KN+q}B4fJLyRpwJ6(6eN8v25Bm$HW+4;BrnJf1@`2AQgIFwr-HI>u|(zJ!VECV)h4 zLu$Y{zJVv!&G~~Yi2r+9hQjrOU4|lDJM>sdL`WJCORC#Slr%K%ip4a+Lr(X z_+i%{s-8-qqHmdIZ!v8G`IFGR3_3?t%HT`+NSs#KrP$x~0EMWmjdeeWS=l_I2eH<0if~UvFKa49s>XuHh7O`IgDnFAo!<{d8uk*5mACjy;*Z!hJf6qpp$sl&Vj%C`s z^zD)=b>)ieYq&otd?s#EP+x38c7Wrl2_(Sx)@*)fZzO-jP~+3kd+^T^C;Kt5TG%m- zo$-lAC)EZ}C8^Y=)t34Y?&L^wIaC*<#5 z+hEdtVOSDWgG5535mHOEsVqLVQ$+#geQ5HC#9S>`rc*($+Evev88Rm*oBdy&kpcu z!c!#9tA6*HeP439uGi-bbHJm#wy=(l%we_JT9?>mrK6Cy_hcDKE3%&mx+IY2)C4+s=dn6+Q}R+VpacckAnS607`#lQaW!>^X@)`;;v(RqRIFV^b#;K zrYfIP#8Nm%VBh6TaXp@zwg7|w7RG)_o1)KKxs0casaL})oV7^DRr+6v< zk_uI;l|P16qup7x7Th8r2o}P#dV0(XPl|seASzFalY2@&r&WG8{U}%sImNch`I9FE z@6=u%_NpbVs`5fnA4xs$tI@t&&K0S~sjQfw)_K}{dn6^I6JsIhf|fV>N!|i&tKH2i zexjF0p3gxW`>Tld#R0!GoXnahM7y>mxnz}mywkbO(HiEB7myCFK*5%mUZ4op=19@YIk_}ZU zN5~_M11Y9w__E{ToT;R7o}VVBu8Y@-KIXT#_IDV4Dci+Wz3C8u+dr*!^wp!Z{M4eV zKJNWh#JZ8$Js^rV>)y5InUqr4O!j|$A# zXgT-js(bs}nYhhVLMfLIA3v^jPaf?ap=|zy<0;&5YPODfbM<``LU5ZS7v_IhFe74$ zrpR6vwPLAhUFZC^Xd&h2Qm1r^x6P4D?7BmXtgKEmpYF4YZ(?ctP$wzftk%&ihcMVT zZ5N5`hh&kp+|QqjfhAj=q(CjH&xOzt1W-?)BlvX=A zQ9rpjZf*|~M}lOa^(>^=+&ad4hGJv1buuCd+v~!F32o0OI z*r>Xn6>PuDel-<;2e%TQg<+=fuD_r@2gP(t5}wumV7w=6*Xex{7q)*{o#$D2osl_L zMP|>@`18B%FQ`yA=E{|I)3nW9z62?d(Q`LTUz|3K=F9Y9r!WuCbA|@qNv`wRoy&;8 zrL9Rl*M+oeFBBpv1f{}ka=^~`GC4Dna2i(m9Tg+7iw$f*O?eXlg=(+0{p$L#5RJlA z-}aS(2e_g#{SEXJ$VY!nVZkGHHxe!$es~=agbv1j6O4L_jA8TOelB2$W%Zno@bkeg zSqD!W1ARRJxK@q&pm}r-I(A-g;s9*;sdwQ6Pj&a;LQP5B zIKl!2srT!Y=l1YT5zC<;)3X}|_l#4g*g1+@>$SbpI>-ewH&%b&>%TAdwkt1n?hca( z+h%eQwf@H95_X1|Wesz(XB_Tv5S7mk9-8a@u~DDqtF&^!51+7LxLDK`)4j)Xq&wNd_3+W*dZ;j|BjUS_%HZjkc=r$ROl~Q+a zbdecz&-A~ewURoW1}orM`l5%mmIfj~t}05p^%TVlx9m|y+eDDtQf;Cx$S)^hT6a<{ zq#E=PX25^Jo53h7(&r!{WXBJN$M-!#RSpq4hkxO(ZW}%Pgu`C_(-W4LwWd0ov@}X~ z^pmr5{s_ebRV&_)ACKmFzS3VTk}I10Fg|}$pAqj2GCIAL3gx4IX*5o!v>PXke6CtT zD{>FDCUJwOAI{Stt#uMw8Q~uObw=6a>nbO^RNQ|HNt%SDCLqTM{5q*_tVqcQQ&CEU zM$&PMaZ4+~Rs=J-Hu2Prg8hvxX4QXiSyCr*mi{YsfixLyhdx|!01a8L)R5Dj z+{R~*^kLf_>m7%}XtcUfW;`6qop0*ns6yZ7JVmhlSbPrr$_BNzSjr`KwK2jCWZ*Q88-5WhN^et{q^PROzvAYs)!g4 ziUg;JOkdBHThPV?r2C6+@Bng-o4m{fPG1RWssUw}fN;WzqRTl`q z;Hva+H@|JA$2gjrCP8~rgxR5D6}0K218*Z#(UU0zx(u(I^mhE!H$6Eo^U;6mj)tLG zGk=L3+@cb!eIMRQtqHERLry@I4J!#xD7 zO0_%KpM06zz!{UXB9Hx=LMVS=&wojLwR5vB8)!kV9hI0+vdFYXX&ST(j^B3Nnj)

bz`^j&mx`@rO(6Ii&6K>5<2o2W2OCSGW2+-?1Mx8=XF46e1BnIVy zsraSyBk4P-)UAWHOwe|f^>nb<=5uO1;6^d-o|y@}$CdHc4jk@Gh^#n}tthB6$1 z!E%VE0;eUygJBMMzrcNWSU(y&8mv2 z-dtSR2KG_p%twC_@BrI&=5=}CmD+ig7di;&I;AQ$8(P^gc~%el{6p23$E6=f0|B)ex38z* zR&WDQ6sC0fq9xGm!=uzl&8Eu^fhnY;g#psi{oK#hma@Y!^mC=on@~DVeEgc#T;Cy; zo>EyL3QgAY(C}xndfwZ1Y!f)(o3qIj9BkYqto3cVpP4!LSO)InPF&)%Mhni3K~~z=Z)s}QM0c#;#np~k60{E9 zxF`I(6JEI*=00%l^Ve;z-NN;!=2i0FF44T-Vf?N%de)5(gf+Ua%Smn|)~#cCUM0u& zqYHm(**qfn5sR}nqWnCaa*{g%`z(^+H8245QyK9|Mj6(jWLO76MyI1L3IyHQnAn_g zMjn35%EFZl0p@+aX>WFfrfx3WU;V_9Lb@~X$Aa8k;X6HZWFIkg|*1+Mp?$~yzRwbEf0U-!Gsz1 zlM#Pc^hs8Ca0$2~oR~@B*2hp;lmUGTTCkMQO*8vz21wEvF$rQm=<1PS?-bV0YNk>X zvLNs`!vYwk&&pMzeDMefDiv7VjtU~cUt&V=7(S5=={iN%F!TO^!njSpkCr|=weCIl#N}m%yw-c5cY!Z z6BHT#gy|QZt025QiZPMm z0;){^C!+g@U4K958_1_=22KCG;7#WoT~92oE(=vCfVLFsQ-oeJxNpYJrnxqe-Wzk% z$TM@ZQN@BHG%myaMlhBBDjOu2)|ZHlCPt0kWW3?b(58B3o_->G1zxpZxBh=)?c`}g zVqVXcU|fpM;CrdWL!tNJ&+4v$c*_3t_pTXT{X%mbW6@bk_E{=xWTXU5L#UTW;%U0$%XK>lbfVP4bmNvgt8d56} z9lRpKShENZYd!ghXL(~}9h97R6_1s@b3Od;rgk{^e2hgN*rPVTMM@hmtWFSS1`2-B z_co1Jjw#Qeih}NdQud2OkTpc=XC+qR4TbgYIFffq9Q|InQ0SwUH*$YKF*2Yd1@9y8 z>24Lcw?~8U0>0%RW;VBS(zO1jOF{7vb%`$Z^iX4R&i4)uOYf#E)}OiasW+n{#6W_)8RXXK+6Mb!SOS4b{*r_22DfetHnyeOf>P> zXH%}|=+bN+5vT#97Y&2|UfGGO-WDj%Asest3pyh{VG!BnrgeXN$opAW@hN=;&FTz4 zJjGDUVCNC+X>a?urx&bVqttx_CKvr&XI5D%>5+JJxu=VdZS+kG9!tHTQv(VVlp62| zsM_xk<|tNfudn4(ro+eJJwrZNA}c{V8p14}PZMOR(u>a+mGCW|E)moY! zlaBYc+Kdt_a74fG>aK<7jVz_i5$lF}QNm0qDvm0>@BhK{Jn0ob zTQ00&!{+0Ng>DVi<|MQYj(JXUvpfpd?JqM-kBNf-MTeZxH8lOVMdcWlhglWboPp%5 zbZ|!J9L>DUAN#rIeEgwx-tRr8V3TDeQxNKzq}+ea0vg(gr_#*HxC{5Ck}UG8Hm^4x z17ISdj{OodMXqWn$oZ~PH zmA8NX^G@}VRKK$;*UaOl>%1%#K7KRJf*6XUNn4unio(^eO5RY0Qs1Jtj6!9rlQKsm z6f7+Pf~{G5Ec!01MDTD_-`2U7#szf`)@MH!HCaPhbBQ)j4p!F`>Mlp|QL@^o!h*2l zZ37cuKeU}yeC4A`DglZZ3*F|ojemHS22ux%ws@rQ9+ z=0u!DBCHq&iNKI&WD#Bs#Hj&jJ@7n$ZPgfH7cQ6)_jXk)0&3dggzwfetQ| z;M2#>SVRmL;=>RnErQ*0c?yE?-6g~OLhgUrGr@8^wSV99KG4__-4cX13x zA)F;316B*~^Hgv=DSR*^V{d;OH_@^LVdC|l^*@VTX<8zqi4lE&&I|>6YSwLmy~o3# zLtMAFJ}-ac9x|BZ7%hXtXVP&TqqbBL&X$8Fwc->XeYK#ccUdHHvw|@qV#hoZN2`iT zW3h}(P}~_dIDV_bl+QSQ_xX{l5!&C~><&u*({baAq&RAy%uFmb8>4^Uxo`BVvadV` zdLb-CWasN(%!)j`7m3Zx5F$klJHy6oMA-HyuH3+rLs+W|t;qDZjT2%IBPcl+j>(oC zuliToiw4fTY4OQ)0%$$qPIrmt5%H6$xf;{>`~f&m3w-+8oAi`Pn`HrZt+M%$Z&doP z&&1Mw(TnT%1K7}_WafWKV9Inij-~2BIS;LjI?2i2Q|^S#RyG>|%glBiU)VJI;5dHA zS|KG%0ls{^&KB&v{6auHK~w1LM7*UHd!Hchvz;`Ow#mfdb)(X_-s?U}-@TkiF?;3Y zIJP)+Sgd3%97L)KF4YhL$&E-`c9FTet6YS6B5+P2IWO=Ny2}V}o4z@WCW} z`%Gad=4KNtKIT+V(p6MBh+x1({lB}Qra7Kw4^W;mv$SZ-t`so2h9-mQpu70PIcHUU z20l`8)!fq3HRl?%D5B6lk~TBB9+sKV@->VJ+rN&G`DvgyCZ@*c0^{z@Fx8p~95t~c zx7*B?+~`V79@V8y@L~YVKxR$lh>;7CRk2wCFm;OtRDd93LclHcBATsr%rtG_T1Fk> zw*rpQh)Zsl!;Xvfar;YbG8+tyutIwl5Kn+8R50O67!YoO^3`3jyOM$_fzachwJ}S_ zyA+a%I7KZ@yUurl49CwAVqvESI0Z6mK+r#(UsCrKe`<3x6h#4RBSA5Z)(gHHC%JIU zr;KpO2TZHLgBsE`TM0(nyQ4Y^XBeX$vjgAiXeXGlJ zHB6cF>6gE68o7*BW4iX_syjU1_6FLCxG6Y30kb=78f?o-*xylCmJH;GzWZzGZyeaL zTbI#Hn-r-m-Uoo`$d!*Lqhk63-Tq0+AlBR)pg7@x^Uo|>5Y^*Jdp8q0Ak4dr4DKt} zr(}au(rsyb0v%ViqnF~67UhRcUZAm-Xk#nEO`CV8(zS>tq3ku$)luQkCZViA}Mr1KF+nUBI1m2q9+6ETf zf^uhYX{N_yct)mI8Gi4J}R;f@AT&{pQ;)@(r{1%F&{%CBCsb4!}p5j z#Mv~J*rA94ERw8Q(kFfF5Bq4lv=acdf5>)k0-M*HnU73oD5W(c?zoiCO~bj9b89)w z-$?<#8B9P^7!{EOx3^!37af_{s%vdEZT-`8!7cqAbhWtO0ZBmj<~N6!x5lz1G2Yo4 z12UArulB!JuKHuBjpu*9mbi~5VFeM|fS7L!KZiHq(yDo^u@QK6#^1v@g7`)cUePsT zN779q#i?MA!AjESW0xC?IG`D1^#~f6oCga$e$KzM<`PYkN>^=IX`zM7CnxdG*jL+R z#jfE%$3_)tY!-{!1;Nz&Sm_Q0IT1c$Pdkl8#(;CeRv`}?oa56W{QmQm3&aXD`BViN zFl>~Fj2+@-uoV_>!RRk8T_hnjoh7OYD@5yQ1ycH?5Z18VCAo-rCko@D!uQAhcekxl zYfD%LM=H5D&Nwa{k#6JNo zu}^?asgWkeoE;9j7NlJ==Uc)na)j5i!a?>&rDYS8(#xt}J=(1v8VxQnZ*f+Hw;Ib= z7MSVz60-FxvmPz+w+#ZSTKY5wplM#RO1dvjUqkvZs4gC2iqRugQB$PoG+j|}z8Kp-ULpms z(pb5-7c6D|_=P=G{UiOC2Bpu79-o~{J+Ne890pldcl>jd$EMmt!&l$oHqjZOJcz&k zFBCMH^Ri)&rz~{qD1GpEX?(Dch!~54zWlfCZHnghhqWw43N-!w=be5F?%$Dp0%uFL zjqg4yoyWn^TX`rm{XtL2r-UfRP>nf24z>Gh9u?U37ul4|(*Q|eVYJArR8Yi>Thk{c zwDK$0BD?atDZR@I0e%aC85UK|YQ$H2S7H(K|OVb~Q! zQmrv2C3CUI{jVp_KLQbPlgaAp(Q5bv61$$VW?uw_)VPY!F=xRGn(2YsYOS$=8nR|O z#|fZhqIc$A?1`J`mlO1L)8rec#a)-=D0!;Wc4w8%1839X&znW2XNr+??N>qm^GXol zACNh+CMKl7k!ZZj2nazB@%b+JE~n;z&UVRP-G+_HVg=8UCPv7Sw{UuUMbJo5sFOW8 ztiODv%*=|H67Ry&ay*?cp_v4LIf1l%%ZRKx^@}akxOH&PjBJ@@Snzf?g@|76+|)yv zX}?F7_+PV>ja)l`x6so1M234-4~XU>MGWy_ zEJ#HZ5wczS%JKcO(HLn3Kn2w(k-_E?d~nNoQH}am-&^9z|Ewp{X5V%3`eL5nCATIb z@o=5c;m}&a#xCvNN{q)VUtr_+xY-*A;Uy15) zlUQV_QzM_ABJ)gy4F{>l||c|MikRBR9Dw5y@ET zEj{PRdW|Fn%Ag_)F|`n434Xbm#_05jSYUo}75@-xALH@wgZtknnI$1p7E^JsQ?EnM zPN}4gus~q=3~1y@o5jiFhA0K|7-$(2R95B`k!|&~Xn$vxi2!rrro<}e$kG`8?*t$@ zrqt}An@DYm>Ax8Oa0RJ=l@p+c6LF;NV{r51)ya3^U!L^p@8GznE>WwmqjMfNNjhoXl^=5x`#Ys*cu;WH!`+YjyP_rEUnrM{+^s zb%sZ3Oqj2+jG8!ho^G!voYK=aQSN?KRM{F!4*dn?Tz(!v_%eu6an>*rD7f;4MpShM zO{lR&Jye&C;qsR*^{fS9<5Y=5)5$eD!CA(%X{)5-=(GiFVUq>xY+jailCVC!CCTb; z3Q?Sw>!9MAp3tC_3@)W_W2sN9e77Rl0NLL`vn~L}r)IgLH#gLr(FYn|vCLfa-|88b zYmQb}qtb8yPK5MAa-^o9k%;_K@}N^LR0(K|NcD~G4(v@pj~edC%~%y6fB~N62L9F9 zMVk?JO?WI9QdQk*bm>>YC4vP4#f`CO4+!EpBcEk-pdUqf3!JP54Fs9#9A628-=K+; zaAsr`&?Wq8_uWx&H1x-En{CWAZuoG+zcI8rDcS&l5RK01?NN5;vKi5ZpSEShMM-m` zeZ&;aD(_a7Mf0z;O;r16Lfhk1BRdka(jj|Lp4ULyoA^H8kMk%0fAKtSO1hbPK4w4S zq)4}D${IbMvmnmn>(Mi~S$PCTerYR>WjMtvN$8Uu-_;1lsbl3NZU6B5RCU<**ohVg zDvD|W=vrod$PEMA_g&53QaSvWRQh{VHXWm5@q+m}u0DK@ziwo>e`t)}ok=t+F5P3I z1pWXAqEFU;vnuu+1e^1LCb{xiz?;Syy$gnUjt`WWqY7m;o zN(}=?`Fl`#C*saH@u?x@Y)S>tq^2|fE-zaE=p-BHX(0S1JxU7Hmy*agB$_ij)&&)*jHg=$HxrG~*XWjX@OMrlHWH zCFa+KT;;7rL`|QPYFheKw6P%n_1aD8t$ivAnNMr(7cY;NZ~H!&u-yh@5Uxe+dU>sXCgGqGs_j@~1a1?@GQIWs##k&j22wqjbu24V zBS+J>IWsQSfI~==%RAf{8HigloH~~Slqk!R4ke*eyB%5DMYiz;D4~Z#@lde(SYM+? zDuy1!yR0BP*cFbv^QLDJGe=u{rhDqdh0t|RuwKaeU}xK*Kz<`Fc1z5TD*)3S`V@NX zXq?`T|MR?UZ*jv2ktHJYFna}BGZFkDq~rF+#r>6*#TWH8Qe$B(rqbE0!uJplXg`DM zmDxmdrAOdDFigWjM4+CBVkZ;Q0llHglx$vE!DuPpG_@^3s1z<=;RUsa@&RUmV09~z zf{`zz+~cOPmgQz(@cOM>uTE%M-ZpUHSlDhn)q$jnyR5JAWZ)_!kumX1x^BGJL_K#L3I)mFjdF(#Gzx0jYU$3 zg|p?}v|WLy^5se1KOnbEgYI*OL)$Z&W>5Y*H_LVw3iIb4Cd%svjm1z`N+d?MiaY~G zv3Ux9ljNnVB&SyZb0Lx6E2Jl{vNj9rFJCW*;GA}6A^Og(uWqIwUF?`DKu^^v8oKCP zr@x#>GsuwP3|+lFopjm!ai37^Y^z*y%|U5xS#;J-NK;Z!DSw<#JWdYG<&tBs6w1Rb zhvRn-!2r3hWnN^p8&_99qC5OkVt}xzpa=MfTBUlehF$2qJaefF| z0O~9h8);BXl(-TWYUePPTs%rJJcQ)rEE|7NT!}U`rX3TlcvEo|UE^YU{v0hoynCny zOfV@so>QD8j)=H4D}J3!rP%%~#(J=23KpKa5U6aCVG9}tyGkMmAC49!jW(9JF|9fl z{9ZwjeP>CS@PHPZt&v+E!4IE8fZnnez+m=#S*#76xC$O2n0W!4j<*&ZO<5p z-VP?0aIK%0FMb!2r$DkBYFP!*v@j5vCuZ5mEcqenQy+qpT79NRwpw^Ah&%Iq*rQRKQT`OoBsdZJ*b4HR$Kkq~JTuZ5I1}0% za_Os4gRyV4C*+epC4+vf3Th2Bz+YWZ{h($Dw$WT-iseE9*)Lp@cdiSBDj;q8J4_O| ziwFokRQdwgo?ySVr?pGqir(kNg3{sjs#3+4!8=0EEfGFkxUgjUI)1!t56N7zeNBH1 z!AMPGNY3lKd&;At!?azo4yK+oF-Q?LJ_+ib9>~yFF0)M$ zHT7v$GdXcSeEB03sKqW4U`~*tWj;+NP`8P@B_MjfU%m=54SKgiMsyCtH8R4VsX-Y= z2s?nvYQCBtRY)y->Yum{4LsgTzfl+HP6QYbG&zVRclFe*`bj4rZfdaL(9w}RfNO4l z!tr$j^i?+hG3maNI?#Hyf)*ZFE+{A1TkT>aX;Zxf^0gc|1xyABa7$*1Y9@sZmGX-& zadY|jd85VQsOCW0f`vd|F8!=#JSyjLADF-zSuw@7j|9yc>BCXua2{w#0H&G%Smp{f z!2R_FQRVwxGID>AzpUcP>yDZTabmv~=|<33>_{bqNEXdGc*FnOWrN9LT&NgK>zPVk z1N%Z0;`sM)xf(?mU|s9P1bZvWM4f6Ns$Sc-;6umiZxGI<^+bWY$ZKVDwG$NBqA+c& zIswz322KU2bD@q?rG!XxMXY$XN&$ktU8QufTWnp&KwpvPTO(3O+t+l1d)cUP>Lh2Q znNW6$L%~4-!;aFN7+?v%+=J%qJ{$hBY6*YoH%$$~_`V?<*0%*|Wa}n* z|G`*|(3PP>*BBDc9Wt4DKSaDhi2%_-(qfAl^K-LlhyTW{IeZ5e!9Pu}4D|OHj8W)c zfzy~;urw>atUF`4{l+1pRLk&0LDrsJ&S6Xg8KpZ%Zb&#b7|-CCh(`5xc$RU$rBIqO zch-Ag%ZPLjK*B0_uVKhPb+NqNZOB8+E}oW^Y<&RB$H|JrO7=$GTdr)oBt@q`kPXXk z0y=IWj(?TUQcIKNz;acAiz(vXUDK_q?$hJbVnq3RdQkf=$VpZ!Z6IWfa2<$ACr(XB zGEoI^?8@j`PYJs0;m@*r%e3!VvPyrGV3}I zRzS(W*^Zug+xTW6~BD59* zq2Np+R8KJ?sb;z2L8cuE;zJ=9ZMGw@3FL7gEp{>4$f(okfc=?JQLxCQMiWOl#tH~K zto8+v!wLA3_n&ORQKe@DkmF^%<(peCM1y(N0l)Aq(-*^1sGx){A} zIe>N>pjsvDe>B`mf}m&9Nc=(ew`iJwUA6z^ycFEUJk$*kXl^kTytPHOtyWe?R_(Q~ z2Ow^2c|LQ>C!tDNALanzlMj3Wbtu!DuR_vG<9dGfex*p+Ky<*}i!GB0Tzo9KWw|Ma zCUx&LYl%YMKlW-;nl*^87WK{T0KUfr`!UD-Q%_sznGmGP7w@F!MEnjNlIX^i581gp zJRTQqaxVA<)$zm=JAoczOQr5FaUK2F0G+0QSyo1^lNx~+hIr>y0EHOOTkS!SiPnU6 zBne;os8Q4>coHnQjj3Q$1cD)xq;_N3q@;bfR7`VXIY%D*geg0IwQ2w_LD~U{GeiVi zzj#nY7OE6!T;1Sb)EoaE;U2gx#%d}b3s5eKPk_l`>1$IYA2D_q+abJT zFMn9KkP?`q>o6!O5*2ugyZ|!+2yQ-u0_}3?+7fq==nRZ{Tq3t1lolg%>+uu+1DD-` zui$`{GiT%Z~ zxA2xVtG29l-!xNnCh;QVavMvNyQ=`%izA>)3eUe&;rgI$>HJbZPPWI|4+1MnkDoNY zN3UlJ+^*T0W_O70K&;D-D|1Ft4Q8^-yIaEa_c8OsoEI3r+iW0Y%?On$0FVJbdgW_2 zo`MJ$TkHTrN=xfa)A4(s{9e8rVyy3PD@{!kesgS_`WEkM0H#}8?UWsUtEv+vV~t z1q<$OQvoT$0fv^_<-kctzwt;7!KItr;gec5Sp_pzS~~Twp{)d51b7VKWzAc;j6cX@ zXNebJ0B&OW24I6S0L`$L0@o>ibdj=AWLXsDMpfo?r%XO}X&M>S-@4cfUgrBGgRHhR zV}xvE_1K~U@4N1Y*=qnng`{T_`asj%lN1!2ws!M>!-V7=Eh#9SFgI4#7y%PbTuRybzm`#N06Y&%R8e%njtFoa)lV|K zY-yVCsmatYVP`I_zYbl5ZOXsS)*C@xl-1E5PB(Mn$9au(ag}wAmfLd&)T@`G*r&CE z*3Y1~EVM-n7BXcNYPgxj!Y~zTZzXW1aOr+)_4^8)2Or>zEfi*^T3~{5?l1h*(&7vDVKw}+)gW(vR$wmiebRc_UJjB zxS6@SBNpnZ8xP7KbT(8i6}0bLCWaCSneLOL;KXSP=+=&JQ;|H7uqTNLszf*PF4mL$ zIflt1xi?rq=2g4*UIaBm`B={0dj#aH0qx}i2LruXfG2XorS(mn@qaEvI=yk_*S=it zt1!Y?yh&N{9ROUG@0^Hkey?r(+*ohUqb?P6aBmJGNuf)!`aImg%smp_Ei-%X}#%T+fa=wKk8ow|5*`tre z3`E)|z&`LOo?`1l^{sd-)zd&8q(bYUoVDOa&mU=Kh_;S#J4)1vEKXd~7ujN8&vhuY z6-^Vcspke+K>4<>12?F&8Igz;9hvTn8>M@v?u)uJM7Tdg$ok&Fk2K^T+*6?Ux?K4S zm-5CP;QcSF@vd7Qe_Hcg>CvQhL@B(!>fBQUP!>QK6}_!#Y3#|161<&98}>>H9@Alv zpXzkOTefe#cvUa*br}~HoxyZrx&P|={?DoydaD3hG(+jpPS2ZcMC!%T7;^efq%$PA zGb|cUjlr45zs%Kj88>@N-{7%hXK1Zf5KqnD#c2-p!=ap)Yku>$e)2>iV{PCxJQ;EJBAHFKGPNp-Kj>|m__Rg5 zuAj%a%l~^}Q69rZaJvygtF#Yusg&v+|Ino z*;?Zd0m(Mpldm=U^`F)1AI-^L$)1Z>?Mvp5F5x&nq#GI}J|iq&MuCWU$B^kYz*P|~ z&jB=@;1|6Ao!KJydlm1Q2P4J8a|G(Rc4&&W%gKJ{3clM&7qc&wMb`ChCu=v#){XO) z4I zMpTN967trxT2Yg|YNCO-c#TdT0uiUc#S=x>m)A=TP<2qVQO^$cRG?jyrAJz*P0%X- zE;}vpO@HA=pO*)l0B&6Si*T|otqXn*-_8~<2`xd>Tidw5LXQ!`2Z(TV=E4>we;Ij)qwcJk@V-c+w1z_`K*q2f^kP_`ShJabD4eJ2#cWUo( znVfG-jisI>`a$2zC|Yi493?!T_YhC%jJ}zByd1DfEF-=weMADvD^Ev9rb2-y?KNO^ z)}nB^QLLa-n7K)Q{_{5Of@1vXZTYr$yt=z)SelEafJ~fS`y~wk+F_4aB+Zm#yocJb zP7St}gk}COP6sG0fp06dcde{kGtdo*Cmg$LdDsi7D z=-s86hHj4$msbnaJo5fU_pRM_a<+mkZ4ZGPIt$(S&1Qxz={(xs_;$@q@cikbR&u2L zu3?GBVIZUhnXP3COZTXWyCFryaUMRQ~jy3=QXW5F?!ygnc*%)o{-Ve9U z(%TG)@ox9yGh;SA0{@O9t&%$(l1i05WNQ}Q4%6Z>GT{a(Ktp!e|Eb0THwb`Ega^Pw zRNAr#<@{O*$Y9M|e1l_~Im|dr+f=;JxOeiqk4>(lcp|?c)#}-JUxZk9z|&(ddw(yx zP-^hg%6;1hv`HnuCLOG)kt!%KCF|8F`rIKE3UAQ=%ld8Cxb@bU84fFdJ`rJ|YvK<& zm*#OT50TnAO~<12(hh&Qz3Iv^#@O~Np(>#x@GZ?F{O2Zf4Z%ZQ^*6kTzm?z$(aq8L z;SvONO&hR{iot!i_v8wFk94!{CKme-RoSnt>!&o+tGPLpOBU_kHB=g%kD*FHx6GF>_5S5?fuAYJc5OaftE)8KV9Fl-`ub>+ZWk| z5tHh!`4wplXEp*~a{urRQC+`c6D zXIv^b`@bu7+)V65Of4mHFm%A+CJydq&S^{XFpSKcOsveD%xp|dY%H`)%#=(_l+-Yc z@(!kAMy_T=RAM|VO#erhG>TBf(iRF zjH-koorq>7#7t~}NPUngVo++DJn%*v0-$i4zn2*XJ5H@9B|>iEnBR-+$0nALrH=(4 z*u=(^gqoU}g_YbFrPd2hgoMY4C8WG-#?>A+BZ9bN4zCOtYl5?#?! zJMa0_$5DMQ6m#ZtAYg)XJJGTeR_*ExekGzQ==GQatypHE=rp_l`r7xoZPZ6MaLX;{ z&J)8VoQ{T*sAv?zgDMeXq(w4W2`DnO=rJ?KOSUv` zE0Yex=0|m2(6;y1xUP*V#fx)pP_mzKO47xn=a`UhV2utfRzNKlkgb^p13>>Sg+8*_ zv*H=u#ghIE-s(vGxL=a*@^@KTE9n5k0mZmq%GUJ||5a7LBv@Wu#`*dBTOWl$F|Xo1 z``v@0{@wUyo3GWl<6;v6(q5mYgK;?-UXak%g{quh|N2);1t*r$2=JQqf?rz$z5>tc zQ^OWQ^^?VlFT*D0V;-CP9-!i4zS-{XaX(o1W96T??h0Bp{)g!$r3N8{b-t!m1@I>h zCSJh$fRNAaC;E87D&|(A?YHcpLHq*KP(sB;yH1)Ptid|*_$ffO@uJ(#4`(qY{Hrb- zbHo<^O}IVw>M{!WgpWDC+Fp;-AGgs@=bU|f$E_M$_(!*UD`7l56YvfRH>m=&dagS6 z+)SraBgBazd9dDn8Pw!EWmp?%+*XCFWh5xJ961_|_l{)t%Wel~+>hi4DSAPi?oU$W z$;jE~&HUXu$b@b9!`ZprM*Fi=<1)Uj31M{osXoZr=9dy@GaPCN9GD`LS;fWhnthbd zxH_etBDDVHXPjOXJwVksdg;C0KZipMQ%ooRF1jfj#)&SOIaO{jsZL=6$i3?tx5FQs z`I{Gi!VHs~X_`tvRM-#R!T_Onm=%VbLz?ZZ*Y!Xk z|8!eoMEBYBg17dmrx1B8u9V5)qq>@t(QwS!vhx7nxjtZ_1Rz!l*Bs0;0z*H?@HvlW zB=Hjdl22#)+C`sx4sl=xVar#K?$$tGz8-9r6+Y~J6P5sDbvinqI-MU`V42Uf!#T8o z%J#>7xIdXpv{QP9ZK?;e=gMwr`X9{Q7pk|UuI4DX&%I}u-a6QNf4$#H`8YiEK&27| ziEh_cK4$gs0D!FL!s9n)c&y|+_z2czoo>diPX~YwA`7N&)^O;>@2GuRfrDVbvq}8G zmsv%DR929rb~d<3N-NBexJmbAKSSC$=NMy8ueG>sYuT^uj|=61Naxjrv*R>dH8#p? zU1hHCb*)wk6J-oYtdX%`tK`b{Rf>RcGDnn$XySGhGGIT%B;y2f1%1n5T0?-xjJP~L z8Av(bFQh~>5M*|!R;xJ@G(3B+x`Os)b@ikUB_QfJ|Fhfvx!T#qB_+SgKyc=hZE}fH6XRmbC6?B_wIC)2 zo~+2t48Uo(dqnLF%yX4N;lnnu4|7fykWYL}IVEZW10J~WR#MsXI$o;0P3V5RLt9|g zdMsM!zrFUk*esrLhvL=GEB^*;Z)7RiM%tVV9L+|t51-!k_elZ5Hbr^oDNMP(_WMI^ zb=^USM`I;OC*Wq4QvTFJGo{+{M1vr8c7|8i4nV0w(fs&8@O+>+)ad3c#bQVf1IDgwHUm#y(2V+Oc7BH^<2&Vj3k^M^# z14W)l!LwTejn~6I>4K$}I(s?bpinBKL`XHGc<=!5^9n%q8SITRp!(M^gPDaEjETbw z091=;5!o=?cA6kPldy6oSWyv7vAAF*sfJZ+5`g#~9>JgI0%mxULk^%)4_x60Bx|9x z8+lR*qLGaI?kvSjv>Qx99?BmoQG11$$_m9@d;{|PpfJ$jS4^=}-3j7$Z$L=UJgAzK zC&(FG{+(2^F@%NV;M4Owz9Ug5dy*dQ0iX!ROGy{GK<+K^>jH3gSo@(sbTCGinDXb5l#%6zNWPc=zXWGUbVZYcST0K2Pc z7&i(z6~oL0|eW?ABkcXAPhd4a%nA8NiuO4s6o3ERcmG*(#E}-R4MPOrX5#z>!>DTIZASFt!Z2zQ=`s_s5;6Y-RU8~# ze}Zg8OhhU$jMDbz4nM7*oa(=|#E5kH*w`evMLD=7m^eh3MYzS7*(5|nIXGF_CD>S4 zg*k-T1&IE?T@VF+CXzF=w{W#2VqxKA`i~rneAUYq&jNeo&dc4yyz7E1@_;oGpU`H+ zeV;b&;8n_Y1MI0zWuEcQ5IhAG|X9^w1IyZS@imd4@sD(iaKXmws-VTOc@y)2iI=ZOmd#Y&R~ zG`N8jT8pJ~o(uR(VqeFu%cAQCIZB)joV|9vVU> z|8ih{7&zBI4ma~(RR$O)X;DO)F|10bgUs7o^ln82!hVDt|GOKtX&NoUN2=kk@rZyn zcbi7ypEzukQ}{8W`?23bjNe?MG+Z+1YSAq$2hzsRVJBWm9Vf3pLM67rIJ3@-Q7uf> z7XS6R4(6vXxJ!4sv}_{&C{ej~9&E%t#Pbxu#z7wA-(1cEl1-nr%;U}ecOzr-gp$sm zx@Pj{f0`L*i0e^8Z>brX!S@JWkDJ26I8K3+NY@%3hOVnt@10FzO&)-IFHE5z<#x8K>`ZPvitJ}PYt*Ml9Jt6o`#12Su&@87VRx+ zD4GJ7-S=~8TWM(@;_u^RYuSq!im{!`!4~5FtnB3_P52pk@Hph$-pYb=a8_L6rwrC!>C5(9P=!noK4fj}c?S|$ zEix65&mxSE()arvu2O;kxL$!(#<7{41$zWtU~>;lnAzHcqwDXNYTsX`8e z$Xq{#?r+U-yx@acxO_5$t3_;fs<9T`r_sZq1o#7*&*NWLqyk}p*L0D)+}EIIXtIDs zY!wS%>9*3qqH(tL5;#}!I`k+q^;dx{VYL_rTGvmJ$CzrJh@8158R!(5@-y1E$gH9; zdz%VioIpqHelw_~vLrRlU<6{l5^#4AX|@kEEPoiesv-&h69wn<_Jgjjw@4mozaBN+ zrzG#Xoh#3=sn^AyP0JMl66t2VDj#|X@PQh_Gzp@Z zUkd^=*t1Z|%NA!b^z9sIZ%)-qtqcr3}54SD}faO^mo%i_k2R_}4BnhGMI`ET6MLY}20r={cIju!%sg zL@xYr!4pR=wmK?)Tm0x_%~6$Y@mCiFki(WQh?q_pJvjroD1?;X{HB->pb zdlN18upB#Ii@=E7 zWu%@#6|HS()r8?~Q+KB;+6!^4^AS9yXO8lKag2CRqqSI1H}Dy+xwi{kI;&4~^5FrG zi9g?xJXqxIZI=#m&kM!3-&;x5@A1vgQr_2y>*tA&COyl``qF|fqPks4-EO~h{gS>` zvYOwD^X}u4Hb-1*!6Y?C!>>q3&bd^y?0 zN*$}#*j(h9=5&@-DUWS_WHt|zWm*B|mRn8Zebse8$P;)$Ey#V;~8pAo0segrDRbSvfw!`@aXm3jmgYNJDL06bGBz!%ii?!^jlRu+)n#2 zNP|Z4OH6{EnSmXSVs>$86^@ySHPwg>6a$Q#J(VO4k_OP;O5A8h1$ue{V;7@`;$C+p z?_KH{YtVBq-WjY)t#}3!sk8BF52!jCm(U zIs=UJHRhWeipLy|LIp_QH|iS{ZkZl)%Glu;@~p}Jmnv~x<8Ud5P!>NN_MY%?3B5vV zO@)gzIS!(>*;?1(PHx&x&spC zE1~zsMx`8TvVM!M5^omGGMUJ#=#mn-aT(S#bMQP z1JxL(vQ*3WpBnbtU~iZgO5i3_1Jh95JNpVic3FZ#y)!;lhlf9AQ{)EZzb;CK5c!55 zSb#LT3QO!imKQc9Ov`?EA|)&DeMqmH;b$8c;mP13BbduF~| z)w>?5<&0LO>DI~-m;48a#jjS{20?M`ckuQ)qEh(Eu>=vlSta+M0`@$g6I;{P=KRf9 zG!AsT{h|2(1alo>15c^rPJVxTA6%{t+XGZ}He}GK_0UaaS>lV{I4dU<2>tpKE~ zvgf9axY?;kInW2pRWE11{<_Q?M!@uDRMEWoFx1x``pCd&ri6Db-P6c+F6u)M<*>it z5*2u97|&Vj?Dly88<;;TTD-erz3vQ+=`q&ybR8gImk~TkhlUesMT166R2`lTDa&dj z9;oDO!6Dh!?HcAyzRS|iN&RYW1~wPuUz@M+TL2U_n?Eq43QR7jM+ksJA`=lc@+Nsd z-iZ02YzfawidjMGJ8pAtSNyYfCGVxTv$$Xc88+)XuC8q+&I?*%lQXi}I-#+kk7Vc|b zR$?lYgTvjRw5U<|OI&_Ope@mfVr4H|H59>6g|YZZ&X~+LAtEp?kK(eb#Rp+klnYj& zel+rx(fIG}!p{=#;d5=n1AdmeDom}bZfr>Q)d-8`?Uq?4?)r+DZgtj>0d6x2JtN^#7^}HET!|~$la-DaJqk92z{*dS4_+lNGCEU{ z8(+^htk{msr(kNK4z4_;ode_=p;XA0H<^-n1u$l}rn;-vDQFUtRuJ!oZ4pTuvt5av z*OKv|Jh0IfFn#=jE#cJwhiuGsrAuo4J!Qc}nUsDE2L(ZK@CDU2ngi)EL&R*6=teuR zvO%njuduMox7GP>pvU?x$pdW5I9<*D?MB97)b0wQ;^o98RsAYu4?Z704Hbul!(L+= zTPRXBqqXn(KBs9~!)n^GqQPJ6VGp+x1$|?g!I`BQBgaY01{3`tV}TM@M3!)ktJdwt z4xagz6Eep;7vT?CCM3T%;3C6@UfrfM)s^YRN!mNNg)D*z2uVo1OvGbK8#xR@KQxEL`f1uO}ky&F!n(@4x|gLD)?vM+~DuYmBMvBe1lMf9&FD_6SM2p zwr@aoLyVo*U>_S;nb$vM&TdwJFvr-Cd9|iaJ?9i2{%9ndo+yl7BtYI%xR{!5bK{2o zL6S_vl=vJDRLYN)%ekY6c?G}pe!z-8`=Ns#?e>Gdy?pW^c<$m3GSngK`}$*15W|2s z#2jG(!je$Qm&9Y?5 zTls~s)aDP|Xyu5Z-GQnn!q@^q8DG9fCZWpBAVrVOMZqvBOz+eA>=2tD+IiUh;$VhO zUWp-HJ@&H#?6T*4pdpoh$s-s$vVz98?Away=0emAha?c=$CbTN@lo|^3YaAyp<{0B z#jwJ_E)A~Ik1>q|j`bQ&(=2D0W5!KL1848hW1fc`(Wt^jf0rRG^rduQ*vw*DJ%$cW@1$Z<#KQ`A(J($tR{)XrUeuNgm~v< zJ1${_Y&d2xZGOCTL2>;&{_C~}W@S!36cYiHV!s7o zmXevd$44)#yCD6VFDqSEZ#_nQ=ZN6oR?A5O^*zD5Es$3>(Ys2nr8|tMm2%HGI-7Nv z_wdUiIMj|-U%AiHR7jR?kc+t`Fte?mD2s-5Mu3ithV_SKCy>j7V3ngI*|2&^;y-`w zk}UhtZ25ZntFx$dph(>*%w&Wxay1SZ7z&eDNhCY(uZU#^oo8JqqL*&V{`iTqDC4an z6_!()SRP?!ZiZ_E({fe!_Oyjh`lR808*j6{ZP>k`;z%Is(oIUnzWGIUF9@hym=1fWqBZT{gQ2iQy3F>lu{T zVsVPS(bpFf46N9F4n%O@CYlpN4l>}S8gk7*?AS1}b;x4xQ%o*-^T%_$TPj4GtJpDI zo`dXre_=K#wSj#d^ZXgQqmmf-cyS0zn$xvLoe0NQQm0~7qDXNgPpc>kBCZ()kESQk z1D*7RoBBsY1;n;B_(v%OaYiVBg;1cXMQ)~#2sax}W9*pBYr62^8nDcnp@ zX@8w1+L{J5)A-c?+WEOC8{3RU7m?;>iRZ z%NATvl=RC^o<{g0?rGb0uGCKGE_n5E{tA;Na`Vhp5DU&FgUSr}tzaC`45@QO6$0)s zs4htI0|qI(<6?UeyK)EAw)^3xynHNW47V8#8}P7D*b`e43rx zbP$BWH^HyRS^7AI?$t={uU0&jm1l)k;UdwjP4ell`ZGyVh|Mj*n{q!+%6c{F5`oIz zb94wl62E=u=;CVtELO^(x3TIqYkmWHqh6_ert}a!l#j+=4@S~F>zsEbaP50CwMwrU z*=iac;g(3spvAfxM&J zJZOIZz31u&;TA7;peSKJ9K-oji|^tW@fZ4?c$a>w%$yJLqg6Haf>uKEbGtNRW71XMu-yrev^_P1`|j!?5$Eyti# zMo=w0s!v{z`~BJacF(G$`;a}0+<2N|$z^a-<*@t<1$nE3NzK^;&g|<;8pQY+E`pIE zxj~?id#y&1sbMGy^icfDtHrysD_fD>xmbF7lg-uSvM-^1aTH}e0pGLdIaaLzlJc{wc{N}L)-D6E8Kl3`78`VrjRrS z^@V&munf5)B0(St;)diW|I=T5%I@^(J(EYiQqi9qYeo`;CiYXF^GbF2kYJhMeW9=V zCn2fQe15;Mh$O4Z@!g!3L`@aP6_H{*4U#&w3Zm%+A!=#mUUb_Wvt;(9_x!WL$2tn)NZIDE2aZ}YFrCLm*31myOsghD@CK%=FwbZYtsY)b&;>^?d0c3u< z#q{L#WDPTUuf**lX}7-yMgfsGJ-&(CyC*I7+gn|4m-xwBWo+g1GFzMcn@Q&-qme}{ zYg1eI$3(2YA2S^{`H2mVj6m|Sbo$S3eui zI_G(elDE)!DG%lfbb_2r2_926{mnf`ZqY)*0U=%&{TY6Jl}DX0|LXcpAEzNY3P`g; zCiaAcsLK49L;1pLZ2x5pS#2e;<88AE=bB&hY{q+$(LS(c3#wnCxq0sNETyL|KjKnd61w8Qku8wu?$uHYYAb2$v{Y(n7AnDs>^CTt8d z>9W~-VDn#|U{{2lC)GBcgqq(+t22iY&Bwr2hoIQKQOs>QfzPdYpMh5z=~AxyMBr1xe}&VGcq6O(|Y3TE?j}ql);gW)?c5zlIgMN>uQjJIFzSg zFy^OcIo;PVbxmT2*gF1fRQ#q)#AU9mS^^Js9dfLvkBr^}d6j|KmzGJZL4mp@8+G1g zr=T3ig6G#kh~bRnEN43Vx~I{=MC)q;B%@^mWi!zoAZ^IX=HPaxKP|Axb3}ge7ZJ|5 z*AC1-RFOebm_e|>T?rLd1!g)ndq-MLLgAe&Z&^a@hJz0(bI8n4@Yo~qC>aMy%M)CD zG`k1BpB8|fqr8an!_Th{i?84be!~%qEDAD0VH_*#hVf}5Q_0yzF*q&PgV%%lSOQBC zvE~mPfQp}Tt&xd~aLh zMB1st-lSesRl9*$Wef7PCjUWJCyX%~goBH6{Y}a7)~qUd9Jbin8;-JZ%yRO{EDVw6 z&ym>io4nVlYUSV0kn8F0x)&bCRU=uKMb+pFfDpl{xfWv0dROHLI|qT{-R_5(??g~# zplwhG{`DT5)^z)ic%N*BmI@GWA>`z?Vf-f8D)bDwq!1 zAzP=xN=w=qzrVFWXhF~PvB@?FP1sCI!DZo8l3f&jFIrn0;c0Lu88<|eF{X5^qu$ft z0FYH~rl_@(_3k$~XrqVGv(lGE<@g7WvKFX%SSN!^}guxLk`Bl(uig?yKwAk$-JAK7c0dV^qJWZKf$FYNmpP8oR?yNhp)rdo#w|X^P z4L7*Z251*B4cD8@P#TvqtmFHr6OzaX3ptQ$qa{hp(pnx6Ku{VfWY{Tp92mkYPR>)K zloPQ+ESdB`(@(s%wZT$JW{c$wOxi#%@%@O;_EA}ta~8#)%voc*3CXa;Hhrg`^DZ#)g>r>ize!p;!+IEZZ)#os+u4{5)-;^k7sgU~(*Uvav(E_| z-uyd}s#ooGht1q@TE68G!z>eFFPmYFK=m(FyZBIw^Q#hRE?XW`(C3f`Vc=*H1)rx1 zUhWE>Tb?Ej96dB2%hn|AvwKG0!mR=P zu(<7f5yI^eKNR266+Ga)0XW5{HLF3>TTs{=^ElRbp^~W9(QU5!Gg$hvdyFf#=ig*Y z^+DLZM4ETOW2w5|-d&_`-Wa7~?|P-!Mx7OrF29*DI0E~AJq8Yyaf4pSTKH~vyza2? zQbS0}$GE9-*=S3-3q8CDB-CTV%b$2U*x%bc^Pe4Z6@B&FIe@9D0p_Gmj@9m_*Yo?( z`usu(=)Y;U?`10x#uAU3P|RTBeIdpEg-A9*k}%G~C~rHzh9aANNkQlI)Qq5>CSPKr zd=!wmS zFG1R*=DB+US}6uaeqShu~4txzMJS)>zRt0W8wn2F${*x%r^Er!O8iNDl);qyqnylmpr1V~^~ombRn0HvP;!nU8)(f z7BH!Cm(SkeP<1IOOPE(5$a7ZrneXyQ?>fx`czWUn)YT}ijc_~Odj*nVsk*N+>eIAJ z(25_Agz^+b5S06~;UyyJ8d9uTXxwzkiO~Hv9F~jC9ehRD2QJX994-weKIygRT^6@! zUrjzzQ#E^Bu_ImndfZ&XIst zomm)<^a12YZ+VQ7vID39RbUFi#B5<2uFmy9j}2|0+`aK!;UIKd`3v}q{?(aFVzfg7 z+(s8kl{}LUGQ+REBU<+reWgI7dvYb@kW)HnEqn`;#n~}h?|Akoa&}Lt6g*J?jR)c| zcB}ljQ%WF4?g8OJd(E`j4t|(NgeNUX`@~ zkNfb>CSBX-HIB#+W9%grx5@^Z4mA*8tgOI)m=bu!IvHuiEVZjAJNWK48K=|q+ZQcG zmYv7NDNXLI0=h@>c%6JdBwmMqoXT2)&?P1BY?wC@PuL^CiYuF{#0l~A8zmzG5f{;x znfna$5dlQDOtLiWDq(lOpIs}DYVk@I$tCEWxKj0PTsm0s8fhmX&ObS`Mcf~{_B!!x zRbWp*vTl|IPKI{E^yIt7gtILBIswoX_s<8GvA;$+c?@iNjeK42Xx+a`Q*X`Zx#y*8 zzV3)mLbF=T0R^X&HwdRMT*n81k#wQCLzy}dEt?i&ZXE&6F@S4}lZ0enjPx+k1L+`i zWJC)0MP4s+^^sg}Ii1;#vBDMj(mTmovu6_yl*xo3$U-QR+xlt}=PXXVIg=QGCH@>7=+LF6 z^R{k1nN#_u#*7XYpHmhTa75f{G|bQK2q~0sZgIha9A({r@7!;Bply`Zx!Rf7l{s`w zsPsIvzAufx(6A-ZKeNZ>?zk0H zo_VQ4hE<-U#7YN}e60m3c9n(eUwu`c8TS@aq_LcmNPX<6SEYdbB)3JnvA;$0&w@dl zJCJ{Wv+VrYJELlcb>0hgB;n#Lh7223@DsUBo+x~jjxICx8(3je5GcuF3_2x&kkt6O zA|2l-WV$M?9}4-Cim&%=PcnFyclsIc=?{AqfMzN{ z^Jq2C#XX|Hc4rGnX&iz)?NJP_@~KzW{z|bTEu*oyx|xCLj@7K$6jyo&bt?PkCth5oj<(PX_=F|F^Wepzf_2If10{+1oOvcFr5DY zx!LbT!Op7Md}6V<&82)dN{#ChlnNNof<6O5OH)KwIbos<%uHZsJx-?hq-Ulw7QI*i6mPK$^B=6q1`0Qf_qXrFsbRQu z1~L8IAUN5@owwysn%83Lbo+ATV*>Q@1rIr;9c6msDh%nlo&qV)=Y$qVj_W`DhOBD2 z!hp0BW%_y+FSvaoGtd}WYYMV)vwLwHY>C)SUvI8yayeNHa7oEPC&TzsapTJ0aC^sg zlVi?606G$bnuqhnk=6*JBO6qYbZ^R8$U(6j^R%gjUE{n1uP@vd(VKA2<~Auv5>j-; zGsYRW3%^6z2$O;Qp>lbkF0JHWZOa1n6cVx`ekd8~ebr=)ffscF#EwZ3O8Im|R<;7k z!uG==j>{&OvD80_p9}$r-gn{I-i8uno1^-GDO~r*nqLBg>ZBi^3=*wN`5VkH&n?I} zodBGUYu%C@Xc8M~FzeCL&Z4ZukVB!zUU1M4?idaw3gXyKALg?9m3E+|-7fgV|j;yka z-EwMdgKiwu6OMJ1Om7`D7RU34WqPV_1sd~N;Y4fTvbXNhffSOO+)IJ!k(vq;0m|EH zz>f12`}gXb7)5eQ1g)a}i1PK>&#pQ?%gHD!nMcx`1X?5PneS^*0m5~v8>W0=xPgB% zLa7w|lW@jGTip(vu&jvS8Nst^Qi+!kl{yknLJ7Rm3w*K<;O;$^S7#zrZS83*=zB>( zPq+8OKMwp%g~Vf0Nia^9RMbmwYKyCk9cOe&w2_@BjY%>-ri>jyY6dlbcn0i*0tDzb z2S3fjp>1Moy?NT;kBgfP-mZk+RTfip{!O)zXm|F7hqIiV?Cb3%5m?47Y+(qBhS_bn z>R=^p%b#++$>@}OVboQ~B2-4`*NN?96acB%aPk28cpB`rrI|NEHS@3$VRoTI+A=u^ zZLg&i6`mP0u67tl;ELED`Z|iZf2ZLTxsZM@KQR6aShgs`VpL3nT!>rJ zkkr@>OsE->IK-e};gqF(5yT~qb8t~ev+(a_bOmAZeAl8IAX8*ENqJVm@hm(e2mmGt z5iJ+^DnxSjd?LhOPwGrM)e!c(Tv@2lKv)(-bk(KEXwU&XP4$7xZF6yR)NP%HRPjmI6uMZJEp?SPC1*yWJG>0)FUYWhe*A&>QGTF(k zUd1_nX%6T+z*<5;k8Xo+Lcuf)M&YtZ;K-ksN!g)DGs|{KRsDIy{yKNf>E*uFIL%Jc z{Ymq9o&6IXf=}YWobsg407~U_`^YBGVC<~3z<^mM%nu@6fb#uk;+1?Bp9er(!Z`S7 z_0fR2K<0QN&VP~NMw0yLH%Gs*;M%t6KIFb5e|@>#;eDk^Z;-R{PhFwQhT+`1_Tu=m zvr_SFcy2`8b#5|rr_5K6o-uO>_sdhs{p9M53EMPW5dwyAYX_50AxU^;0LOFP*F(vb ztef6J+4;!&`}vaM5m|0i2b!g zlyTkp-?);D2FU1kMd8m|4Ba>qaCnsO%-(U)K1)^0OPyNvYwi2A6M$D!jyrGCr7(Fz z%k^iKo^e7{fR$vlkcAS&qyHfnn}QA0RX}?~3g1VkUe;DnAhv;~wzI31)&5i*|3qi{ zhi6Z|o3Eul{reOP8P(;BxGT5Ed)-<+&40>>iSlA8jD%;9BwDE(T?f04gA99XK*_ifIsQP7(*zc6F};F|2heJU;w zn_l^-PL16dvgy{}Gj`vL%~-PO(!TnE%lq|Sy!G5l`h}>#15m^aJxVWR&+bAaQ!b zjNnQC_I>|)WM*J`fyH%Pj^KnrB?puE&;K+DS1~K!?w>H z@-0+7#rDC#N~}`VJO~)BAVwFzhR&~Fm%?O~PYv8|O^a+)^^$&=}B_nojw?)RIp>;WS}JVq1$1)5FOug1{#{ zoSA+U{bf;_zi_OBrVSoV8*it{ttowLE7#`1{!H`Mk?jy?hMGysLJl%+DB6c9)Ql6) zd@dx863A{L3a5-Y?HiYoEruy!jwkGfF05yg0M=)@LM#-zLu_+a^cKFfHKO8QT11&l zmQqIciC6ag`Ig68_2}1kmRrC01l2G6oMUmUxFJ~A&OwK)ky)HkFxEF?)L|n;7q!zY zi$%$6x)SxbM(63Q10Biw2Z`kkCVn!hm!nw_fDKcG70%lkb_yS++Z&~qE`1UdEj%k6 z0DHtrN|A7RO2{@4|2@H* z)iKi+Eg9G6%qZ4e6vyL}oerHfmS!$V&*P`spFJ0eD=UukMVNl|h6GEgJ&j_QGZs9A zoL!vyX5%uH`3Wpi19ZtdVV+=J<av@BZ_(cW23|=P%jtuR`EKI39VhkZ+x_6AyQTZExMg`0r6-peA^OsZLLOc-6m}@M?3@c6Yg|7^*$p*0NTj? zj|peGnZJB9_df9!lMcB@UMUw^H5P3-|0C>n?*fjXj!jRNboJ+T4d!(X=XH(db&cnB zP3Cn?=XK5IbE#`GC=XI^-b$`$6TF>j+%)Orh+Ry7c%pIQrI?s#F zMfghmU9fz{o3blhvuERtF$ZSmE_pJ1N7C|%Cd?S;FC{X*M(c=hpUDb6d}W60tBkP) zamZq-8XrufM_oVV#ghb|nE%kh8?o}|@)rH{ zT!?SEE{zT;OD;v#xtiDho%E0*fU#KBmU=fPDx^HKydo6|$M3J#c)GXk(q4M7JB-Ah zhBei2gHLLBJj)YPJ?ppXBpH_u-R;RJ!()AENaz+dA_Yg5MElnyMpiO9>QOf(F68z5 z)HzbDxNa4k)?JkND;998*KYdyfMLCk^n8B}ag28J9rh#s5$rHJJ%u%bn;!T@o5j4K za$=TOn@^|A)Wf|VG{Ms5;*U5Sn`gMY_(*`4yy@r%k%dorB6{gj&E&j}_7wa3-;&r; z+L8M!N;a0mmE2DJ-fU@{z^*jP0;1OE-RhFvDluv3mjA9NVmKg(;IuXAb+XHrAn@aX zdEBz47(5G!<#~*Zud4D zAjD^lVW#BMbzx_GM8W+Z_Ak$ex80(W7JCUVQ+)yD$!B0z>Gzk%_<_wMVF04`MWKsS zjBlN?&PVF9jTLZx`Itgsh3m*d5ID38n-VYXxf<3$bMXg36;%wSP^9&YgxB8uK&R8o3CcekFZF|R-Vag(me~6xo)eUILYz&7gLrn zB9kbTt9RO+EV?NQt%rmSwf~$3A?J^fr44f>M&Sg>&jh>wN@m2#9Mp9Qg)<|_U_o(e z^Vr`ZL&QhZj|iIs_htGF3cMn?4_m*6haK^;;O$|zi9&~z#nStG_raX}Zk`)Hf3;Yg zkoBOt_ZhaVBFY3N@3-WAVmBjgN&X+XEoZ8uD-=2eD;F2r|6IC26W&@3S<9Vok1?T6 zZ`l_4hfB`osx_BzoGJaF&?UraJX?aHo|3G$ukSheT9`R!CPWlLy(dQ0P)ujJl!ifSK7?*xCcEe#le6bR?bZx~O;GQ$v!Wo+*EC5n?xY zE~rd;)k}Mre0SLKIoYQcJQk2XCSWvjJv;qf|2m*iE00Pvku0yx&YhKkbm%Z$DK~n{ zG~@ih`}rf{@@;E*2icT1eA6f>nBF5p}xvk6#aAUq%xIu5s!t9l#b8 zy>Y)K!dtwvRySodwmLMovMw-Dc{x2btYDUr6uX1VbIxq^P_nx5S;$1F9aaZX0$ra> z0>>kn51j)5b);Z7NlD7*Rrs|saE$}KWh?VNVwP8wmS zTj~H?`p4NTi%jbHR2rvhwuBcp`@>PN>V-mqR1d||NBk4}lZV5zpxN)$d>k2@@rpvQ zrT9_?M*6WfdF!Cx%?g>y)D}t&8I)RaGS&B{eLmd(JeRphkL^TDoqLuIF6TaR3}9%=?F65 z$ReZuJLJk&N0(0^?ujH!{NP4nbEKDUJ=}Vv)sd0{a`oN!f2`uh^5wy1Nj1_l=EMQ6 zFonqzYq2mjj{%+vqx0QGw-Kj_6$G-zoTF3fRXR)#Y2GnrS+yImEKIYzs_j4_USAn; z&h4i32ES&73sNGj9-r}d`L(wB+AjoIpB1cq1)-@DcVFY@aW>AU*O*^Z7-m&(@+<~9 zaZW`RQY4p;U>bb9mW|Wi*y4S7O)x-Klr5Ijd=G7$_?n5tCzd&RK++pNp|I1eaNNz% zF1RMC^}K7P@`kvtzNaWYO_S{m?H)vuN8%cnfZMVU680ym63^Tjh!%(BtOu}gZ>25P zw$=P?b7j$LIC|!vHq~Gp**DrzPx^@pN@*F%n^WVl9y;}XeJ<`H;vY%JK0VRvsgBc> z)@-z1P1WU|=!L8lVOuHJ8D+&8M~PC~V(Ew4XI+JF<~j4!CoMGJ9U$SWKLJhXPBZJ~ zbeWt1uFQpRe3P68y-MCN{SHJ(3NS@wNokO#Md|NK=&r68VOfeFrf^+m^N2N9?3KvX zOm?E@G^n8?z)ePgMf`m6Ft5U?wBx#q^m&f~t7QM_4pq@rS{L;m_sU;lw)FY37iznE zYsReFN)7YD7|A}~*l?ZL2lZ1Eewmf>RbFeKI_Y_HsC~t}y!zsUL=uQO&zCIZQzqmt z3-bN-iw%o61^Y}y!k$+gGTpz*c;@rSLs8P4M^{Y7mtHOcKf*hFipo`b<3L%1?Y!W$ zY2aHb#HVW4;a+%X_HXh%S9CVw;eI&$ zGQG9tCkwyK{)_Sejv^qZ`?ttSZ^U3j9gn_N%L>H#$y?8}pL>uJ#0^+~W%40v7Da4W z6e=3_9fdsMLMCbESK}t1?%4GXaGDI%6#b|^=I1c2DwsIAQyka3hicJRJhucFNk|rr z2amm&Ap2yorG>Q2Fvo!y$nHK(&bNK-5K~scNu3SGIrz)uhYGaJIE}NoOIi$86bdV@ z)aM7n3ukS&kldRJG^4z^3?>tas0xjUeXT(tNuWSG%?Ly`)N|*y5rz4oaBNybfkIF; z|6IJIyC#I_f7~+ttRx|0`;#o-l!PJN80@JYiP=8Nz0Lr0_@{a&OKeZ-0~fV2R0WM` zF8Khl@(i&pHW;ufSK=>$vq|wmHw^dtV=6Ui){;_qokXZcz6-jT=Uj>JfnolZTmrQBtxumvQ5wK@PA*p&}w;Cv)aU{gM3ikCX$9 z+&M-j*@QYI5Ch74`#$z0gNX$f-YNlBnDA}ILM6ZYCEVY!g#wb)xZE-{79nq|zA-CC zfLr3fZ|jr2rdKV4 zJAXo?(|VE&HN`U9-YGifV`<0gotf3eKfZN$$cH~uL#^C1KE*R4)Kd$3u-Lu`>wCBl zLUHzSn&)yN5q}m;j(@s{5L~(N7l^RR-Co6j)D}?Kt1^Z&u4jk%mJ`2Q?4l*%?C9&r zYNpM>^AYdK;Eu_()eG={B)z~gitrKK9E+VNrdKsjD%?)IbnM5{{F@k_rUONRf9Gvx zb;4F8U1{2rW31}%d?WPizffDv7PDnV)}kgn?>uNC3XbN&6Os7LXl4{BoaJOnnxrH9 z(gDJCwBvQOO<(x8NfUK6Oc}JzsqggzTbLH4QA77k`eS-Fx{5raF75CN;)1_aitZr- zT2w#Qt-nWKT@2#I*K*~;yV7@D??c|zk0iiHO3o$4F1Ia0W}e|^dV|%n`b^sfiHLmN zD`wq4T63s*QjQ5%7q=P{db|3~CoG<~P=L}E4DVpi+((5eSx=ses8BAoFG>*;;0RI4 z{FamJZHWuoJI%pStnGwit+^9})%l2{oahpNv$lbL(?0)2wrIv^2So~X5xW2@kYI!c z74HP}r*kc`ozXt;6qc=7G_VS1wH4VA@iAw+KM1~|(q7y|+fMRgB@R2`u2DPj3VcZ4 zw0zNL5aVqn$EO&fv(2)!flHF!lR8M7Lh$6&^WNWZFAlnYtz*~Pq{=>Hl3y=^tgCH% z^i`ohkT=Dm^?RbcLE;)Yx!fu5kn0mQLLM8MI?}?8o2Z=n8#9>*X&Fi^;YfH_C(%+q|oA)Hj7RsN9f8U zkIEyj{1t0m`^m3gKvAd|{58$4|9&UdhGm`)^rp^eQ`&%G6TN1l@e!q z-TEtL+HVM^Rv(n_O$dr0e;B_+R%bS|u&{QauL+yoF{_&R$Z_1HmgZ(!_rl&!HgnWb zlI5>xE8J4Ftx5r9AUIs0I-G~F?!|7zv%ZRmZSbW}vGJh21714>UaAT`^VHjUJp&6^MX+VfhFzxWUr%TyDm2%O0gdH!G<(eg4 z9EtcctR89cwC&A>&M$Z>$nJ=Sc7Rzezyx-W5FnzCO=F3J*WK%33ZU46bZ$V8`k*Pa zj3+h-!Vi4DE+nmf;0}Lb`Y{QydLGuD3|sx^SupHAu)7a7bkeuLZP|VhI?YQs^i~5q zM``mR$Ef$@I9#}`6`1Ko-kzb5@J~cqjmvGoO4@#aVI8OqokL~g_aa8o0GPX{m$oURXaSr(` zzliiR)6is2fvV>xAJ~k_=Ognb0iysjW?92#G;7$3;h`7UVl+fq3ON;m7WO5FV==iw zhUSj*2=qR+F@F8E=)n#Yg48&Ts!LN#Py3p%f&TTE?9Vh;fWU}=yVP%Hw6!-2CJ461 z#ghiEE_dR3K6PnVS9N{!xSYO{aVcmS7hqLa&3FBJv0<}NjS8OT+UyFoeZ@n^>!-JUf&R2n)e`vBEZQI$}f1DGginv)3Z>-=M_N zV6S3jPz!_sEEwJkm+)HU4V4#@OEMZo@<;j}#8y*WrAJkW0-^CM0od`F_@41@q(=v- z*)d@MLM{`ssSC8~ahbz+rB}yjbDx9@BglW_oaHjv=AOBNw#E!q$ zU=z2QRuL(yi~?+YU=rhL{zf{alSZwfWbbaYricg!Y6X2*O!>VmmYFV>J<$i-Xp!D!cc(<1aaJ8V++7?)s_y)hT`D09IYq_g`SbblwZGH@%JXra_J6Q8 zcRVUO1Tz;W`~Q-5wx+E8h6GCYhXxG4lF*<->dkXfuXH>Z{j9nrk~Fe)M9S(Zrn1xJ z=lgVGM*{?uwN_ zIwMtXfzA8NCpHR;7PVOdX$>5>jFcre`rWp^?-8ID$HbpVxI$o9 zg87EI?wlCD^Zsw^6UtYL(%jgFkjl_bU)2$8nz5PFz<$4i4qg_mFD}I;SFZEBjy-YX zYwG5_%NSfVZq@@lNBFw*^DlpGu6?kXbO1w zCh2FV8R;5el||i}Y*>)MRn$BNRg+a!?uv4-pj8a|s_IkUu}pWm;Tw3_%I`v3r*GPS zHDKJ$L`a{_hTa5Rt?CzrYQU~A|5s1(1Z}XpAd>hCiai=stiW+9887DUaf%*$LSzIw zfdvJy<*UDRrP#>27b@`j`B|_H+Tx-}m9p#UWds9rbT3w-~RWrRro4-A0M6soQ{r>fmx4G^_ z;h&;jiKT=%BuM267JGJ?GET|KWy}52PGW6Xl8h0?e>&4=)W`AnuQoLE;t2^}GMf`{ zb&GzIXvaRcN96F9YnjMDfpCk97y&xJfX7tzbp2P$NML{*Z-20{sfOQhoxSjG(5%sh zY-xk=5L9>6=6WJy78CTVh`$90xoR^mJDE)7=l_R{21(>SzhrahuYMP(aZ~hrj6W-H z!Z7kz2>)z}cviqyKPq0oHyuac*F+9r71>$_O?WS3w_c=Gn*KU~v2Nemfd0_#fU4)p zTSlO>LN3k^Zn9E`KoNacfFN=@?(a|EBnZP@$fG|xd2EHSNHd5f!wA_QBI0aYJ4`OS zPdf1pNZFW;7Vp^I;r)AO; z$gKmyj0W?_*cB3C^wfSok8+ECaDa2sO$XF)h&qQ>Lu7QpG_Pki#xJ{pz&!>sg1*|o zRb5nZgsb`t@ecRdk;Fl7lNd?cmVBL%m(QaFUnQmf_HKb(k^dQ?h+X*wIuVkr$B2xf zoMg7$vHU8c|e!&CO}CnJCqn6QylKS@gY_s4l}Di2v1u}Lun4DD5Xi? zp9kc`4YMSP+$M`cw_zuGn^?v|zspjDlz{=kk9rV(R1Wl%mqoORJY2w{|K>-FR83}% z23B`He4b0je%*PdH`dnwW9uz|B3YWY(Z$`}b#ZrRad&rjcV}?-0S0GrcV~gc7Fld@ zcelmeF6Vsj`~COEeiI3dxNMXeZE5SV-Ck9B+|2BO9kFjXth2QnFwmi8K$mlul4ktXjObF#Y$_#FoSLa zu(f}aFL@F5yU?MQ{kq}eOLR0N2ZCA#4`aMP3{4f=;nbIbqu*Lzs*J$gM@1dJrcD#M zqgE3q=qs(ZD%^M%g2o2YrK`|lXQAy4M>(L3Of}w zjbXn{LHT$&%nZ1NK^kQsfkCgijdu@!0TeB?o9vjvjkqv#Znl&T7V2C#CmCIq^PVj0 zX1Nzz8r+OvxRT%C@tw$uB6U}BiR5zT9isLxM8n*!#9iWh@25s^WegWn!95!KzPljR zHs2FIUQt_@Lq+)JoM~`rs00tMlxOD#%I~HiMrBugIVIou&Zgh=&aoQ`B~s~Kk%zeSX4V`5#*z2TVzDkH;5Zwd)h8v znIyRjb?_BX!v_;J#`LfnL%eIPzJz7er^bM7B(Q7fc$644m1{Ez-4FQ`1DhS7yA6bGs{~zTDe=3vT?G4 zSozj33s=Dn16};ZGR$m|~rTbyLO}tHLrh*{VzP&>3J-Ahlj9GH( zU8gaAW1Qdr^NY{J)*Am#6Yoys?cu%-=KcM6e-q$Ww9&mbUfQC0bF;rbEc9y7fy=b{ z78)3T0%#)%RM&kww4Bp>eRF2V$C%r6dPNcL0<^9=rpegEO1?c;Z{%R9Bs?^_%o=vo zJMq#rMNXJ!y>~x*TUWN(*JR}vALA^3yz0SBul`L4J(?9+*>Kz})VMov^@EH3xh6}p zac;TJx;ALWvvu_HZ+V^_KkST=Xi_wReZ#u3ZxbSu%|07OrG#O>bbpLwPzq=0)CcgI zooCA2`}<6Q(e{1V8|M&i+36)v-g<|lPqGP$oyy%MYs-H{ln|qn=cg#oO7BYK?l%Pi z4yG3^&eAuocWEm3rxac))8N01j|1{%JtH6` zFlcwJl~e~H=dqJAxGto#adpPsU2~}Zv<~$~$Ut#nML!0VNpI08& zG*`Qm)Ctj(S@NM{Uzi~;5dAo&XQ@y#%b*?MBrVr7EZ7Jlxb_}GT%4_|Qv4d`|UxqL?*s~*&F?K+EK=$US zRtDnHgr7|D?{+gwj8(D&K141Cc?dxU6pRxvnk>2<)e{?nfZmb6PQqD{l&%pXAIuTr zcl!^`AXG((tR`vgMi79_!``$uWrQC>^b6t`mzt%RH5#G?YXSvfIBp=&>>;kFqaI?IfskFjN^C2Tti~zoQTzU>H$_SGMKapXP z1Cu=NR0K{)jFo(|2z`zPcE1yk3~c|>jOaq+{=9(LcCnez_70bYb@ovj= z3Gyq$gOO#&K*m-v$#<^b>%u_Ryl@W+hJa!f+&1Q9CIs+k7r*gr8*g*Jgh7P!3U3=D zNeu^we*Ky0{GBI*ny}iUj+jULIvWlK;pY_wa{8-fe;@h;6=C&mlHlj@xTy(s1<{kS zE)g?b>ix*+kO^rP$W&1X?T`GAd6xb>&A5C&lJdvn2^#_hi`jn)ga0c`g7}deZUFlU zVub!SJjMU06*r%n?Wb4qKVJU`d;bWI{~zHW$BrwDAu_7|U~Lx;WlXrtNFtcXIJ|-2 z(4xmm6vCWCSM#lJeaYV5!+lShc$+TH(%xY=pDxylJew|Cm&YiCgsR__V33gmMR@?9 z)!NdZcIQis{-=acgue+*j=RbNBs}fCCjMJn$^)j`y?uR3Np=7Dhr~l8BhymuL%=Mi z-1YXgi1Rx;r=Co}!DxF1P7-z4Y9JEVSY(}U`u-+g^BPan6AW)#@RFaG6}+mwBT`GR zp^p3x=VoF68dNa{(*{Lpnx_HOa4NAff1K_>NNC=8$Y@Gv=6XaI;_ylP8Xb|o{S}4D zBxQC_N`4WA!Bf>FQtSyV2$lbh!#85zaN-Pa_URXhwJbYdg4+H^#IhWL$oz<*(Yp$v z?6b`;sbmK!gGSTj5>JDEk7O(}gB|H)L30E6yiYeOrpBH+xfq&8fP9VI8C|1Gja;&Z z`A-ednUT4g3aDA#+(94Y?NboSa28{KeWL18&6lHUk__czVCtSs1pH~S;Q%ok(Hc)j=&bL09p~rl(C}5JXPhNL7sH`Gi?ddFjLm--b8!a5@ zaK>iz?=C-@$9EG_&|HTDrpdNjQ~oW->D2avt?FKI@e6zR6Ci0!FY*Ys`uk_pMT^IisLGz#P-pWb4pi#dF_U`I#E|aT??iHmz}0HG zw!nfAdy3C|&*50qhe*k<)o)0o>A4F39y-6zbZ>??%ns1MUkPE=c?%7|=0h@6eU0#14Je%fe`A%iQ4;}?TYBhPVd3Jw|i5V&7FmZZ9g89nn{c5&QH zi0Ne!4|O)c>W@sw9>o1_FDc=t*Yy=xJUzH!tl0+;#JbS)tfR4Qd%NqyD{6%eAGKrT zIKn1QyL{y_KqW5Pm2~bT5s*wo=z9uG8Y21e{Uev^7!;qk))-SY{8Zb368l+HHYCPo z4?mTD7O}%mg;zpCoBWefAE9evl_KjauPI&dIu|a_Eq=reO9FqVf(PEa9!>6NpkPXB zEPNXLyjd-#C~*K>gEBXPtY(x?OxUymdB~KgGKRKAv?0AV5BM?@tnpfcGa|uud)Efak#*hJyhAw<2F(KfFVy514zoom$^Au$Y)!wwUeZ!HRBup|b)TH^uDEDio8bpV5H#DLgVgurzh zJfNJFAs8#Dp%EGbXl^T(Vod<-wZ#W=T5Eh-KX>k90;TOlf!o&jzydpbAeN0bI4c{l z&k_TS4O9>V#sMU;L;?=l;Dd1iH*H+NSb@Ekm|$$c@75x~GFyBwo=+=!iK5(CRj+}!ir`bli@ z*RnA5J?W(VGHgD&f);d5g1LC-jU5{aWnv}jLJx1>WOCt9(iymL8fh4IvdEaP?pN~2 zZ=<2DN=Iic->v@u{_Hxy>(`57xb&IGmu`3ep>iREJ$K!uX-07Y$L}o_VfbQfsvaCh z7J=2wvH~v0(w`BQ9EvrB7l)CYSVqVB4Zh+{P%1BWeWyp|xiz+#;ydph% zUqO|oE+k5@wS_qBQ6-#H3fKpx@7DvGElJh2^_Zx?7)bKdXkHKGzQggMQDiP)@fZ*g z)5WAF@t-T+fQ2MiwsZttN`!YCc{ZO*<nf9hm3rN)gQzQ|&iByF;iE{L$9MuVBb z{;p64=RCq!7iDePddl5Z-9vvi_54HF&tGnvg$s^#4q%}(gEDa(ri(*QO$;)iWgRkHEtik6ns@(dxSV0GUR^<1K%i`q$r4Z&ryi#1^(&XZ z(3>d+1~3sTiCMO@&}^u2edQ}uw?e|Q@2vOwxC;sR;7u)+#aPcJQ-p#PA!S8G{D+6y z#4aIh2>)x1r5q!LeS5qioi=7wVHodb0oT4S9H4|uxO4^*&zQ!SfhA$40iFa50grz9 z5^oG4N6;_{IdAG)aJOsz@b!1cl^fsb%Me+L9k^aFt%cp-DzLiS=}YUadX|1c;&w_r z6VrzsAN{x78l;Y^UR=ASJtXZekn^fkuqhT{A7}S5(xxE`K>e<(QYSu*VEqy=Ba86f z1E3f@O?!wEA3<6c)oT}o@qLi*Jo2W|#xhPKzT}39 z$PfUPj`kM3AmhrwZR`STv7=?&Kw;`)NIth=8q8-!@miuS=?=@MJE|j+_^X$e_HJK0 zO0wAtdU~Fd0ox;T&Xqa6Y(Cr8M<=1H8o=oPmD&e_5XTrWJH6$dGI9aU6AHHhb7=Th zt8cCoPvT4dHTb8Z9)f+2I~^e|qK>#~SCB6U+))zaUx}Og_tQpw1}V>U>6vZJw(;pb zRCPimW~?{Q;~Q>veFZe5 z{2udtzAx+|pMtaS5O0C70Cbz0;la#hJ0d^gNgO{N)sZ8Jj*r3Hzj_ic zF}F|qSE|V)vv~M+D|K=1%Wy|mX#yY&nqJ_Krg|vEhpS5bMX$%4F&q-IJ~hId42;)S zlno7%b@TlUTM)Bo@ zZlUBO*0*0qPuf)~>q5nM9bceUDqB1>4~mkC$W8<+5@*mzt|>DgyC_pQ_*(%pN{dyk zwF?(`@kHDl6SP(^kz#6TN<#_YQq&Y;_Zn4i16U5T(@opiDgNTc@B64 zFh3I$Cg$?iPJ;7EeW7kaR6 zZuy*!x-axut3GU+ZW1`9T&@R5K%C3?l#oCnxsLn$$1G7b{w2S5C43kMT1A?B)%1w#t<-DBWFcJf;Bz5%D*=>45EQu+AzTYpihnKILDtt(ih~EP^Tmg5At9IP))Q z-T^=}rLh}Z)toOXE!vFX5pv>Q2V;#oZL_wSb!2;5iB&oV8+IG9<1#RY$6FH3BhNgY zKDw=>eQn5G6HN^1xH*6eD;CP0yau1=zj&QiYXSx<0^KKhbAlcZ9P3R*hy1=h|NB!= zgu~}sP=vwdZkSjkhPnJ4{f(8`NfsYW0u_uYt$Mja+;^M*?K z%AeY}cc*|hN(v=(ztowd3;|3#bxhTytk+8y97-q_uuMzDPmCB#GkYUZ{v~}cK)97( z(;IohYz5A2ygMbyHSGAw8mlTIksyU^+azi9$AE3L#>@{pZQ~e5&3o4RjGTU6unp-} z@&)Qq&EFE%(VH%lS90??MsWd0;8WYielRQ}6i7x>{Lr|eU)Zhbb0)M3gOc--#T)zVL&QDU<=iUtqVe9ljk>)5rh$|J9#CXmhX3k)l^}9$fE_>I)MkjUgW+0T9g%i^fMo*mOcRo+=Z((Fy2rq~~R< zYqTn6eK5SSG?pTnc#?_o zy8$L`>Tjs~b{-h6jh&_W%Q z0>~&Hr6m9PN>L%lHUlFwB%_yB-BPw+iQcC8d0Pdike{daAbp!SYuYrNtK#huz(85K zJ}Jy&`USi@vF+~6U1rl$s!IQf59j3kgP>|#tIg>;@~f)?$A;<`U(K2Gu{Zt`IaVksb%24Ne5n1w-3+y0MAnzB<34f#{g z&K`kmt&3OaFN}$bS~i_jTeEtyhQA)(xB|R= z8Vqwl=g+Q&wOSHx8w{0gznaZ`$EsX)*HbBOJarcwD0oVebzgl>Oq?^rtQl1Ry=-@)W)7Y3_s-O2z@LTJuA#xYWNwp+%5kE7^dC$#>{KCjm?ir zEW<;{2-E$5x4*P3$UmFtJGQCd5_AUhj`+F_s<#r*2R`aG*tTaDv1Mjw;VkBzF7iH5( zsJ)=nJ#}|J;$!Ip0oMo2o;j~~0X=WLxOAl!J(<`alW|vz)=&@U&5t&S+6ACseMWRViwQ?0E{)g1p=|D$4d+hgbNS;Kbnva3>Zk}E4WTmp6{*wKe0F> z$esfX4#Z3iP6*7m!38GzA~hWyddq|Gsll;8BSQbj+}6qgCMx$C<^L1fbFzGf_U!+7 zgPh>O5!-Zwz*ynIIR1qe5bT_+|E+D?i2>sRd{%s-3#?D9@SjF)pN&@V!8!g#7vP-i z|FNUA6@CTx0{cW7pU46P6$DPyCPxi!g#*sP`7aFN{0D{*w(SXl&m)8LvVgWrpd>+H zis0;ETp&(G@B&)yf1#9+(Eo*=IRC|H|3y!nylo$f;MXjmv53#B`ll=3KvF^A_~2Zu zZDv8>R?r|w4hU>8p0@1>@C-0;u7A%C&i#492yp7QwkU9Puuly1&p2@ZpPrJ$fKNv! zqy0}W3EIdoA@)fAMPLx@-2ZftppAPI)8Nt}hO zWu?pYb(G5ogtK};6Ui2amBaSOguh9a4vLY~C*Qs&eClJosxZUj$KR+^3)*v!)Vq3W zqRYE=qBGU0=NjRG@t(-3d8x-)1k8_|G2W{|lf{XTfVV_uqf@%>^cubS=C5m6OzPV? zTH)Wf#}?y%%#<`*I5=PT+m@Xah9t^%bRKYGdurW485(;4vIdS(ocNCTz-SaWe+*pg zubnKv_&8WCb)LJFhrNhh!01o!G74`gyud^>(yGzmtl(Pc@hoZY!}q<%Cc8SR6 zHa^&?8`uH5`NiMBK>n!BJHSw_A1?6Nss{%AOiIrsKhtQ`cbvYq zQg>|DN>zX9!$m=G(D-^p5QPSJ?2VpOUs1Vyr>hB$S>kYnA)@aHg-|7gjBa^v2<==s zjFV)6` zw zwcI#wFV;#ORH0rI3hrC*+ty%LYt2!bliV{2V41}M{fga`_Xr5)C%$yL*cV|ZA#8oP zDThV&BZecVY{JdeLz{>v#d#1;e!p~Jd^V-9M(&lj2*&;1{Oc^3V-#J)=*Bz+b;V+f zBHbBBRf&_d*~4t`Z|8%%1D6fa>fpYK7j^x89$NJwO^>ibky^!}X6YVg4LqcSx(R1K z;KldvZ{sN?5k`49yWGa1Kys57Mm4=TwAt=_tvf`HHtn18I9Nx?tRVKS-fbZ|)f|_; z%BwANZt@RLxx(l_RL}5{6+YB-`(cBxwmWFnYHsZvl{UIo1;DB&ASyWX4HqQSO71qQe@B6{tZF_)~>QRF6dKfy23Q z8B6*;E!rVN;%6pSn^4C)2_u1_$-c0eY9w zR_9VqtL&r=t#q;n~^d!Mn!+Kb|s30oQ;Ze>O!D^J_tZw>I6gKlBOeNbt4W^ zt1PsZB%Ucz+m!y{cRi=8!KWXzR_2IG0huB>S;9$y8HXjWQ8fg~|IN||wTz}rzuC36 zr&@oG-L)EQqp*I<1SAV#^HEId2e6Poxf|3%)FR5XLL;r~WfRIj8W{OiJOq=B*uLtD z0V!-dK-qX=ny5v-MEpuiaPc}on;oDziW5!;fhK(25LqqWT!|cIO!4F9k6@%9!uSR_ zKHdXBH+tpwWC)J;4%aqu;3kEt&VJKaWNw?3s>p_;W?bcNfO^ytH6#RNG$2+stb&lN z8>21SQpkIP{@l$Yx_f=EEATUw#RyhfZe_RQ9s~kxws+>=MmMT3Qa<4Nk>+~*13CHt zXf~ooGu};!PVAN_WDjl@`Rp3r@ecs4;{!WwxULXFwQA!R4-|NWE4}^%6Fg)uX`Tan zYttf5E(R0r57d$^=Z3?VW8Ac*8qnwR`_37Sh z{@ntG9M(1-qgP|UZr)6fmlq=l8}bMpQsR2Qxe=){pw;1~SCxZ@*9>!2^!m`UNG_Ve z9W4^2=mM>-RZ&_^4lu7l;d`4H1e+AO0%qg_GxF}#s27pDPEy&Y0f$3D)DR&yX&&=x zup@c{YYmbT(IQ#{1W$Qg8^@-Wsm0!G8XXm!G&OXACd z-$fKS@%$2%iGv8`2Phl6pi@2ZuT5Z4tS5~Z5Ty%Itu{hUtv7_gt{mSc90x>zHelM> z&>}eL5ktx`OifDNcECz&<&UG7@K6~$6Q<=zVZ;yDAwdQ+RCdAQ^C3{HK9y z(N^43u^d9MNV9#$b$C0TGZ%W{T>hjDyK@i1TdKZ)3b{k|OfnS2MTV*6Rb&Utb^!kM zu@Z&`)j_oN_zM@NEw!DfB1X@R4F7(A~K0*yFJ@*yM!?+U)fI8DjLamJAKi`=j}6mF++UT2Y%F)&%O zM`+iNAvWTdlV2uOOGG_;sh?1tV?=o|tgY~bF=-cT@nC(ca=6V9Byu7icn@WT9MxsN zZNm1*bdLdK^Qg-^g5{t@Y+_F86s<5H_&6>dwWtbLOD`(T+ z5dL&nF%~g*(aGwdn79q8->BaGzpdncD0I=sS;zU_ap?Y9WW>iT32X;BqjxIy&9b2< zyVw)I0yxB!5;%zu144U1qz*ay1C|Lp=oi*eQZb(9aS(c$w#{f97Xujc8AUJcroX3C zUtOKq-Z_+L3(;CLO37mG99b3lv_x8JEQNg)R-&>5 zAsLysdr>z!?!;FQ-U#UXsQNGC#LQ|UU@M>XQ_OCEEujaaVFQ_JO{|Hq0Q&Blqun0S zl}x8}9uoxyLUxxT4wXZJ>E0ORpbFP7fFQniyzuldT0+g$wi;Pe_lPr^2gg>Zu)*2Z zhVMTgPy{9o6Ws9-WPWGZoRJ!PWQuFEz@T>L`~)|>u(*pUqSSw73ZH0`&tvBI9zfodZC zxcN-MT>&9J{gO0J0t0I-`h5(zXv4ElZc($i78^FBPOoxm1bEcHa&9^<}M&7 zj_J#7M4kAFH4y4M?NdyaJk+Ly*c!GYH}81T__o-jd;#_N?fo4v1n>-js2=UxVaza3 z$<@>ojJmk;u#BmAG1Rl&+Z1H4AwycayfJWLrxlS2_u+w~IwslTyv#Li3BO`H*ixYe zQJXgFhdy^l$$5fxlN>Bz)(FZrn9W`KKxvQQRVfpahuXorzAPozFgq_-0)wTNcp)_& zG*?}}&ys}Se@~7+1bjg6VxbRnWR)@o;JWe^?-o3+a4zeoJHqSs6wMLNU*4#?!jXX^ z5n5cDU`o78#yA;3>T|sHBJebS&FsrZh-O}?yD>SMlj>Kg`y%}{W0aqfQFscdL*kDV z5GM3fQX89p7WU76RXx;`Ab!5Clu;HBSV zBhjfCq6aX9?C=o|@S$yh)Y|P8)-WJypotU)JZ;aI zxUGBYe@zuW9M|xytPW|{S#|7FGfm+i*F!}C87wOdEWi9iZHaIuawE&$hbLNVWsVGDFt z8_JLb((V=F5&4!(kwS3wB{GKb<9Tl0%~=;SjRzsrbV}XBV_wT@*0(Y1Xw-{)b7RB* zW+Nw0aXQ}AyAo$ zp*ZczXZ(@b?@}rGTn#!)*2jwCP_Xl;eP348+6RPEJ#@*F0&kyCj;645_3)1De82qR zyeV2;`UtinI-bXkM2~qbqEO$V1=ht2yU$#X8-UIg6sfPZDda%eyYM+hkQQpRUFFU6 zoHPeR;Kj0p9uP63`r|Ch=_1}EyC0|vXwOY_|$;99h=D0iRS`cV$h64>%Kh9bsbb&4^Hk>_HK(NsrSr#41)o}nsd;QZxPr~P z@wI=?B1D=Ju(^_>0|2pB{Y=|<2~roVlErm&!x)g&CMc{Bgahyrkb0$T>d$JTA7~Q@ zkkndTR`lHJbUi6U${3Re+xhcV!xa+F;7uyI(N`yBpBMqSamJ0!c^AY(4Dhgh$1qgXU6-tGKL9OXG5uQKN?HG}`)^jDvdfIj| zf~SUE5|@R*CWDK7kF=Q} zEvrajHVeds0WgsMlHSbE5ST<<4iELJl6uo`3o}K8Fg*IveldBV>vg~amCdEf$06mh zN;7+9;7A*7W?2ib{78SVaFjX}w5Xr3xPF$%9J?Ugu4pb!cj@sxVNUF8MMAKh&2*zf zC&ea$Vb!o11BdfiIDk$*!FzdsKXG$iUT6lIyB&6R3~&t|SEAWbQ+qy+6FJnSjH+cc z01dE&XeB}+BN*P6{2LsLYq4h3z9DugM!x6MC-KczI7U#q-OcfUC5(@Kd_o<$H}|q( zJ1++xy@u|0>rL$uD9Z8AhA?+G&nr~v`od)$ar*GUlhtpF4U6+jFU0s`FsxtLzaIv$ zE&s}JXqAzmxl@$7FcYi9`PzHK&o$`3)R0495D4s+-1@DR{ zd1sUDO)B=)wwVK?Tksn3pAyWujTOFIZQmCLBm>@>CYU=8;^x~KUdZk4a`3@x{IG~Y zC5hS-c9)lR0|WYrksZ*9<&xfwM+M>w~>b1g3RTnd$n3}KTVf~ z>Hzgtp8l9DftLOoQMEPzGcCL2nM_%R)X@=kPuv;iK6>M7zeA=EXZ}s3=d$h~_%~!X zuB5X)`<~9J>RNE3Il$kK=IAMY)_WSX)%UjW*FRALtLc%?>H@w?L^DL?XEi12Hdj9} zqz;-bzQxqJc(GDRT;q-YrvBptfgb2bMh)=9n_l@^`1*B>p+<`Q#&b+MfcnS4O?sM{ z>ih{$BSVzOR5{wn5xwl7v%jO+BZH(;icMr~UiWFrSto+E(@JqJfO;!FxT>1qJISo~ zqMF?XJze3BEU`tS>=7o$6{S@yYd*|4AW)QJdGiv01|6CQcC+Nw$^U5M?Ch@dwZPZR?3dW%L3wQ)Z|koQi!=NjCBA20U@UdmdC z`u{iio`dE8O}_t6n8C`$`Jc2ix!QV8t6Z3$6d=@VN(fFl)j}q})AZKUpaMIqW|VVi znpS+j76v8PtdgIBhmvycsm%o7QnzGN;fMZ*hYF?^p}-3Me1>q-4=X{pd=7^DQ?{qa}e_ zlX*0h{71NSz+~=%%e*?#H3I$UNf?QFN{Ce?g!Jf+@1e)iBQL|E(Y6tz<|`MDVCYa! z#Q}f;;Pc$ye$|pvt?rZAUvn(Vo0n2fyxw5lAr z8J~%sPH7d$N6qW8ojC$D=5_}QGq02AcrV8$JI9DyyPA+&wYc(GWE(|A);j&s;`AnE z-r0-fjgr|p{P>QM9V-6Kb5-ea+$^rY74Jy!m!%w}Gag2rvZzsAX307`R_G1B5a`n9 zj(=8JZa5gKuFeUUaL|*_F@^h*W&BcwpOH9Hc)##3-}P7J2BM8#6}?jexWJEW zssky)fNp`l8yW?p;-E7)ub3<0E<4o`=>V27(#=y9j~w;|36cAIc$ie|9InO+IC%yQ zD1E$jEy{=t{^k%oGcDPCr`z?YH zCGh7uI~3kSbrdE5@cm{7*P)&Z9CB97)`jm!y>~-ELeUp<-KP-7hb{BGZ~Z(iEfjg4fbL6FYB!&;)wPl7jSD9ol+M6ps$=Dj zE!5TpuR8Al9?sblc2!j-V{+XL`GVaF?tL&l7VvPJ>W$@ls4kB>g(n$& zXo`@Yg8NmKB=f5$BSx3Vnuq4MI<7|J1sus*GnDUu%vgE_Y@>tLt?Mgu%48yA50`@= zW8{i^9RGv&+=k~w1OF3@GFiu`6RFQcwy*ffs@&z$gbE57kB-xo=jC^iHO40kIBJ2r z4ZS!>qLf8o1J6FBKKv#+FUt;6k4h$vH$_ldNfPPS5&_+(_BsRW|OtubR#ZBPUIEv*PCJ zDJ5i4R2K-Iq8=NIh|z>IA+abPw`$!r#|xT(oM89{{S-^4iFgjhC2XzfOj2yBZ_t51 zLBW)Zi=e!d2`L9vy2;&`#KT(UDWdFp+pizRif+Wc=H$MgFRy_+%7=S+gzG;%g6roQ z-6r-cDtH|Zxc6>7hGBO-?dQAbejzIka9Dl2FQ2RE)abub~6f*{wjd4+ojEBxj4 z$9OCvX(Odz7xb@oo4%Q4%vqcf#J{nj07#qkQzgNRRXk0_LS}A>rFc8oG@s8WmE5*g zIkXnc$p)djVXNYbTT}V2vA6qipoSOVW1C4I9vL8C+IkG_CD|m54~-aEyZBFfRW~l^ zQLOk$l><7dYV-iDB33~VijzTZw>M%$H2K!H-l5suA8(>eENIVj{bMqyPkNa-fVXEn zMbhhKm13%-WE)s%Qp z;f(@!e%ktcmCBuq+Mw5135Z- zs6Jm|`8K1CFA;+M?4W2g(s{XS!g+JQpWc@L`3~5(JFoY*?fK33*QfipsvlitsjZJM zCoiC`*Iz(CCgG+>Y(!C)F?3fec zY#lGuY50DpcCw-D;5`2#?-9x3uSFO^HId`daGqs*-^zAG(;0gtK{X?HbLXP?i(!4* zVQSf)D^ma<|E+VosUBOv2wvu|O z&(vh;uEY4k0~)KwbNu$65X0aY`Mz_PqnP3;cn+WzRj)*R`9+i>@tSKQ@oOO9ZxJcn zRJSxdk;jzv0S0@2meov^OcsSrGs-RX6uG;EzSq)(*%>v1lyk)1O>{Qe6ND4CRNlPm zRlU^|liSy#7pAwo!wq9q-Ul`LtY0ubvN?wpRLv$LW3(5M?rRSV&FhrdK7@WGD?;}e z4(@;oHZHj2*0Mv0e2iGhBG!ek`|pLDjMdvmd#W_w$>N(_Xw1KDroZ z3z(M7Ko$TX`^c{toQqIa_K-wG3tkQvI6wlJh_cZ6?6)t@wgnkxVEjs`t zj{A>-z;&(P&2OC09Ag6FcO_mUmdoC0*I&t=G@5I0%wMwrC8zdd+CRs8ibmm!o{D#W z#ic|;v6mXOz4pfO&Alf-+{|5Zv{+>bqb}6GA>9n~?6+8(eV6Wva~wCqwGQu&zY!r_ zk9-olka}gYnm1bc-~fED0k%-xz?1@#6WYiFQ?B*I4L&20=CjC9afAS z-evXP0J8W+@zHY@7CGtpSizWngA*&YG8Z^aw}FtdS_S4$8mA^Bifo9@Xu=&N?@GQz zR~R2;jP8M2UD_G(_sTCUn@O5g`KZ zMf%+XeViYvTfR<1aEE-QVt)a_r_T8(JN zht9ZDyJ5W81lP2DiH6u?x7NsqoUM183r3%z#*vu*$vq7bUYUJveZjr`zKSTl#tb?i zCo=z{1(99wY4pu#M;@TuEa#MT2(2UGRfT9zA9Nr0*#YKBk7y-$ebAE{`C`*Ae&_WC zIM;5^n>sflM^@yN(_Rbu;M0+hKQ)r33kc?mp)&h>hQ*`&x-(Re z?``Fse+kLQ;4L?3O^317Q=C}15b+iQXX_LAJ?QY~T@)bX%BfFCM2Q?V*$Ge9HMK;E zWI)#;e1-xqrq6ri62-e%rE&2`)}>$E*Ve$N_Lm!-nUu3)57K854sfLs$%z81xLKrd zc>7;j&|E)XhO)f(!`KkX5oesh?vSX6pRiB(_?}?MG=+RsVZO6KDb>vcp8FkzGq%;9CU;mPDixxu} z(t9V&EvQVCOd*Jx=@QvFdK3+8qr-)?_o@qQ;Y&h83gGbfg^oQR`lhMB(QP^+N+F>b zw}OHvH62#t>cdikqxiQrb@4WD+9W++M})PMayJ(OEl#Is2bbfH^7!w!UEdj5@ne^7 zSY*V^$I9(UJ((!K(_h)Ev7_o)C1DUCFJg^FGs+OBwT}#9MlsT#=nqC{x!2g7URE=zep;4fEnafz#fEAHYtH8rqku!sv;j#8c^p}n#f_-U?bsCpha$)p z?q@br6~D1>W(rWW#*ZcU>4Ihk_314u%^WRlYw6&%+t^iP?e~8N&nqjIHhS&oYR|h% zMTHGH(81rE^AIK4#l7!vd-Eb?qD1bELv7uHyS1TjIHmOMn8{y7g&_ ztIfssO4drYUzMC+ld0FdI(%lEEnP(5;zl?pysgtKZzKPMYgM<4UU8eXM^&`4pVdN6 zZI&!2)@MUg&c3QrSLgl0``U0nEZd6{0B5$#e1@cPdvV3G)cH9l^5O=nL`{VIT{Z8mn^)8KYTz>cYJXc#5LR|m4 zGZxG7{}~hq&d;CDR{2yfpwIubw>EG+D5l;QH=xw*Ieb?|eS#%aqe zeg~&TpCdR*I4x=$WQ4i~#^_KLwUwyLq`4O1TC}RzINY|Xs8>IB*`Jyvxv$@*eXYg{pmb?8wi7%MA{?u>OENlL$bGqRk2W!ou9 zNSjWx%}W)hY9h|CR=S{iIlEk<@zsVZMNE9pkh87IVXW;yk%-7HTcoM2FG`U6H=Z1f z)`nsRb9nOCcCYRdA^c!OJw>gN*-$yhK$DYF36PDJ)8k@_OKT>azrk1o{`J%7)QbApX-lOQ0-HfE7ca%AMzn@Bty|f}89Cc<3 z+8A(aG(L8hYe#rfe1ZWcGjGEFtX~HeVck(zz3KL#Es-S&!K+5@@?Q6S8Tc#Cu1}s) z3|Mf|5znDTlOIZs$&A;Hmir%L1+v^?9J+2D8*&FcZ?F zq7S}PB$cnyu~n6`B>xFKlZZDi-~JTgT9nsQ&)ug~cQYQK^(o^wLEZ0UN)8x(E^>jP z&&BwIvG6?+4%)7nf9Nu)%<;}qL19?`4zT^vTY<=GtJqvTc!$-cy6g3Ii-t;QjVG&( zhBORhe3sR>@~{&O+NP2owJvjp=8f9>)4H9ElCYgH7vYvfQF4AKr?{mi3eR<#nsqOW zaB!0 z5vI6fz8#M+#c@><#=V0!>OTeR-YvYJ4>pK$N7uHV z2{){ht_q|KUvJhNy4ro(1>R0Xrkds^cYLmdi&^eI2&BtSXOM})lmdrfJ)60!OG};oa@9qrIgg*hN%1{brFQVV^jW+-X znSAuF$$Snrb8;@GmB!b}Ws!x@e*Yk|KK9y{ta=OIEn4cE=i4DrXv%&LNd+Ce_it_b zyoUTs|3;rGX!`tlqVK84(Lv|3&9j;Rc1P!SZGWQg?ykLaOUyw(WooTcF(~UoHXoHJ@(AYZmOky~7P5?6t2yEtkBl3L#30-~yd7Vr3WVaP)TS#s8qe*`X zGU!V+3MBxAS3H236UMbC+pP`i<2~OibjALZLk5~s_E}ir-Am8W?g2McBcHkum4$$; zao4qxk6HI+7N#Ep!50YejZfEPpBPudpZH15L5M%` zA}G_|H(va*K$j7=C@19f=?jwrb=GYKjkgzATXf=1JaoUl>2%4F{63A4(uOg7zp6eW zszK~5iYq92bfxjMSfjhWeeS)^*B?20wVJG+@_Srf)WS~Xh%*rUxHENkvUUFzrG)1+UsIpGC$r9ULL%j4$gELyZoN8pSL>S&i-XOG}XL0Vl`<#g66alacX=4 z-$CG73C_COI?D8)fISgoQ?)I)E~9tE8nySRZgloLp66w{^))gw6l6+e!B9I)>u)mwYOQ9aXD#lc7QV^c-mUk_)YmL=>jJF9Hyb68Bpnm_8x!xYVjk@6J=)e&nR5Ifu**R|TpPa7z$P`9G{c0S!1${ZD=y3J zZOzB*jfcuq7F_kV^U_(z#Y`yn(3@9-G3{`|>z}Ykl2FtoMXVUz3nbDxS|yONdvLoD4>BZRZMf4_P$9n?c*e(O6sb8`pz2hLmxyCVLC(8WEqKg>1#nACFu9&}! zLSB85?t|bzwNURyja|my4p?*o%V8OC^oPuD0He|hgyff@{(2=}q3v%qoOe5bV>LlQOrca?*c{Hw!8Pi+wCK@CQ`1p_ZRXqQG!^7OWaY|)op z%6xr|Ah~v{;J0&-M{PLc+s7u(KLqz&ZG}Ww3aeJqwK-IIT%T2I91x9 zDmN8ewke4N>TjZD(O|I>2gKs4w6OUV-+3NimFg+hIS=m!eaF>tY^}ui zU+v`zC4SKXUsM3{!Q7p+!0ms)bt}v^LFk!(hi;Z7RPpa=Xa=(NeMQ=qQ4#>lsq0H>cP!(*>CH-DkcU)#i=9I0dD#4r$ECcZLAUjmh)FjQF!-la#@|2KOT9ep)5y=e{3Sjf3( zWb8c7QZ^Le>^@r)&k$@PJ6spS!9dj&3|oJSoFp)8(M;C9yPqpvfF>+4suQ@5_p93F zURrslAW2xeybM_C+{K{4akto%_zs;|G z7$VdZ3y8m06rx%^IEpHKg=2PhLo{vErPsx`T6*FJx(}Vro&1gV>%96&WE2{?jr8@3 zW4X;n_NDL<-D4m*WwHd3L@@jZ0)lJfAQSkpnWMziW$c~0-?B96k-C#zf(LfG)%X1G zk}r0_pjqppAPI)cqafFhR|dcfqL4|jV5J-W?fK6R9Pvd#{?L?O+rZP=1G@!j8Rzk$ zX<{4$LTB*34IFZaM(|K*5NdSB<>Fu=@DgTfr@zccmP#More|o`Z_HUD%iXxoORbU{nJ6ce<&&}j!lbm-Ddw!;*~|< zdQAprSkbpl*Xi3eLQlgd`w<02C!uuAqjqo_s?ZqRShz!_$}Sa86YllzDy0abLJNR# zH_T7;HA69+i2pjwN|5ZGxftO(%$yh|r)n^SkAq#mGk`_~b>2GcFZ6R-D2*vnqLMZI z6ZiLDC_-ro&HCZru1RuB#Ig<=rXB;@^r9Ybwh(SrH^ml^kiC{}OY*pQXmO6!xsCH| zjzmRDRFmZHT*q+`TzPZ%p?RU(7^VTIYFGj`WKFwTM3P90Z)?06c8F4@UTP`0bM+VG zrmg1<1FrTE^qx7-g=9C02TvUjLaI4u>y%S$@2Ksc{3v1kOW=w*YhMzNI88=ez8T-F zOMW(X6s48RsW?%m48%y`*PQ!0oUWuURjn&81SZwex!;oEKI+n^chJJ4y8qD zJ`2oN9SWq{V5Tp*`u$VkH3Mc0gS|p8icNLZADrbVyLiqH2agalm_|K*2Ah$Mtwsxs zJGV#RPjO$n>*S6WMnkHVuG<{7COMr>MPycGL9$N1DU`cxzl4doFDG` z3;b#92Y326Fc7f2s*lf{6GwpsE4wI&)EyHif(5%x$v?v_JaF}b5*4&#<`nK_ThHB3 zZnCtC5^4n&1?kL&trjIVy+n)^7H|8oIkTI>9Y0t}he%n=0Z%Ty1`oket&gajF{?+C zSEr!?CH+*433dif`wq+_mm~PmT}QqnlA>b1vc$~EKsRR~3Y0-NK{{63hnIP^M~Cpq zFYdwfxuMR9`bz!^gl|>17|7y@3NDp>EtMS$SAwA6-lalFL=)1+NPaZ)-Zk~1i3J{MhFj*sl`h{R$pjbs z-aN?s<{4X2g<#~QMPKts=faUq9u6Z6C8~(}nRxwBh1`{1O$MH0Lz~frX6mr7|K8d3 zFWEPu3Y}&oJ*qrppc!H|p>BPl2G83wF?^geW@UIBibt<5P(tJ5+tqc}*sW-1+eKC| z4&+Z~+>6e)#RS$*3}E@j_h-M3?}j_{8nO_T$74A2u~Y9d*LMUo|5d8$r)nV}xDGm( zr|_$h{hg*1Q(+>=4$vNMj5Zi;7M7Wh6px}VBq+@IR*q2Gz|ET_G8i)-s=f$EK8>eD z;()9V6cZIRSZ(T5-0Vk+e>ifixMs^HPE2Q&j983D#REY{DP_?|GQt0PTzuUhSAqVz zl`xD-JT)UJ!K`(R3BFIjnl?Gx>OU z4j3360_U$c15IlKbS_WE$W<>sS%NvIpf_~>g(F0T<-v%lYQV3S#3lRVrEQGsQmNf< zIjpU7sz699&xHkqw{#gq7xTIpeTGSSz`D*aK1wFEPfII%akOL%_EGcdNd*bp0 zer|LvzXTf>#BtBD?IodgVS9bgLc@@I1qMelcz+gN3lMJ2)t+?@?CJ9U+;I00G|qC~ zj|yC5)I%Cc-RHO%0l~?4n|fo0CGCA&DV`n}J7_8SC)ycJu78@abn; z95<{S+Og8y5BWaV`oZ@aIPD3#Xd2@K1iZKTY#0&ejdoHcmRCyCLgLiuRRmE5YgyG* zGEl-2i^$3iV)nV^46sVF1bvc73Ys~ zpNmglT-@Ow2XmsmLsSJPkOjLC>BN>Fe)T+Z0aBtX;pWHH%bI0Uk9W>5cw9YM#)GRU zDt#vWDPJ}Fd1;*spXS#cRc|4(IPP{H;N!lp#1BdhE|x0{dn-$~s6&FoC&QA#y~@7T zWp8Q)0IzwT=lH)ASg|zK?tIoleqyM99brPUaxnjY7-O5by~gX8clZav5v$7Z!qOpv zKvHXG5d?NCm(0yDIb|!m>ZJ%G-LcMz`z3b+A8EPEy@c&;U0HI=P*6^O4P{ih2C)Xay< zR8|N8#Lb>+zMICz825Barp#44FM;gjq7lZW8o##S4~xqZg>;5%PdBy&4u3UVMfzek zQ~7%i#LJx~F(_R<&LcP&k*JgOZPzeekNT85y27J^c$YY20o9@jg%x>=!QK{H%J^xr z<+T&U(iI|FM2lv0JR11>hxsJ%o%@zIoezvVdJ5*BIgZSjd>pkQruT zy8TUWiu@D|*2M`u?aNeBdIv@`orrtZWF_Yc_AAw9K!sO+x@U>M3e6YVm)aU=LM*>; zyh2KDklSn?swn5Qs#h=3hK6wCdqH1iIWeL|cG9n-NB0|TpQ3d|yfNS%pmZ*YcL^eDd;DCagb)P+Zr%Hx?tYKZbGPDd8(aPf5gC(%w+(we^vQIGM%ELwEaVPuctO~g9X1l@}$ z(sXSS&Z%qf>CFv_!FDgi1###zHM7k@i`^biAz&(N)n>oiZYPUT4u2JjCphyYN zS#l&DB2L9$E|pRoR_kG=X8;6TAhX>^6pNRXO7>U|CykC05)MTN4Cy-OO8HQeR1$T> zb?X`T$cCJSSKGJ!yVF>IplK{?9C7=-k#1}GML;$bTJeRx_Ly@&fqm|rH&y*MOAi$c9x2PVw`*MyDb3~P z>cxR+8;8726-8qWF=4-V|8HZgkgab0EIV%g7;JqNz3W;~-B2N=X9wI0cp zwm$ocrYrpvIr&b;m(4E#3#3Jz0B&L`ia6~3ZG(UGn4GnA6{>be7=K9v=2-q(3%}pM z*cUgcH3SC@?wtku3pB(AQ$4q^7p`x9!iXgV4kwRFN-L*|(UpAk9tXQ3Ejw4$*^y~% zv5a%$QxyHgU=G4s$S(NMlpda~YM^!MF%6rPuovi~XrGahV-^+i7HX{L@y z4@bltgtH`S94~aY6f09n)%XKPxDgbRb;(u8<$~fHfs;uGruG-kNfe!O#kg<@Bv)yx zbMm{EF%-}3Orh6sJZ$cpX*~r(I=rFAHNppk&25Dd#oXx z6_lIaB^xu!oIz;h`y5eiKvHJtd6P-3obCG-Z*@}Yq~RoaHS$@L8vBgrT=2ByO`x!P zplWs*Ei6G^dY7@VrbMFKTR8(@V~1#h)$rF4ym`agGNLYa z!)1nkj8@nby`u$Fu!p`HErYin`_?7?Y*6ByVivVMdC)g-u{F<+zc@25g<`#+N@^lK zK3oSD0V~=o>D@F1mwO!MJ?yfuUp_r@Rq;d*G8|xZ3x6+3eRY~h&0i%9jyWgYkA({& z@A4AOUF~fF4qG$u4tFPPn=c#wH#$VQBeXOE@82?y zs5Du#hUt4v(&(RZ$##}IVWpcqAEChalJ4bp#KCr1tUTf%R{x!Ueu%~6Zq}ldeRQ{l zlWVO;fIx6!6Y+aH>#Tu>HTPeZM5~2Nn&4VqxVnI24P@a{j5lQA?{z|RI+lPUD^^m~ zJ$;K;Vj$04;{>KM8;W_QU(oY*fJ&pM%?DVQ(M|d6QG%fC4mO(hQk+IB2N!HRPd~vq zl!2p*hc)q-P@=qFQ@lo3)Tx=D-X7)L&T1eUAh9TD-&w*eW=@0$9k`-x#vs6&FodaN zHA`(ds`*CIG*4kl|-xjN4o zz&|7~-_vrZa$Rza*61fDe*i1oZz#ZIHMpq;(lAe#?S==E7^PnA4w3j^c@L+1Ytd~0 z0X|B`F}_UJpT-6HnqKFY5*k6V2qw=p;S{xW8j8r#+woOOOJAyzF|kr(2cET;!?~0J z41!2&>0*`TWa@_kAwG=ff~-%c>qbEVAYZIIf}$DVTrLF(j~G=z81lV;#~YXVV*2d% zC8mKrHXtEHtu(ZJwCK!zxKdAKwl@~CHVha@my{Mv)$4(~iBGusIqE;6@0-F)qn3?A zEl6`1!-k-q;^ZHRK8mUc-CdEo<71#4OkY>W!DVuF8zhekZ!oz_46+^jnKn=aBns;{ zVYEvNt2L^gBRk}$%qv4QQ@bzADWM+iE~1IxEcsR)Lms6v_%4RX=~v(l3AjCSVTeHJ zE&cMhOghitPumZb;{}8J#s{_Pbue`BjmGPj8I2dgsA@0epc`cflRP$q=IoWFnA;42 zkcU40;*>czM%8JP&xc}IFZL(K-3>PM#W5j=R1m%oEYho?A{!q+YTolm?ifv5VdxB+ z#W}@cB!fiU!4Mip!Dx;y&+jUCFrLHkKLFS}OrE3`3^y z4QV&qSWuS@W__@4nLYLn?q%tu*)B>MZh0;>kr~ zXt(Oy{YxcjwHRc!?$&CVW%*+OB27cFP{%coaGM+Ss6w-UQd}e$_WbCdzZt9qro zq9KJ10(Pfc-FY>y)WZ*|xej~Gr{_67e0Mp;#hiZwU2=xdXBo&orlfCUyoR@^I*% z`}gpq{vRr6mz{~5NRhX*Uf^BwT=JX2j4Hrh!M6V8G;(zwpi7;pko{a}8XLSR{_vp8 zYl+M!Vpgl1nJ}=s_4p(9Cx3H*i?`3tol)jBSID)dl4lSI*yzSR#q6Ux1FAR?J^vnHZT1rIh*ZGCvCNmwzKEkNOE|G|s5vG7(W&1R{yf}aJ@8Wjp zNKayvZiJ|XJdDgm47H^5Q`QmMVU~_NJ0!`E6<&ZweQAK}dk#sC|`Y^VHL0bs9 zN7ml9Q_Th`Xw~(VusIDqOL1}CMe=Py7shj!c`ffmuV>v6(i(3ckO6^X&^yn*k z3xTsHZZ5rGtAV*yfdZ2hA~o%ob9-Z(G-%R9d*u_Vzf1`aREv2`rUA;fXEqDl0f-1U z!=$buH61blC5?o|<4<;>&qGJcM#2@qZy6z%Ku${PG=7F`bopz;LbONnRAikB=O2&V z-~XcDg!0k0$9BfH*<5D*M-gnKIGc|?_sr(T`WoCdnNI0Uc$C~15nx=Ru>WekO%5i9 zwUV?|aLwO|@mj@iaDOr7CjPn7K5%t|d3I9p>LOzlKaY$UI@;FXlM$R`368G;Or;x$ zq~sX0Cr0wx2z?JBW79$>W}Vt3PQnP?UR{mD`gb^3A1SzEgD~EIS&rC-&mx?L6V#C< z&!iYPVE;T1iBat6P@aYQE)dl~JtZ@4hA*z}XI>oD+*4T~pRImSx^(0z(USKRQ9$hb zl@A_TJ*i$qWp`ewMlUqZML@ktRTQ6 z3`e3C-8uEF&ujR}Oz7jAadybrB3SdMZk0|fGt;>ja!}ZctxxwbbunXH^iHc4LcD(b zh|s0C=}N@(*Sa1A9FCbSIp+-vQN=Zt9kkcB8tHm1edl_Ku&rmP0554 zR6LvbhdknIQv&*NPnywFkUfu%H>qlN=aSl^c*DHS?{-;)=a?_;iQNG#31;z$;k4|@ zKOx&1|9+i^ytkzp-{+%6hgpKZhn|BC)#D#ywR18HQFIkQsr_WhN4wiCunH+jE7&QG zrB2^xaDuGCfRe6Bife#42ZmT;u6^hQn@b3%uqe?@CIXg1g@dRVzXtZs{UGV zjhEo9e4(c2U)!19q@R)awcuH{Ds@Oy==I~}WjMY<#f_}3jD;?~H@Z6@@@tjYkYKJd zF<3^u$s^);@mXV_(%{^dEOgaKC2pcE^51dn6Mz?5(U906pnRpCR?Y1^gu)Z=oF}|W z=a&eVqyW}pej@hjKGSA`qrV-4S*Ty;e1b{bs6mZLQ{@S z4DHQboSwTJ>f#k7x)C7$cWp#kbo*)`~)nw`fkOqX~Vl-G;Gjj*v&$#EqbN=rp=rsK^a&shwFF zma<(vmKe`H8=#xSw2u4z7eAzR#g_o0DyUo|ghUQ=H($t^mnrKW)D4*zXB)m|_sqI* zBfpc5UZ2?>S~IhmL$PKWSl8bEIlQp`HoC~+89Te}1RwRAO`CWD70j7qGNB48=t=cJ zJ?dTex{HQ`G9?G;z&yIDlq>HDZac5ZsQF)zv^vpT4K;|EWK|qWu#>*Ufc`V@x}GgCa0_0o=?+_q5DL`=t^lh@ud26hknqI2(9_ z!cL1NOp>xBhh%HeeKR+{R)Z13t)@7_z;v=V5Tk*EC~Y$XV7KUfny^F5l`J@ zP^}FWUcD-DQory25Bb$z3=;X&2SUwgx)>H(x=!}uWm)ab_ys0UlXU^?eC;GVmi-E| z^-SDOx}g9Yv(XV-d|uS~G!W;!C2AixU^_ap)zl-ez2h%!ga2WI=$e!K(~P^8 z6`0J)!Gp<#3&7iJI+P242?Ub!C4iHC2p*}l?;l%>h+q|~_er_A<(5A`|I8p%wDuBw zdpRY_eRt3~G`}ut;EyvOwaH#G_wr@)-ssUJH*ze5?~K zPb5b1$(@5QnKK^XR8NrQFidd}HQD7HLIzotycp*aE-!j2l)&2NQdY*cFqOd7Pr&N% z41HUq$*9Zf+nAUOW}t(0Nv~m6LXP$^NZ8;f)FQ6}aj=F)-uQDHC->R7R2g?P|2SKJ zF$RojQu|g9RTvjD4(kZ9oNb4HG@Rw7`MaM z$V?PaJ|M7BMYb6qAv2Acp<`RY&h`hNe1-OcPCeQ59fR8qKE1YfT+EKpeGJAQ0lNYUI6$ss zTmO7vRLD$F4T7~u{hdG_%8o3cH=4!#Exlf@o>k7+VnBD12RlOtr1re7nNcD1R|Yk` z;%j1nDQ9Fo{m$jIpu9Ed4$)P(P3rH}M}?$h3GhYrx>77r(c03OL**}+>4;JcyLMj~ zW?a8WFGI&_$kG0#wX;&Rssh!`{ObhPcs%?J4J{P>f5)7CaT#chD`P+voW)x8Z>zl0 z{7U#XTjAe7aRzm%ZX!H)vIceebDtOaa^z4_eY#C@0X-3QGwKk&1@Tv&1^Lr?@(6@G z68H+@M7Msz)@IpPEO=jB-1fqVElqJTZGU85j`%(+q{qNf053$JkV4p;h9nB8)f#>- zMuCBs$c#Tup}xVn>09t*=d=LV2^p0ZrTooGD%UqYFR621DaLwaXL4hA({9`{O_jo+ z*6N1EaVUQS3Zh7`u63*f6^JG}8yie3zABHVzJ%!H8>#;KDs`o4Li3wW4^LClM2sTo zds7VBELFM0NqXYit(} zDGdH%Jg`r4n>g0lX|+zx@96-Kn;yAmM(^Yh)Gb+hkqX|lrp zoxW7m5ZsqP87X3~nl_B{5lKi@XxxLNCe1$aMxnh^{av}AA5v^5@Oz+1qH#}( zLo3p@daAg;L;!;+SmCW(6|OtYgz26sNf_DOQ=j4||KSXhzQWTZVgmH3FZw;IB!%lK7dL-5l?Ei4Ps2 zeqK$nGVzRSJJF!|F4ow^r8Qx0!!O-#0nCmm-nnT3A>R<=-1W^g;YAu>Nl@F~WQ3q2 z(#BQ@^>9}3kWT@Tdymf=k_evJ6~=#Xbv&D_L(jiAx>X_utn zB&rsHIW+AKil6m~rcM?++{?*gIou;GwGfm{TBS*gi$Ik$xMm4V!3d0NTRiO@%7b?) ze@as`CdMKPqLnW=G{r*S33y4AKr$U;5Y1R&78uhQTyhJFq*gtvW-9}E#FD6pqGXBI%M@mw)9cVL$ev0J!H+C|2hqak)Q;{0M8MF zklzCQn|Nj5mklmL7gvQ5+d_z}66xTRSjwBuE`U>tq@r%UU&e!$sE^8`!K+hF-lD;~ ztByMTaC(RlTc@{ zvbm)fptnWrvPh`@6xk@S8l6n~5vf@Ew^&}))g+FEBC3E!>@FhDYP2RTP-4cqFz=@- zyZMQJ;OAQypDOaO$`teBqt8@&p}Mx;`5HhAKA<^3rK8K`af6HG?reF zn!OrSkcu}?gc0+st1$67>-8kNNXfT5MoQ`o$5M~L0upbP0mYf5j!tZh>|L(F&nFsH z?5N5<+ufFsXI0W8GGkrx2;^B64p@wGO1LZ)UqVZ;G2C|K9k*Y$1{SCQnPqDjG3+Wr zD>Qi4vkC*Bincf6F(Nbn@(SkjtmX|*EK95k6V~aOE zYOfgmKE;_@IbIdl*X6f-P%iA!kt+K5i)-hp&(V~0*2(hE21UNkaiO15>ttH zy9?0j=XY0P(hM0NP3(jYbg zC2>=*>M|#y18_-dU@c>Ur7AJSAh0>UI9T6zjdWfYh*g;UXAv-9NU z`^ZavBzEt8DQ4p^*yPFK$sOuH=Nn=>+DmH&j$99Y@angOh$gedUV5X!dTZF^C}tNg zu{0J#k32vwPypxx11iG{=T&1M)`I6U5{{wM-U_uq!?9v(wk4tJko3lIwn2V`sgm(EvPk} zPynnz>6v|J8sVEz9~5=+&46*|8~O|^gkoTfTsP+wI1t*snfl#d;=M425E&+^_5B$X z*W}>W;mi8=ljsLx`kS%W>uzc;At2xzBFib@cjx-NG*Qi*-fl?z6c_zP4Jzv)+}?15 z-|ubFW3&D7DYg1e=fxFC?+x4O$29bRQ3R@stFyVW-GAnPr;+R&%uV1ukP+~y#_}-a zsZ3^&5~-juNN$Kvdi?)`2H;5j4-J6L@jn7!|0|?m&Fi`! z-+eXt*Na7?k&l(Mvj@DDpIoGf=S2k@w zJf7#m0M0)RRYcW);D565RG2jaoty7<9lp=!Z9*SA_2&z1R6&eh-?}ah1TuuSuqR%Q zfPKTc``nW1&dpok{ycG|vwJF&ifqjH={BcGsO{v$j4?CmK9j)!w<#SZnTAW@FYTiJ zUTQ|C`>Q+AYPd$vUCi;AUmBa2biS~i$ zrSd?2kekGl&rhq}>7J9)f(u$TCp{C^`c^VndzUR`%$Lw0RdcY)$POM{OxX5v?i!Wm zu*uMH$PC`kQ|5U?+mChP4g0n^Z8%h{TY-#<5_G(!c`=a!yl+Irm0+W{nWDnG z&6|uWG;yI|j`z$RFgoBvw9M&D3Ua3PQJ9OFKWnTq!fO>h`xf;E^4iz*zEeh~Q5Xp7 z7(4ib@6gLuAY|b_V$Q(-&SO)@=8s7&4VaK3Bc+B~ky%-Vf21$5&(8NHeUi}V-9$kN zw5JOI9KSWGmEueIQ>aJ_b9O_^2_x%E3{Bc5msvsd1|)05$EM39%0uT0BMs-QtKJ6F zJmXy=exWwrI}N1YHG=smA#ozl=d8v$EXw07VXu6TaAB;E^1A`6tRdU#Up;pKy)!VNG`WO zBGFXzzB-L48BV3Tn$R>FS@uTGNEJ?_KMV~{TD2(}M^>lZg|sLAt|p~|z0I;5c#|pj zCD8Z`CfhL9kgS-?551ipnIUU4)B^%EydzeP8I@ma6#JqNCFp99KPkmmOaCYaPD%_) z0{M`w1FJkyWJQfVY7A5hR(3s|S~Y*J=SXp^=&%@KEq3<9+iN8pqKJcxU|k$%zfJ14 zYe}8!nO#`)V$~Si(ZwkB)9*ZLw4xy`$15m{zqWFYkg5y3n{IMXH875AQkeaPTd%+F` z`^@Dzph!nv+4cX7VVdBTC)dAS-_0c8y@eI1gR^19E_PpJ=wmkCxRr7sOtEe}j8BT$ zr&K0Ioz)quG5uaD(p?bDybTXmI%zq8a&m!byvz#mCnDeE$FDjqNKqnHJN-K3D1iMk zqqa{HC%j>MhdsvMsr=qO#lYAov=^cype42T`=RBb$wvzIaF7(jILR6bI;-g}a&{Pb z=cZ@g*OY)Ta_|OFeBG7DO*gSN3}HF&lhCMd|9T%>LxNC=5Cwi@Zv;~HJD!Jdl8ePQ ziJZ>~e_kn)=Qc^|=w1*kH&lYd5fC7m151cP^}PqS`MVh2$K?yp|xI^EXG=!lI5}d|uv!Zt3)sZIAuT;ZZ0~p^Qx(^B)_%c~%RQ{t z`xAOcJztw#gylU8aQCl_Jl6R=`1B5V;}`9QKMwaSToC}wL-uB6rl2`_>K zb#XRmI;8;f(D0{sp>c}i*@=$o{=^8-18oWtsmAH58CWqdMId}Vgs{kEM31#8-#ALV zb{d0xVPG|+eULEAksH-D%dyG+wAerdL2`EO-A0V7_b2cN&5PJCJ_L^I&Re2EEqI1j z%&8Hd(@U_(bs&z_ZTwu?(#Pj*dvD8oihxKRE9Rqz5SS(OA%!3t8lRbMo#qYpW(^E3 zxvZ(^7au$!`>^y(ib?bPUbLeO$NU~2*^MIv|Gl3+t0isydpavj>}eI#Auym8kK9ZA z*tGz=+zY0-Gh>td9=!)f_B5s4&!Nb66@s2#>!u$3hW^ZcE-Ux#5>F4Q=yOs4+1%$G z8x7|pL3VwjK~Fe7b4G$i44?hbKl{DJNQ5vGwL65Y`Y`K#b~C&u@^ydt=jrpuIb#X= zjqdVC7wpy=$X%CzYj~fGG;iD)HO^OnXpQPBUa9GbOR38{C0&RgqlGhsa~@b%hJs3p zbhCPG&L6J$vg~H%ccMHGH2RRtprLaw@=FSK^BZ1(1tFeqy`b;k9U$dIfAQ_a{K_Q7 zp;v$BLyJck>vi_@oxT0r-6AO8%7gRt&s(;Y52sh|!7F{Hy@PVjCy5Qy_dRqqn;e6} zLG02+K>_I)WaYT~Yk?o@t4gkc)W~cLbynCK-`-5T#hrRRc#t@m0K(`RlRN&$fp^Q_ z>#Rya8w&+XU-S*nDT8m?Q^H|7Bj;(R2=q>$UiR+X%Hp&NlwLDC@to;QTzQ&6_iF&~ zVY>Fz-G#>8X@G;Jq)6zYEPV~fwD9nh3x^)I=Ns{z=l6U5prIeAS`+YovnNnL1<;58 zntihHrS?r)5xw)#Wz74ZB?v&ww0GP)R zDty6}s`He>7ABXsR_M{CYvlI)Y7*T$gL3f$7dh{4{~TV#r14V*9h$bjDz){MP^eX( zMwBs!$PP;SGDI+{89O(G{#FPZ=^5p!&z-&SBEwyscKiz3i&c%7pD(__i$Wp^kR7ZM z?_tR8lMRT0F*2~hcmp?Nqfy?4N@XC_GSHd?;SMmNAuNYM+$*=uosG&P7DiCOz-``+U`Q8u;FeRpzc{^GeFZ} zLX|%Vxp=Jmlj|N#Y2`?RqW138voY)g!d0;B28}atrg$h+B}0v45@U-I^r(iL3Yv{Y zLerb@4u(6d_R5=o{iTBR0LzKT}gN{0fyT4F|Z_BmZHtcG7e znG-QQM#}kT5x9QZxz+Ww&`j~Fm14i}n8AJbMxHj@rxz^GA0*v_lse zNQ++fTra!nVUUcQ6#JwI%RAhoYT}L2w5^HF;Kg~Y>84ON>(JGVLcu>5e?x?dy6bS5 zMm;`Q@P4J@;pl_5G}t#HCiUO@YhkMBk*EJVT|PLns%JR%D8bMasq#EN3R-{UcX{>{WLC(;r!j$MyP`Z{X$DqdM1b$4H-SldEXsN8- z4+6~%fIM6Wo9PYz@KB97-SpoeNS>5m> ztA6rR63^ys^-Pn<>xYT9`HvdedGr6IVzatW+3UWZW;#Q3KV!akmyQ<~d|v!EM+1Rv z#j-$jl$cKN?mwQw(w_5b6RcO2mbA;HXCjW3Icu%S8WbOjLN?`q0y`xcyPUD2AlHaP5)VGqx zbP$Ov=L88ssd4WFo^>DzMGS;w4hHEk&GUSI*Cxrd|;&^NM(H8@%JY0o z>M}XcUHyj;ymS3sGcx+RVpP}4?czq3&|4O@-AHMU)oH=Gp}L96KSUJsl9=D&ckN2_ z!x@&9a~cZdUoQ64R*w}J^%U!Dw%eT4(m8-wo4~}1hotsokNnnPUGYIurIIRsH_6XOoHt5fH1eOGsR34|&Ce)GlwN#h>b-3-W@feGr}KVgS%M-O=Jg9cZ`x4_i&2P)&#zvQ^dDR3BkDa*=YvC|$OnzQSxk>N7u8Mn;_d~NXcc!F71mhl71Q0_3LcF^^aT#tNPR zhe!lLOQBnzvXPpk`Vok3d3=W`ehvoh{dY^6-%+e}rMtF|z=JSpT}>qXF3Mo?3~Pd{ zUbD%kviNWGx3Cn4d_1yCbC5DQKyQDyM4 z1@+PGPgG8-bEA}`6A_T_5NN&EFi-EpUTs7(mJd#fb&U>A6hoqG#f{f^_jYve81JCat zV-E5Y)CRQms6Co%1dh53G zx%U8mK7ipguKU&e({6C-o#sDj0sG4Ksw@}pZ(LQPyY?bS+V?u@d|T=-_2uHhRMNxU zu4}dDk@6zLbmsH5Di`lg|0p8)f>}~y%4VkWtxK(XO_z2cTiu?T&w(M4yII+U7T02Q zgMVe(sbz0pX$J?T8O#%4l*QlVJmJ>i`a-aTkgZabrbJxZrH~8S=M;HVdOhOrqb9B4 zA24!9`xx+j0^v1%I|`#`CotYQ%zg8a=H?h6vbl>i=jIUk<8|%VejxaFe1F11aFZqG zgmz|4LG~Lrsn;;V z``pb(LvaOPTq0VdS7K1_#~0CpTA!X79NM0>RyrbFNa9h?g5Knl4LyH+=#))FKNI_X zq&p3R9huVav}0YGtlgp!JDJ{t;_v70YzcP)k92~0A#;X&v#CGRL$EwqE^w>XP%8Bl!=BIS#=BLys zi>|B>Lk1xkq^rM%!ykq2Te`FE*hRmOy%X6_xRp^e}Vo$E}H)PvkzJBSNK_n3}%~(7~*AyP6g^ zVadgKF{Eb>ZdjGhIKTZ}>eFL9apq211$}a(ko;qC0hwEM`3*rWMHU$(4cJ;P{teShC^EB#u3D zf>gEnMD*oh^B>EkLdzUAKceG?qwCAL5uB(h#(T!{2>y(T^}rzXZ&5Q(J%hnE{gj7_ zS!E;E)U~$zuZUOS1TQQQ15N+6%_rx!I=VU@SN%nn)IY8F9vr@b>oh1}%D^lvcTP*l&osGKsS%rI)oC9lrkPxar~lb#;rWW=#M!Av(PfuPhK-X;;nDe0)1l*T zjGV(kru(^W@hKqr79)y^nmjU_)!9FP^x!o&6Oy|7s1tst5QkmYxO9HZVPlAy9T01& z6gZWOuZGG=$;sHWO)wP@E^&5??9<;p|8!C*BYFD!N>@V zFvt`r!qbVQ9iA4-+L;f2HpqO;B-GYZKZc{OY^Qh-sE~l%_0dq0OD?hOnr)nQ3%#NAvgkVMFlTJvWw^AVF;)nTJLba5a5A3zC>D8mu3cF{ zWTCaQED(!xbFc9J^W0q>mX(Q)h+)O$md_vGno8F-dGEPzd)JX<@54D#WIXcGI=1J6y0J%rP9%Z) zYTx$7gZ{l3Qg2(qQ-15&O5}$ibZi>gcbu2e?mQq!yiaqcNsmQ@eDwGXTHRt67w<3C z`S8CGOB=1!P`XHKImO?qqQ3XN1qz*qmSR0|F4~DD7ltxpWvEe=icadoY~Hm4z65p` ztr=-*xRE1cUC5=Pssm*~|NZp#yA z)dm5gZ*h5FVy4dtUw}|2!sdofy3d?roa{dGLr4aXn4vyH!L(F>=f_YU!;1s(U%6WP z&y<+U;mKDY9a6u35C)2gy7tLL;4;py$vp$Qg)02B|hf8H?cfi71=UlE~U=R!oU7ogav zz)tY#wHAlOE3xFup+I>+kYDGiM7Dj+pKfvF;_=pXzRl??M)B|n4ZuSAi9Tvx^vaNm z*ZNWjH7MXcVTOXUpyo`}jdq*p&v$cB8y6GWT%Tf!Ydo}&1$;PaL1V*8B8vD@$E6Vn z3duy{pEpy_)BIC4@$;4`Ma)!-%Vs=ePCjAECXINBkl$P=fm~^5;?_o@Qy=S%)6`-m zSpZYu8~O**Dk4lor?V=azl&n#o+n(bHm{!!K{0$7Nz=8E$-6j55GwqTfbm&uO8Bl(T{r zLO^SbKQ5bU(DXSj*c*&&osJhVq0$Kx=>@g4v$03`XHS=bVjd!O*>lctxB}%$GrY77K0Gc;x zP+@(2YtlT-(gJP2X$u9lBL8d%ntonc6UM^nwN;QCeMil=A|aa(aT?uvt&J6k_mk%Z zcP=AxYs^$6@aJh*d<4(=!kg2bI+|bM(17+`Fs5!g!45LIF2O0>O?56;$A0 z{o9g=K;$n~^bVd$EfkZMCJ{7kU#lRVg5$)}Z41sD7ba0z@}Ubo#=iY{c@3Z&f+{AN zkj1iw@OP0m(8PItnMipFrst+lKZU8q&nnAL==3Y0ee}vd>na_w0|t8Lp3vz(Y@u`5 zZtLTML6rt>itqS>xSJIqG1r&Ek3onXXGjruUu6%eOig@?IJbC13#M|lExup#=f9U6 zvbtb%F;O^aj@=}}BU}HAmz}LA$NB;^BMxmDC&!3JDqz3kZR7e76HC1`m%bRt=YjJ|r)W3MrYq7(VO z^uy&Kq1Ssj>Xot)2FFGssl%l4BH~J;c5x(#yl0wJc63 z(2nTlRB> zfhmU4)jR7o%L1XZ$qOT0;t>CN*#%VnV08GsxSaN=^Xb4eVY~@v-(~b6SLMfvE2l@= z-<@YT=j+@_3D&gh`#&-A?(oy3mK45(t-d?iwCvIIQqA)H;Mtt{OsD`8O-S201W~5h zE4?H7pDbnC4M_6M=o?!YFAdg(z=z36$-^h`A8T!P87@j5{v;z|R$?B0&i~Y){O`0S zHy`*v>gaS3M^b*A&WT!>x?l)Z)M;ghJOHF~RD*aJoDiHi6o3{JlakER(Bgs>;xPBqr8astBJFdZYw?!|QyLyo%S8mUahHER0&Nua?mkJ?U+2k`+P9du~UnN#{39E0Lw-A-4 zqaP6#UM}_!e%5C6T-6bA}ss_j%*HwPIxn&h8U9wHeN8HV97F2!OP zgBq=#OfQ3Q1-jUyx_zi@^d~!FPCRms?2zfZW3TRWTVNn6fM{;Ws!k18 znv@^>D8jOzkeJ&&^7^o3XKLQ(z`_(3&@X66R|-T58!{{Y*;4M+1Zp)$t!;ufvS@8T zS%2r2lU9rVcxp`W_V@AFoiI-M6FyicsNPo71@pcjhrRn2& zh`xP@ANL|p*zhtAJcbI)n9MGnEQwQ}1>ap2tA(>rbuljGvU?fDUtS67WGYkToAfNv z?0xb;$uCeV@lVqj9O({cSH_`V@kEoADzXk{{Io#Wsq*#V!v&3+F?eDP&r5)o(#TfV z!1u#^Tzq6PMs6Q0Yuz)8k4J<2~NurnE1asy?2od&w21yD#KJe`R< z^a)Qrc0rRGbA8BU6z3aY^1>bb0&{^xyYCHRaDiK>mdwKG5b`-A8?_em*7OQ|meCY=%Utu004Mo*`rFAlq1}o8Z!Hu*fZw*-l znGh{icHsZP$*c1?{n*BhGJ=p+TT7;%I;_f7kT242;;rW7P$&mpv}vn?z}ll5L$aDi zQ7}YEEN!;#iq3?yewfZ!Z(?I4<}Oc}>*jq(0T!rI<4RCyT87wQ?IX=(bI~Zd%E_h= zBnS5QdLnISi0hdmuZIi@#QGxR|>TcwpaHS zbuxLXZx`nl;(J@eA(N1@G-4oU#H_T~(1r93#!TcNnz{YD${3>p=SjvrR<&rJx5%Ef z+;H-|VtSAQK9h!p1L<=5iRPY%@}M$f2+9fJ&xN1V$5z}aY13^}hRDB|I5RP761P!A zX$n3$+a$|PO&7rY4F!b|h(UZazJ)OV&p8ZE2}??nu|X3pRTB`VZ9$zpe7rSwlRQq- z9hH_0-Ikz7=_5_#8+Hl0;DDfsp&!p3a)hf&$1PVUzoq*7r_VWl`4k`hI+^pnevXr4 z`1xWyRCTJf=^Nbm5EX_w!7^OMzq5HWT_tAf?AQEY(qW8Y(PUl>ijEuuIRSpJ;S>!L zXv!#l6N$v$MBIY(;}lt3vbJslXmUfm5(fYs;d8^$gOco}@_+6u_;bd*amxlIpGGs| zd<4HLz?}D_s|SBF3;>nliDnIFy>oC~NK-bXt7YZ?fcBnk1oBj@I;gD;G-yKx?&%_#a&bFNz5tEvsx8@`Lkv;GIo=X z_m7&o$vZNPxzQ_%FKXGbq!hiowe8kVa!mefohn0R>m@_`0fqWh;(@n>hgsvlTDv;i zc-E<0*&S*ir|b~ULw*6 zI8STm`kvUrSBto&#c+g|mfZx4+L}G6VQ}fQyvC*?9-YWinL?!KUSs7q?+wySn zCuYI37c#ft!PY*2T4UqaeoCVCH1K0Q#qo%@mF``(SSKyKm*FzijjU`Q>XS;Rq>?Ny zA^Ryejk70d+bDteQ&_86!SB(^uNz6I6+PI4RV|95me&X@J272ZtzIS@jLy+=st=bq zWzw$B?vH~impb3qIv|r+K}rm{6y1er!(bJ9f5>F~jTiMLJUpMJP~4rJw8iUSNd!iN znD7uwx3{l%Vc%UmmRClvRPL!=rzPCnVZ1THO@t5)7jc7 z4ASG1UGKIBrtFu`?snW#3?Fh848FSGN3vouOr-dSc0BTv8*gnZy2SM_kWpnbTHN=x zd7aoujVz5sYe>bg>1so#VrzLK4J@1))#cyO>8(qk+K0I$KqhGf@%OxY+J8?n?`mf( zauEI_uh!wk!{%<~sv5m=gkmzZM-S19QW+BA|BtKd^84OvCr{|(+zrg*n)x^T%kE7% zla<_`%zGyf-vZx@+b`j+xZ$P3Jgf+VW5UY5u;TKxm{V*#oGZxM5qj+L#KEh~-|w<% z#;EXkK<@vb$=;NGz-jdj?{L$Aa0o1r9G%s;p}f(OKGaC+Gn-bSw+}^E#@}>f!gN?~?>affT6=8;H1h{n z3OacnD63!H?jnCE@`pkLtRCGT`PFN2v91-Gci#T@gFLB!kh5P7?35pz5CWkv{;nzV zutrGnqxt6epc#JqdkX)c*W|WDyOi5`GnLXWBl=5}toE9z z_^8i@nvHFR4Zp|vsyFL=+x`*=SQB1rs^a22eYJ~xSIORBMw;&FyR%;7pyulHGICCj z9H)eI*MAw?upm`>d?s$KP^Aov{B}e_RXG2oVwlJ9UDE$C376pUzOo_^^ zm8Dgw!mvloTv?(W>64^@goS#XVZK|Q-8K=)7kwZ!1tBU6KVMU}pP!UcdXRa51INYrml@R)#r zLa;mwmaaQUa3oawYe7xIcCokrGzki<&H0(;t2y5>60roXvB)ss zKl$MzP(Fc`eT}3mBVLjkhx*#YeM7gZ<3_LZEvpaTc7KnVPWi{QvjEB8?2vH^xy8#r zxh4tNT7LMDH@NtW%t!n<^LnIci>;3y->TJYe_0MCou@T?q`5JiC#|*+5zoXVZJ|KKpg5OWD{SO8zc2xqeC}5qAkrcg$r+! zJviuAJ2zkWi_zT{2L->6|1WXUr_LB6l)odyefCyF+h%Hl4Z;xB28?VMd>SS?PTCbN8Dp!s?O za_PP=?Q~P&(&R7Xr2XmvsX7{v^2LLzCEH5mygqCK^QoJn+Arx9O3(T4RMuYKOT*(s z%1^*v9m=Ui%`*CD13K6x#7&!CnkC2@8gvo^;ie63VA$pwsFd5vnFwPmRKC^kOs{$1 z->m01MIt`)brm8kN=11X!|fW;J^A01L2i6vr%XgEE5`837TF#hMexYA5r|S1f4+CGM^o5=sx(4C1ZJYIJrEv zN8*>2RxXr3tQdr>lFiEh0y@w8Xouk+SBb3$F*i@186_j^8szsv#A%*UkCg}^22D!y zcJP>t9WoO#TB+(q;ngjHe9YIn!a}!R6y?w^L$FBWHh-n+Hz<8m0_SwixFd^G%tx+$ zt-Oe6VXAMA8CqIwM$^=K+mR^OwI6=k(n~b!LS^25Lh26Gn6P>$MgD~9W_#`QU^&W& zwcSgbyh+aMA$Ka$V-}aQwhFF7=6c_<%r*EG!$P$4lwq8_>F{n^1X}&dwwSWEG#5$< z)8&=aTi^)+MrBtriKM;u+~9t%0+onL3UdqJr|+l9r(U0b)>@ywukqR0l`Xa*IF0?q zmt7+SO8hPH2Wdmeq{tu`quP}cS2JN@6*fsuPJ!0wRbKxL|0or^%*e0l;)v&TP% zce+PZJc(n<8Ln$y*QK9_cvC%ZK0^DF`z6dZWJ4&758MKd&zwOB@FU}@)=*6UAI z%l_k4&WTbH!?XIPrL%B+x6QQ)w!14NVl~vZ1`#;#Z5^Pgt6?=0yzkM5Fv}xlx0=Z& zUTSKmg_yJt^rTy~bsMObZ&O}R&y9Bb*ogmkUJUWSIQuuO{D1ZKgYn^|;lSMgf7}WT z;sNu#v3Bv(*a+YeY5tQv`-H+StHH&~Pss!3`+r67Tao`i#kBY_5HZ}F=RKzd2p7ct zCfEO8LH_Zq(OCUv?65bhKPv&=YLQL83 zGP-dj8fMgAN)$I#I#mw`R=oJlj!iAk?a!A76KSt@CYT>Z736%B^}brfx%yoCJ}%!p z9E>(Y?jI_T!<)6S4SIIO)?bX{DAx5oeJ}4+uON`eo72j&b$gS?b0&v%*u;c(LipoN z2aDi|9pYv0mLo7>2PvTH1d%0xJVMk!f|Ua-+qA#k!xe>nchQgfN0-F&ZwFB=b*rT-E5#kYe90UM)t&yeN|a<$)BA5#HGP8+@$!J zMY|6F5699sGV9SClg48~Ckvw zx5ckt=rC`_txqVb;%xMbrg}d86`RaWbw*UwL*PKRWOzFJa!Oj1rr;PpV^XYZ5e{+|ZP1mZ+6Rg-s!uBMj;iKbuQxOOySDwPZ| zV)X>d#!~mwx8xsFJ0~KYd33q%q)E#y^3oBaleu`A!FS-8=b z`qQ+@jC#;qSDD9u(6B-SdsZOKRA?`fmlF|A(L4Fro7*d)3=uv(Ey{UcqXRLxYx84@ zj)K4Namq`5vF0Jq{DO^yy}Pz&(|5&1PUC2@Rc_!(Jwz2FLd@Nxhl?ZmZ$;QfWF@SkOtkH7DNHy2he*1L^Jy;yAHAO9C85s&d3~%nJF{NFlG1Q@@2R8nMCdKm*<%6?^_9L z7!JE4qfbFmZd{_f`bC(A@BbRoBp{XqoO?cTT2mwM)3`#82x8KBFQx1V=O-rapBHQY zdO^VZGf@n$Jx4h8LxX{k8;#rWr^;SYCu~E<$SQv7cGYDGw)BFT2QM?43H~$JBf`IF z^UE#hkc<`b`+SUyj&slL7<< z>MQX^KGkp^L92+pA8D^e@`-8??qgcnokrN!w=u+-*qmX+S7y6c=mfS=%9HZ# z!VuDH-gQ2AE9W4@Xh6!$5;D9pAOn9*K~90&>)D69Z&Y^V82>z^%wzTG$@~cO-d4f< z^H<3kw1`kJ_c1)oL|pd8HDX-J!7p>K4y1!Zq}YoR`XO|i_%OI>1Se0#2}59P zzbv9tI?olt*>~cS6F^NWM}UnhWI!G(zpW%6#*zt zcL{(`z}O}^Dg^M$W&nRc(fi$xATX(BKsg}X009bj2A(}ag?+?L2MRN)p~JL*wScm0 zGGS;2ga#36;)?+Wfq{d-?CJ7FS%6CGm!kqexMbWOAohk71I8I4$qZ!__j*A&TP({) zMTg;O4+GoXbmD-dmuaWW&?$@=V7oOOXMje~0S2t@6&cF8#)1KJqIn~QoUtJP0mLp` zybwWM;{kjDxL9Q%&vH0QPz%{S8cZ6u7y_3)_5oQANS=6?1(zNBwS67MY3bdK6#|$J zX-yU6sUAQA?Uuhm1c9H}JrK`G>0HoZ9JqrOK++dhaozw#dDuIrs_x39XlhBY?(k5E(kxg$Co5@8J!Yc?<@Dm!fe&;3H;aaUeM)h}I1O z+}c6Q0rVu*!D)oc-VDB#1(HwM3o?>3Qb{2?RrFqQ2OJhem54qqfwG(Eg#uLnqJun> zk8Dx&`!nH$_ov}u3B|@nfVl7Uf7(&bidV1$27k1Q9=T00wz< z-yWeibcf>9qJE1ay!?mN9_vQngE&_}Z|7fTr;FD;KrU|qF*NFc8w7r7W^e(5A2mb(Dldcv z_CZT_jv(+xpL#u19sr`at4xIc=P0SJyE%z-Ow$_C@Xc=C<_Do@TR+4rBwpe9BH zc&M;UCpJJk7F4Kxnb|3KO_g>_ptp^)Di)vDQb~5vOpDC zNaqd!qjKmLA2u^z3!tyzLv$kfCKynAt%ocf7oZrZ%^Ju6&59`lq*1tG1O(%{g9GA9 zCJ|t2i*PJ&)5-!4sKp6~BfYVrWI+(sG|7VF&(DO*)1|J%iDcAgF3?1%N>Qt#HA89t z?#&V4@v95yD`SfaZ6l9R4n+KR2ZBLLm^rFLK9wl7;(6luRO5_b1rR~c(~^1cU@*AH zBp_*hFMu<^46-PO2KzkC5fBXc2M_xmz5{Tet1W>>rO3g<()koQL16W6P`1LCDPa3M z0TVWank{1JxvwX}S_Q`tU@hoBw{;x8t&KB!%Wxp6gLXR{duDzbTr&=2pa5zi`ZkqM zH6ao=j&C~&26F}ybiZ#$fngMa4*T+kDUl3Rt%b_xhjIsO5vQoZ%Ff$E67QhD0I_b} z*N6z73Kv|=1yYR7jL<0oPlVilZhV-$tu)G66}^ZCObdex#fd1XXcEs8H>LD#y(VhI zKI*Fg$>%mvoZ2(;AZ!S0RmFuU?ea24C@Ts`C}-FDKyvrqRJetyG`P4WBVg>Xf*#O8 z78B%Ip-ZhEuyub1py&RKcvj5&3FN5{{fw}tsh$Pr_?b%`!O;D=6skQeh3|TLX{Q#;jE%Z*Y+0QyHQL?^DO%D~IpERbhoV%+oBMX3T3=sr9l zDRdtKKS>HrN}WW5eGB3U(2yTyc3KMC0OY=n37`yjH3F`7X&uiq;aLx*y}XAP!SFvz ztU0rT@yyyD*pl-D4OZM@fE?Pp2J-EgcpKu6b$FghA?Fx)qWNB{*%p{(zTph-;0S`v((aTjnRKYO6T z{sxr3-IV%;4sSg0ozufqy;o^&OkEVD@)CJ&i;J4$8} z6T{)M6{!1xD=xbgf%M7@pC#(wPPTJ>{8oVQGO`3vB@sFY6)=tjf$6|T5<0Tqm~e9Qw!m( zgNbj4!Js#70RuGYCCm)aOmR#I4QKL2@SJM%PzhYv2dKc6Lcjs^9q=Y^4=qFVR49Gh zS(5Y1-U#9{4S1eYIPr{5tRL61^~Z!;bu1SMm<1XQe2SMjgFR2##A8YLU%m7Ldy<|PgsUq<-w#{FgMfy(BwA^~V zQJM2!{mJwvtu#Yt-?dImU;i`$=gigP-${>ozSy5~A+?XWFArrtoi#oW%XugFzmNCd zcu^B^qq#39Z?z}h-t3}LYf%vLKDlps$?Sq1f*pq5E_s?l*=y_~t*UR*&ZlX)EA(^Z zWw(mYpD0*}$`1kuive(LUr^3Z&cEnjXZs7S{cS^iEp~eNj|Ko8>*Ukb^1ZOAsCjMv zY2yS0*C%Giq*Y!QnOZ!umWPSz)^|3JOSC7~eqqA!<*?Q2 zv9&i3)BipXTMbK6K#QQKSarWbii&fT_@KL5c91kXB8Oa%Fx30_GMpT*D22a zjpE%Kx4AP`x%TkYQL(sJ_VI^-?}D{;=?;IF@KXL%?d#?9 zrB$sX!b32+5+QEc`@4s8_O6$Q(L+SwQvW##;oOQum*351R~+A`O3a?0^7W)Yvj^Wn za%2O+q-d0nfN+DFx5aS>mW9Rs+J^RD&lvbXd5CP+vVC@)icLE z5QO*Q<=_RFQ`+6V1#m*&TKHu0s`Bm&Ipw>SV<2Xlo@A-FyCtV;>(g~ySI3=OhQ5rV zbexfDLYGdx@t7*THkZ+XKEhNfztw6 zk<`0`8yv^y)(48sxGcN>9f3dI#T8hy?yGq%uDb@Bbi`J9Kp7wg#}@NM|3% za1pA!7iqI(gW9W($-v^|rkIK0tJ31~vm0|3>K7*u3uSLeZA;^b~-lw zY=lqX8;G5=%eByX5)t+foxpw*ajm2Cs6!%AI;AaZ74P+;-{Su?Z&2wa}MSIO~%Zc%* zu;QLFkh<3iVNYP1RaAcbH{M@QrkK_IN95HLd>T{X1=ieILRMS!)_+gFL{$CQ;*JfN zK>m<=j?&LoTXjZ|l4eq_*mAD;icuO8)!>-#&S6_1L6_%OKl1w8>#4wi2*pFj8G%_h_dQCY4wZknCLQs zE=U~-#9$81>Klhh6PKYGbFJz(#>M`nQbM{3NH`w)O_fs|S3h+~-Iw~dEMRmRf>hY7G*svHVdM zJ%?Qu4F6|elYNW~Rf1nIXN#1K5{qbzMV6ZoUXqpSko3Iq@3Uy1 zK@d;(suk*}o1xZ*XII*)5q_HFIyvO^%*~3Cl4WPFUFth?(Z7w81tFx1C49QucM~g& zkfh>`jlp2$!c$d`*ATIkYO@_mFY@oXJBVHb$-Y=vt9cHEjGH;qb z%_;!6htY&5%}U8kN@FOd7*)6VLox!QL1T^ejoDOb74KhQ7PV6H2<+512}J~8C;ee#C~ zPn)>JKt^^%COL8=+*}hPte(lt?+lNX;-hp!FZ{xb-wg zcP~`+tzqhe$|%)$q%geBCb(HRyYAvic$?3nl04*d`rbO5zrpDrnvbOZ1$+R3xV3oA zhZ)$4ha{UvQIL#P6KoPx3cowoATy}uGmSW~OZ=|<*e6K2TCPAl-sc4zq|D8(k{u0f#R{$(9*`EmY}$i4XsX` z(dT7|qP;sNUmv~f6MNo|A>9wfPiywtQ5tgAZ_JWdRgbh3R?o9bd8XbY$*Q8{^e_tQ zsF>_^GYk8q?FBwswRTVXEL7Y!lQkPpOZh?q>-B~%q%Ikw^LS@EU^qLxkmG{)lRtJ# zY$MXb2Flh%|57ESj9E}AZq#X|bDfstFzU(8Sa`U}!CJHoF#xRQ$ny9&q$c(dX*RX! z!Tst})E;b0Q-O@MKg%CVh7Z0v!4MPb%ed|64bUCyRII$l6f5dQ$z65#=?RjPx17mh zY^CQqru$P7PrvipeFrI@LLP;rLLuj_r!g8B7)@;4ryGrI3SvIAJTC70dyVvbViOfT z+xOp1yl7^>0lwA{iEG+laA5-fZL|M9`WAUt`7#tXDuf#4lkCI&O97eYz$rln9x?+b zt`3L#d=6&|xxeNwFM#s!YHf5HP;zH5&IVRn>9AX4AeEoUGXT%q^uxh^S_L* zp@Xw|*EsJ=(B8+G+d(27$0>X#1kkwnZ}@##!K*dC^wW(&eu@(;s#0Z{m!IYFG2>CX zNlz+@-G6YRQznxXM3?C0@PDIcS(-0ort-?kq@^&pmZXAW{jJ0RS{2WwXSsA+(jAb$ zNBHc(FZq$(W;6kfh#Rth|Q=y!}BBz{kxYe5wu$fVl`2@Mg5EWB5aY13kNxHIu z?xj$Xc{?DW0wUlD`7HLm1bv$b*qvK)B)$3BKeE0Gxz1UZVu_nN4RoXOVKgccveV4 zrv;_zJ=Cx3%9c9ZPTGT;kdP>g#DJ3cWYpCukHF_*tpWmdhS5u&K46>tJlXi98163M zZr|5#uhH4iEX>I=q_rd$T0JAJUX@C{=HK#px$8hH=z~W5(ZeAt-vw_Q+dF)juV2)> zB2u08L4l_;>Y?{W5;xD0qDJq*et-=tnY~idTvYt%0WB3`TE@DR<{aAWjx&GVg`kZU z^0mPNTR+Im`_J(t^ib#~rfe&{GU0g%mIOO+jeXm1_oL(H-RN}7i*8jY{ba%@5A(cJ zvC{(K#w`#(Wi^!7jAdQok2lJPnPJD4x*``$VGZUs{r3%_@}KxkTwRLaq$J`S#TZ%5&l9)EN0m=l(b(u+E7b=`uQQuj#LKDCfe3 zE)Dw;k3_oO^T+Ym=_qQ4D_7vb(s}lH=%1&t)iuLx%#K;mYOT$8&kr{&-Yft*I$x*g z-vXkw*Yy2$JG%>C%bx7)6vrx>qvnw*@5;T6NB_dWHKjLQ+DaOK)W4D7M67G7eBJ#8b#brx z|6}W`gW?Fjb#d1~AS6hT-~@Mv;10nT2^NC8!$5HN5MXgjkj35I7YMQh_r+a;Uw-%A zdaqv9`=e*3Pjz?oOiy>6(|x}4QU3q5;Iy?ORJlfG2^3Q>5*JrmH9ICVF#c9Qev$7^ zw6o_Y16kyo7vcfdgz$z%;#P^ke6l*mq(cewQEmtS*&5-@r@P0GiYcrUTf21kU5ShV zbe;F-bjg_tT(|4PjcIDxovfpgs-M$Sd|aA;T7OMZs8Kj}(Kg_32CJ=n5@#eP%0V_X z@`flrNf=ZJ3qJ`&RJBASE3{;OoW;eiq4OWmjTewk&y!)3ZZa5WbOZk=1{V81oW6wc ziXgR_!SV)sgP7ateFjDk^sO4!P1ky>6*L15Ubk9P#Kz-U2;;fD8^>Bk4HwW58NalxA#ILWx30m*1AOi3Gjj2y{+&r!6MojB{@bcO z2!vRiFRFB_yrDLE9y@U)%EIcj8E9mc@{3i_&aOwqO-xx7sRtBMmJ znjGaqhYfUx+7hCy7LiAKS{0s@YJbH@$&xFQHOCZ0&9(F_RTf~H6LCFW8PDL1?|Mi% z34z+7H$0zU;>FIT(TprzUPhGdvwT_JJeaF0ZKg3~C2Uq0p6|Uj)g6_8Q(vQJk}isx zLl==Hh_)@!VZguI6%$tj&Ah|S?&VNcR~^#I``LSU)#<_XdKY=g9ev; zOV#>x%3GeCcw@OUWyBNF!M@yWL7>txebTh@Vu72Hh!bNn+lG$M1y|JxPPb^|y+8En zev?{j|N5<8T&G)`hp(|Aea&U!dAFiFr(9NNSMQE}*#hv~Ra|>BeX-0Jrdx9wohOjt z?d0Er4ZRC;NZ>O_W|qZvojhj3K>{WwUccPdtd9+J{Oa?}%O%u1qkyh=N% z_$_)$hdhryJo!LpDxxFnMA@i?^8~fq^^mAYI6VVc?KHUB-OO^2${q5mDEaiEawj+M ztH*2nOhuhDY#B-?P71KF9tg^P5JMv)5l4n4I8!*V$TD`=-l)a~Qe(5l;K(XY{NRlY z+=#F=MmxD9o>{`b%TixL1Fc)Jx|Dyf3b*y7oedIHb0B*U0yFTZewAH)ULEr9GABps z*Si8@Z?#z0V`H#yS73rqPpSigTF%TcG7nJ7undEEwu{bv8rLd1qvE$yA;uO6vhG@Z zR;=+b$?fDy{TNf4Im=@L3+yFK8~2q2b0(QK&W7xOVtP6a!*VqjdtIDM!^Dcg+cE5n zcqa5iI60$wqS2aR0Jd8o3z*$RYNZ6wJqOtO8cE6*9#pFC>oke_30qX0X>Uk#YE5Uq zHq-{byO-n?I!-?!?Th$urScf-mF`JltR^>qMN>sfm7(K5l8r~_N&IbX&cxL8(P%`u+M7@k10!PHDvMXSA7}bYNB;~6HR<<7>+8R+5K6EihC81%-_5& z`qOYIEWR3~l(lZ-2G6v`1^TRr#nAVIa!6(}ZOsu#PJj3Ad(_oULl6vYlzN?w4|W%m zpH2myvHd%gsQ)c{cPyjPVN@&unq{r1C#({Dhpaknb5W@L?#sAgYaW@-r8CGKbI-WD z3%PxmoH#KCwWEp`F`sEBie8iMva@Bgow=8S_hvWXzMbY31mWE*tW$d}l&$BGN@*|X zQ$k?fr7GRw;`^aCd$qNg87n9GvPM?wT*HhcX^D?3u%Lro?)P*0`8Xj!8s#YG`IFF_ zpU>?{K_|~C&YY#%;6hAu4lzr<7Q5i6^^WmzXNn;=lvXJgb00`T?R51HcA$sPVe9wi z*oIch@L_)aq)mkeDSMem8_7o4t~qN*_nHmfd?*u8kxu;tU?Cg|AOw zpV_3N#_prS|3(3A5-0;gA#tB>8v9_%g3Cfplbm1fkJsua7UF7w`nv1gG0XRY2vT1| zqDvARS==yA{g(P?KPonAT-E(yiB;rUE+e@9dil?0E4-Cn&m^vaNE}>kFSy%EAgAv$ z=co*ARa*zdhf1#YN=`|yOgGzZy!EEM%Q{v`q!b4O4E^TVs)OGxo}Q}xF;ba-a*YB4 zfZHwLF9r(oi;k3*qJTp1q6{T>TfO*IaSHzLMxcRV{9@DVfQiaCTSa_AIx>+hs2EXh zi-l}&bLW+(!tJUi=ikORcJ$Y9r;A-S-hdPoi(QvVy8BY(DJdzl59Eq;N6RkvXOqR3 zL~fUrv_R=R-~J4MmY0lXcS@N{3x}i2GW%y2Bde!9`GHW4fw10~;iRP$iBMU#Vk;Lg zuw!+O5X*lGqx3eI4yipCs&+3*&Q|$4x09UtI%d`FboDT2dO=ml&v-?x7E3+@(}T6R zR%C(!2-%^1uy+o^#mjU>PtNVWRCZfHZkw;PczAW|ois8q*qWVAqq5J_nXxUSnf8Hu z=nz4nm&aphI#`6(WcfQ)+>C+>$Z={4$ge^4bCm)Ye@YeGu z(eFP8>I@Gr5c3K(QXEyj(MXARyY%#DpDE`AcNupl)Py`)dzn$R50 zH@bv1>0#2*8km(UUpG<-4Hg-lYf&4h~^ z#fxM{u%zP;rGNc?8+Sxer9MB*PLboN+!v}+`sbJ3F%qpgFJE$wLU>yQ1q&%|wd%-N zKlsC+dyesa1qSklLxSN(R{;HRpo?(^RsHKLQQR2kB_0KBeno$FcxP~@$=U=~?O>OA z(+j48RNE{=q?lAS0gAHr@Q1E4EKH&-IE@L*V9DG1Pf(lc^0{XrUnXA_r^_H})55!} z>w3%vX`a?S;op`nb6MO5r+r4)Vp<7w(9GbOLmPL-)mXFDGH~Y%SThMn&QWFUhwKmQZ1UIj?-Wx z_XPi{Yv>f5o59_cdOb^S;jrkuDKL^Gdz@)fbJ-MIX{0QPwM*Et_A62Tjh6btn8^>^ z>9SWdytZ8vDGHdH|73xSod8T)nICG|7ZWf!y8c6bbJ%x zBj*C97lwR(<650dwISGXUmry>WU}Kviw->8WU2L`XbJ+omnKd>C1ex34ZkGeL?3Ck z9ybZ5^$^5wQ{2~6 zF{l68&qJsPYp>uJewO*#Mx4`-Rp&J_?f%w*b^R>c4{90JbOK(qJUkP5#$MdkcaUSB);d=^)S3A&X~P(axiA7ZPv-WXuR=^kR+)K8R^OWYiT4X_5Ry|1pKed3Mt6`i7yAG&zabV?U|p)1SFE#f(x)O8g=# zyo#!xTM?(m6gw$=o4-@|cDL}?HMO%+aR`aCofJam4D2K{#i@Vp-wQ>wS7vpv&kqhM zrck`!Bdugo26}8gDuA7+Bph@tt?8L<-iKEPeJLT7*H*tdPkt6lN0zfkTWGDg(d=V~ zg+$fky{a$mqf1S$Dm9fB4BL4KS|=nkO88}}+3DQ6TwQ~K>Zn-WiQbc&OB*Jsfu4@X z7p70s*7TuH0VDZ68CE&E4X$y;sn2rcJqS&3OCk$L0Fc!En#aOK?a$fV{+iTvUjV~O zX5M#=P1dKuB?SFwJ%&m#l8a4Fj|7LI?u_HTh0^}s41tlF@$ge^n};o1Sa0D#pLL=x zr8R2>a_u?SjSbR~F1xX-G(wyAySPFu6stil zk^*oKS%48Gov52%PQ_^_4lz^<*5NHZ%MF#t{QhiLrQhk|K;?tr1;^pYuG3(K-xvs59&-iDo-LLHW?%Fi0+ zxq(u*vPjZAuM*uU{+c7c8%%9s9dMeG4j37hb_ALTbFfY0gV!fse5hiVvzu2hK_L(+ zfkcr%F-jV6!olE(X)^ce)?~9oj?-RZ)pBhkJ99LohIE0sE?mjqG!{?jTNa2Yu# zeLa-hDMm-=t3e&;2w4R(oV4q7-6@N|%`8j7j zjZLYu+OFSFK}Ln`kD*Rw)jHEM;KnT|HlWvdsh%%b{0mc|(3E14-1wC??lL5|2KTaR z68m+I&ZJ=^D`jIra2aSM&ipP{lk7v1cJjZ`_Ew41mEKj?e`V}jjo$=Y7`j7$*f5;Y zhQ%rEdh5GjF6XOrL=5}Ufg68O`Wv}={(t4or zlBP=e4IR4tktAQSu7}N(%u#}J!c*iyKt{3DcS4_ZQtl%zl^UzSBHP(kLQITt1ddm2 z`{EBoT^3Zxb-S#*O0|U!cGgF0hg8Q`pDo}f`^iooX8UF-D1$l<5|7Jx_aBG+vGi;w zI@z8tw~V!WC_2dvanJG20nUwUtf7rolZ0}|%4C%Cwx8UbzjDAKx6q2>2=?0oVNJ;{ph(Fs% z=-i%%o0!5pF@>KROb_EPTPF?CMerZJc3JkOem0s{J^bvvE^X}@gc%R4r7jB-%i**$dHXXBS%;OLI zd@7Z^&fjUlfsl|OW$&)sb)8g(aa_3q$uuZLiepZ}OVG8-e+I!ZdDB3{*l7*1v6q{{ z^eWfI-^M!Ov{>eiXwmzj8#Rz}3mus(?v$ZdwGIXr)lSBQkh*el^MOIfp(E>${z=-Z zk1+%h{T%eo*GtDgF;x{Jc~$1D<$W>3in)A90!Pc3fKqO%Jkr`F9`Xmq3co|?n7T7P zGun5%KKstA#F#Dp(mcSnJq&pXUA7e@VVyMBoPKr7{{auQ9zU}`>^CS87S2<6H$dSg#QQLB?TcAa!sUD_gBURni~ieWrsLk7jcqdarKD3d7ZLUlS3AQt;c)YoNVhvY zJiYY>BXa2b-7I0l@324<)RpJ$PO>E*p^wj_BSSE`x=2@6Y8s$ekoI%g>9N6<8o_dR zTJ2j-4Eypk8d|vU(;w0>7W}v8^-5}&n9#AlOo5O@G%5Z1gip~cgdR1=VHS(#ry>#p zmX8sr-gjJ|%x76S=3QIYZOlHC+?2BTGAnU46fb-1DKV)+M|1gho0FkLqm8~MeY-#V zndk3%5Eb=MeKLV7xU?wF#`tU2hlhBQT_7JXf{upfA;29%2@H$}Cg!*4363xf(UfU>B19S#n>lcr7 zMWxh~w${fhQC$~#jcXEI92>aH&RDwZqZK);N3`kL@p1e?>wTkl2S-p2db56cBQ{ZZ zz#F<~9@WRCf#@r%v+jG}b{=Ezu$*EeNI7o1xO8ii+vz~O-uv)UstlW-e zz*g}k>Ll+r-jW=}*qPM;#I6*n%W@FJV1Ih&hV(H~m%=T6(ieGr99S!P#S~yXBGwzX zT&?%i@U-_-5xH#sm+od-&muW^!UH%jhF1w%|?q+zaNj? z#n~OF#zJhJ#%&DA`+9JhpEQPq_C=$~x-3vN)O2W_iM_HQKma8nmBvzr9SbNr9CrbO zxPfafz5wj*WL{fvM;S1izPw4qb#yxR?wW@ngkJrZfBz8$Woxqt&S!z>U{YdW;R z!as4_zeg(8&?_wN-Gy8Jl#JdH=0td#xy14+pJgH>lBi)K_)S&bRtn>An;f2F{wH&{ z*)9imLW1f)LjfsyHXZ^(0bpgTTHEn z+Q{3%>v34crqo!Y!KFvEaBwquizsG2DRHKL`lGzJWpkaa2xOh>9Q{hW1|CF2d4+l9 zw1ya`JQiZP{_FdWjV{Vr4&#TtDAgPw!l@0eelfu< z(3{rQkS?H7>wZ+0X9!PTMb7TBWJP7Y@SCui|4JJ~%c@wB`7>+6xV>Ku4p z&Hi-?$D}kLqsojf)jm0+n8z>GLq0a8wvzF{L+*hn@`9I;F}xr0*$xWA`qtA0mE-Wz z?MKP}icZpO~-C;7@1InzzpoZ|=v^ieNktfP#ow z*Q|+$zL0xmE~xDJTQVD1;n;Xt^_a^HPTNopgyF?&*|3sp{k=zt`XQBn+Y+Z#n62fT z!k+?|Q2BD+OK^Rn`kD~i*P+E?)p*bb$$4De+PJ31fRxZoI%au2*PHD&gef%fg6Mb&o4Y z<-KGlhPROTM{$AQm|WtVvNtq~Z>n)4eZywenB*( zT$f;b1QmRtPWw!4VgvD-y|m@^_d){W_OJED@yBOW`Jh8GWsuKBSd^a#NJ|Wa6OJq8 zbw{Rb2GBes`jEwq>smwk0B_I6e-hCJge~Uwg9vHQg|DJfu+*z6n$TgbGJGWAHkgut zj=$!S7anqli6*XnDe|VhsZ%3qVW>qXnscs~pKclQY|(eZKa28bwYLlJX@id~eEXCs zMz+yjofmBn=u46+=L@JMk7*FvQZvp|#;zfeyEwmdQ}Z>N9wKuRG%A*UZwd5Av4=PO%FIobu};HAuo%;_*fHf0%WZF#25e!F#{` ziFDQYhj!o=xHIY&sWs!%c5(`75|y8cH4s?9{YtUpYH8{CKgB z&2;g}zI<;icU&?9DLjb^-Zl|9J~D2G9`5dpgy$71U*w$i=B-^A6-`cS&i6V4gokUZ zxW9`&Pq?`tzv48tMH2`+!$pnLU{gQU{zr43?X4b&529@|YamD;!#_GfQFvu7>zaV_ z;g**paZNtTJrkZ_SzB#LMeHm_^kEw;-sRK6lO_-4+*h}vw`I{2O@a~h?3#0K37xT5)9 zO}Vvl|3Si*xwN;4Xwm=u-*umPuzMWdjP&on2$)3TT#rYtwD+Y9#Q96-B;-+D&4HV{ za|)TTyFdKF_a1enX;_Xbu@!|yy7}7+-SrUnUV4Eonx zR?n!?MtH5+2K)eIbA_upuvZe(ArQ`PaEgl5k0Tzr=8zMz_?J31G-c!;E@=*Ze*7OQ zI**CVnv%pP9JLqbLG@EPK#p^;!wL7Io}0eJzawe?nT2KO&1(ysLZy36_5G1t`bczY z6SREC=CS!|pJJ0v&gvtW=I1B0)yyxq4{Q->ZQoC_ShaC#Qk~uNSja2ix^8X;H|;pK zF+iU6ukG%X>Erd0872Q{&J>0}k>bm-YSbx2$hx zac%{^aL4xcT)CL67Eu&XHYEtT(9=GebS??Ojc@+(zy9@oL?pt3LXE>%b-+u|vde&6 ztzDU?)Oy)Jujeeu*d{RgJtO6IOB2~Yekqk6;;-A8L(&%Ywlm(Ss~6@)Aad`Dj_>{+ z>KLjC&)Jul%xXig$)KwuD^Jl%WbZe)EiG>hFN`I;AmuJrk{e!qL& z&D7kSWiFeB(Xvrmr9LZ(KFjixlTu^n%G3RLr})K$xZge1^TXoe*VPV%n}ZgIvy;Xg z|Nha=``ySfrKtLM1I=eFCe_{M{4IR74SjeZNSbo}!H)#3Sk;K>r$Pd#B$*}!{.T(I2Xmu%2-vvL07k9&^fx_p6y`aTCqaQ41U6>%g_qW!oi1Sz3HqB9=Y zKe|sc(;>v))n2Wyl*WPS{>fg2QHLEcjr*#VAZ+X9dmP1SA-S-g1tJO;4Zo?FWJ-{( zj16#PA`T=MlF4nh*(Dyzlc2nrlSUjekT9N;DtI6EHr~5)p&5^}VYk#^aoTFA`qOaK z0u!QOO2TyYQ|>@_tqYa@_$A zK~F5xG1$(hg`5X2o#=0eeN+SVI|giA8g5bxR|8+50Y!lYBR6Tq&v`W(KSxj$XcjL% zRl5ygyC5pz)-l%8j%WTTEpE+Yh&&j&tBXBoA})?C!)X^yX&RXy2v}2Gdt~u(5>H~d)^`+M2X7RDHTzPc8%FZ4N;e@+ZGihbIiTQENVmx>5)9{ zyW*lXmwyJioFGGLfij9J=d45C4Iv^*MNnuUrBmwPp^Q(UP|9h(K5em(G71oCeG)!| zCeFtx)mN!&Hb)tRo=WMQHDlhVy)Sa!tvyt0L!e7|R7MR+HVAJTd9JuxE4@hxK3MxC z1KQqi!GmbV+ERjw7>DJNOkkNvJv8uu`B(HN z^Y!QGCdFV)%%%}}W6?Ecm3M@?dm`4-8$!%yv=DP!r8VYlDyLcK1fp*JXO}K_r#m;0 z9K%ze_SVKiYLiaCyF6TELudyXa)s4og#+1a`6V8C(1o(zq?tnw`U-T--i_8D^uTr& z-1iUSJ&FLM^#%7=@;169X$$UOjOid;lJj+PAQ>JCd5{b%lQk@qwAmwz!!g3P(b0Re zv{pf^Y2U5}50cN0L$I;He!8=eYSdWI@~*A^z-=>dL3=BrY^Yu4sO48yWLeusIrCht zDFMiGEb7LB;AC7G&+CPRrjbIP0?u3tZCXfVP8zV8andda!j64Cm})?~hIQocWE9!- z<9Cw>r|h5B0SW$l*Ube%baHc3ltt~nh%)2$@0?W}*sYxJH;H2qXCb`Sy1m=hu&@=6 zBcWXJV2>loR^v%Jh!K&P8w`S}I8lEPsz2#$bm#2=LWc`%Fo#-ALNu-Cg3GL{Sgdus z-2jS3UK21V(3Yh9unZBRPhufB4-s@dcRk8@XD_!_>Vns_V`tRiJg#Ut3Ax>XtAXs( zXOSRfn!-f7-_|1<%Ak5L2Ax&ZV;VMI z&CTe1#_tBjHq116hY|jQ_P~NrM3X9{E#kn%5*uXc z(7S-uqzK!)$E5ZXc@J5$Ni2;tm;O5jX|jUVf`3&^j!4E+jdjqo@GFbt`>Xkjm649i~ZW{~KfY`y2WDwX7o=!Ww zU#m>HgEx1V-}Vg?u+3>Oi48^i+cna>5NHeNdy}2IQJ_mhm4tizT!yE~fB zMST{TSSg`E{OdKeKg^g=0nBhC{-sY{!u~g+cl81ZGfK7x9%lH|2A^-h&vvfMmt{C0 z{smPi@1s-ZfvjAN>oVBfWx0vi3Gz;N1X?fbfd`w5>*D=E5`Lgf1KYB8RZ-=KZE-kd zt*$Q=2M)Od;Rg>x^<-O#2bXxTtHgs@%I567!~W_tKm#;W()0)73sd;dQ>2un`cF=XKBo*1>FZ_Up%8 zQ1nuMht8nhmJc6``kzp;xr9rldw?=`so@j!wKWtuN%_PlrUG*>P%bL%m+9oKz|26`I58(kH zbo0i#iVETI!wq0c8YcKTt3&m^lVID_%C7p6|KulzuGW%zo9AvOy6WfH(IM`DbR{}u zFl6coq~yDJFXCcolG(_zYwM%X_`>@&SW!s|GB7k@cY)veXsjmL{(gFWl%VPgYo5!B zV{tkXrh>Y6?#?s7Iy5#wj)ScS2mSR;2ey#Mm&#yLBNKLLCZsYAK%apHbRO7#^Z?ts zhsp?lI>4LFwl^rzo@CMga7hh zN*F9_+sMRoZt%+8Q#kal#awLVF2l#-!1mbzWa2p=YG{JL5Co}2KYA(2Q_$E!S=S0O zQJdGFXg&tpQalkh883aKheaWaEUgKd3?C)6^WsDP`-`vuV)K^mwDyv%%s78do+&tPrgbTs_HPX)fzG`N7Bd zXX|j*JK~&+vqW2=U_RJFmea??gRvU{7|n#K3H&I7a$~w{mj7ll2;P5_ zXv@?|1EVpAM}wk4y)9EK04k&?fW@Wa3fBXO7@5r3GDy4UcFJn5-2r`d*b;)#z^fd1 z9%x|cPeKkEw+bjVe}Jnou-<*uhw#rf+)TYvF6|Cj8->OlR7UQVEG88G?cr;Z;8*rwVF zfy%&?SwY@EBTs$4uH93e5@iKFIeDV};H~QJ%nCB@c}W`r$ogi&$oez_$n7D*$nAa3 zXjZ%t!Q2ThPTZ-Z`9-R08aL_hc9c(TKOFHh6SXboj5zX@f4{EA!jjsv ztNCb_8nb4hQTWE+{Js(Oo#&#yz9#T@ivPN8=1T}4<5d!UzT+FF&o$jIS!@}@*Xt=STT(w&ztb#iOM#P^&NMURRzHKNI)*q4I%crD%_yRyG;4(o zRU*?C2v~JY=}5^2vHsq5&V)|(e(!hwMyASyHj%1B=%;6Tn1TlQZ~oAmnwe|rv7O&tCu73@2!l1rl0j1lJOb#Plr7FUf5B;%KY!HRNgV5KeP(>5Df_e;-A!aB zVV*cWZ+xXWiD|jXpBe749+1AZ{MY8Oks!2GRfOL-+XYVu?mzjg1}zf&X*cX`0%a34 z0WW2_zNN~%2=cLmz8g#gvoHWmPRG%sUf zu)nMEd?i$nVRc+E>oD^|77y&Ix{`T;WZo@E3Z6rTW6NIc_?tNm8ta0y9jdsDxLq~XXo z_ra~T-GV1>3xe=0RUvpAx7~=~&6yWee;2IgDn$uxbd3i4CWN_)Sz8FhgAp;H&4XnE z@L=IKXtPkQ?3$ko=MGp+$Un>V_du{1O!SV$>0^Vy-D~ioSu;8G#%xm1B!LCO0edW<0*VvQvAlF2#Ezlg-%-|X*>x=H1UrPWz+|?z32CgaYr@Q5+ z;(`TrcaAz=1HmXOUZx`_t|$WZN=N_+Zu9HXM*#Me-+>PAc;g1;C9gt*qVoHdZ27r3 z?}60-F+nP9zn6|8@?t&rb&>tMbEp{)_WIIW?NDz20xAKn zm|z)g6+6*;u=9uZAJ4(vq#U8_C`k6uWLGsPyTD?DSV`@y;0cIa7=9iQfWFqot=TJh z0=S(Ez!z31FPkBIRx|KupPsAR5F+aX*8yAyf3e%FW-(9Ui7P?)O~nK$9NXpNNe;IY zBx|3+yR7Vh3G@xxDgv|Ho;fE7YszHJrF# z6GX@*!_9Z9`O9W^&yqRZ_&tJD4sy-QHO(ck@KULBnJdGwq9ex zg^NZ`!bJ}qH5`*|=(@%UAty`n*-gGhwMb6gaWSaSQmX7K0zw&k9|{duWa7AGRF zKFW=Pgi6frb)g~8n5zS?vDz&k9F zM8`5JU-=)QUclSM!{p^O>g-V)pD-M&h`h3oR+`?rM#3_Izw66cE*#Xf)AdJ3m0NMy zz8F{Vr+#b9LzuQ{N2qmB3%dFGcOg8Vu+Jvv^Sv_G*G(HLg*=tmsAvmrOKv44x=c^P z2i>NtRo1QQNq7YAgVcDjw~dn0!d<1IdR`mFuqg?t(Jr91J7Dm|pyXKA{qgPi3BMh; zX5%YX?&>ma(3`bi_v$JqwN9kVp`~|dwVAx~#VgE>lqL4sx()&niUhA&L```O8y*=! z)Ns?a7!Cx#|>anoDoM@fsrTbDs6p28<=sPu$rG0Zqj>=?H=vTxuFp%slk1S5j0C zQoDOAAyr;zIPc5HZfHV|2&EkQ1VWH=GNp$Ehp0h%FaBKh0!SgLVJy_2Tke9?FLVZi zsz{7cfsl-W4SkN$7Msf1?6C-53_`9G0-GxNRRX@^Jp`%c0{M4R_`pD?Xv8Z5zAz|a z&p~kA}j}T>qJTyc~ zm0#4RmeK-TcAOHwzP`D{&8w_M%sZy6$IY{%2jY?Qb|{R|dsL&d32f*s1Jz;Gj}ihO zkB^XhI^NJAV+hj*VSFi}WC&^5*2W8|;1r=G*r#j93(1(CK@I`8kV{RdA%^%kiV^Iu zj0q7mRNUe83~-1@%@;3zo!Dwdk}_Ozzz|G!#2BYrA($(`6y-}J(7_R*-Vzg{9s`4M zfgV{INtLFmdDM`6$3N+DEc^rrc?Oxb(t7Dhco+wv-AEy&QGdb+@-PgExQrq*s$5(= zRP-o$^7gcBqykuHLShbb=?}BbP*rxd-n}ok6QwT4Nq1EdFv^r9!kbVDrN8_Q&Y0A~ zM_dmeeqWB$!w{m0+t-31RdG%m5`lmg0*oVnM_d;Jf(cqMB{4*?vIu;_$f&n!;s~w} z3X#Rq+p#e0D{a*Vam=^e#@i31y1}qY@bJ?&IcFkjB%4<`>Tl_Uy2z&Xi!mvH8=%RV;R3Xa59-j?x zanGtB{t#@tqV6QxmoEOHMfFh`u$~Dg_sN}aWn-Dc zt}&tv--JeC2TfFZi3CBv@=a{PF2gf{j6*&_KjxzXv9KL~{`^@tVL$3P^mjT$T6H|8 z11&08Q1g@XHF`zU#7~5%;VgX%jantcK#el-UZEE%wP35d!x>Y$eM6rr;FYd(?PMUh zKUo=9UIqFNedt&nQ9c%aI;~Y`;%68pas6QW!DFHjA>l0x7kbzTcUO zjRVPkl>~nMSrf>R8-}C|a9kmciNCDOCi|#A1Ma~55wR|w1dDD5+E35727PPVj)boGQi#Q=K8RLXVMt@lS4hf2{&~7Xc=L=6 z3EXKvkCieoV8IU3R`nsKngw^E@Ur=0Tg}9?Gp*>?$7_p;ggXH9cT_$$&EJ!i4)D1y z(U3P^xc5J8wJW)Mob}58*e*KhX$j|uyDB!sSSD%rIH{>X;@`1DFZvc;wd#}}q~~=U zd>lSp&T1CgWLLC6hL~+fn$O`Omvub)sc)R#)!fT*m!zA6P#<52wy>Q}2i>1LQIPz{ zrB)Mk<|xdJhUPoqKHW8Ugdw0~GO<93jy==gLZKVpJm-a#>2qA`9i4Ef!?sqYu3?om z<=bS#AWl57R_8=UKbl*isZxv{)xYu*t{DkV0?&5eBx%Qk z8OnGVz_Y-ycYds@Kw(;fHkyuL+2y!E*@~`3>QpI|IXE7yeTOt;v)yfLq0JnOPgXt> zh4)v-wBAMV$&c$-^#UvFm)Zrd_xSqD;U9u#p`yYyEz2yX^$co_96C4EPNwx2VDcIt ztaoM0Dj%yB(iN0-?l&;KOx?Z2Ug=Rxck}7+s+Ry1I$ymf%WAzKCDo;0vFdyG-(*+Y zmE*k~wy3{B*YQ2?8L)XbmVsCn3+GGJ&bl1+*Q83|* zX??30dYO$|W3Npme);uQe)Ya&ip8v`mW4ZfHhA{OKk8uZ*zb`KrU?@cIgY0NI)SA4{(+P`L>=84rw)l#o8=s;U8 zjXreqN=@S=`|Cv^A`NqAX*b!@7Nz~LJNIOYUUy8->Q+`F8Voxw$x&}~X_q&ab7$%H zqEJ@(P#G)<-*Wdt;`O4qL`;UQ4{`&6unudMU>H=I5Y*`N7RTMzodvP&Pm!Flc7IH- zZI(l74_Zmn$9^Lu@7C-q+UD1v;p*B3L z*GVu=k>+_X3XnzuZ0hkB#-@7}{_sJ$NDEQh$5mJ4N-T1AB0c_n47+m~)F41*sm{|> zD_;e})IAkv7R`(_9+j>-8e;rPIGVxmZS~vhZ2aop_@4;X1wI2+gu)`-Bz;jDJ(j72 z!lL68eTrqSlU255=CpGU-@3^sgO_EijI?D75Gzto!Wa5QNiNQ?=kuLLs?MnoApsQd%Ezjn6Peysv8U0B6Dox&aeBN_pIP$HQ!hSO-uGLgJZmD{WEqesBk{bKbw5Y}Kq zyULR(;MR3Abza%&eShSMZ~wGT-?q4$WpuY7+xd)GXOX!e{%EUSY1i;P z6t|FRBG!i*E@5gM`=y%`?`IlS&v)}TRVMKo-PP;jiSUu>ffNFn%~i4c<08G-@4?ev z2MkmX>(aMQ{eX@z_ceZ!_gTrChpXp*CaY~fZB$lv)!LpNP5QoVroDUg2eZmCIv@?G zh#Ld*$T7YF0iq6w18m}*l%5EHEYcwyZ{dKTO4>Jo2Dv4pb>!g9q_o|ic#E~_){kNn z&o9Sdetz2_MGhL94QU1SkN+;%S3Am-fIb#zFyOlK12`1Gz)k&0F@D#|Q?|#o>*$zc zKac-VWz?KeX`+B=#YhS}=AGy{4(ze0v zHQoBt3yXHnQnVDUR z;vVbOL8exT1!`;bBob$o$3KWe^Cy{TR`EX6db&E=b~i{WYpR!yBhaDiJTm3NL+E+6 zIN@3abG~tTns9k~39k2#C?sJ}`>hAr2E&pmE^ZQdNaKN7;rg|5#~G(GjHxOk87?CX zD_wm=!#9PBz9q7+2zW`NG|)i%svwJ%%V`r=Jc2b%aQfr-(H#PaY`NbEVt?XXZ<`h> zhH*AwHzbor5ePiOe9_y7@J6fj2vCn>dX!f54DuHxy~Ay^#F*&Ye7--F6DCDa%WI4_ zVOklckx5`Hduo_zpqi;yOAJ~mqHa-uAm~}6#pkpAobuGwNZti0G;|D=UUam)|w4uX)c zaRNU(8I6rDBZuTDZJX{pCu<#!|2boGR^MMv|9nz$1A{O-(~m(ufj@(fg=A;yQ-qrz zZeew%MUI>$3kcNK<)Pcarbb}F2y>Z2cvO~-|}8&s162+g>8c7Dcwx6&0KI0yXu9+b7W<6w&Brg8TIQj z2;!M93yOc>a3Pv6pam^w4 zaz5KD^XM>nzLLD`7c@A+Nc)`JWBNc3$S~PeYpEWV_(G#cLQvC=Og~%|1g{UMz&fH& zBm(VUpo3{LN{uCP(T7$d)QLziwQ-7ffclH6v;GnxImFlPY{6ya(hYrKevH(Bp4S** zq>cRmf&uizCJRG9vBBGeFRpe|%3~z^H13)75lo>tIU9+-IMRfRW)0rZAr>riyO>1A zWX*YjAg=B=jG2!X9A`Kx!xnhYF#o{NT#6TxwbVU#ig4Fa6}unH;^+EA6Nufp zN7+2SY)o<1j*)eILps=uA-NK{FE8u;?E{*7Y*%H%X~AuFQTAIS5;HmmhW^7SUK(x= z`k<8SL;oez{tK6L&x+ZEi!C%N;L?Wk4%4LG#nMj4STT+?<$?>LgPNi5Jvb+_vbZo@ zOA)Z!?Lv3SLNM;!ntMo~DAHV(M+OaUZ6+)%UC4PgJdH;+U#Uswef6t5%zHl9NX@1K#s6)(j<(vqF*`mbT<4pD7d$mRS^9Xcjh%#K$m(Ze@)%5ACX3E)T~h>8Gv zcT7oCuDtbLWiPD~1+va!WWT!-qT_~09JXss+AREgMj=Mzzu_tQE}oJZ$plO6PG{HS z5c^^0ZEp@@Z-O~2UN{8X#kH<)5xN`ACzTXldQV^Vg0NQQ?S6!+pvO?eD2Ka!d};aXo(O>z#4 z*q9uHRO5ce0>?q=(=%Xe!8dN-ish^$FD-|+fPG%0^bbU1&(q%Pki*OxMm^%@=)*mf zZQEg|iVK}9C6sFF>bShra;J>*Cvvu^^hJmAX~PzX!o3Jh*0#R#Ykt&e}dqS@2|SejHw6P;~#D>pnAre7PyER zI&4Bh$YI3-vxCj5Se0}v6K+|FNrT+Ol!qykPez1QY#YYsaVHA}8}GqrRQzB?%rQiHss*_6YR5Bwe&d(n%H5e4*Mh}o zZIJqW9Xx}mLxroC3_v*vGtJV?oM7iO^b3O`;q~tNQgxD>;?<@%rC7xA3w1w>poG>_ zkB}B(=)91DYR};(sAck;`@cyE_8tsA^`aQuB(?kk91e_yiJ3K3xdfaRgq`_+$mY=Z4zmFwr}G<$|V1MhW*?4mL@3$rvqhY|6ZXt1xE#8{VxDgW5NcP8Ex~N_G;eX zK;n9YLs}Kbiuh3ogBW(Dj<^pI1y|(#F$!e%09P)A0&ynmzP=kJ9M+r5&Q7g4fR2VeHLPFK!zE z2qreDFB*^wT^R~^R|pyI=Au%=DTgS*V*#Ef(lnytxmU>lJIZ}vuf!#07>CIi_7>U{ znqF-qs=ThfIlL-J5V;vd^mFN#z~ANqrzYYS$V)kGwi>+;B0tDeKCKg2Llz$8k7-D z)#LM9^_@*U$J8%DA7KU>6h9-brbh$UOIoN3@RPvU_HxlmW$f)z5%YYVzRwevhIRX8 z%SdJ@n8H(+Htcs}qX!OOzq62EI$^{DNEhAtig_5jWQ5lg)fMCb!!k%*OmUCTBjo7Gnw(wm`*d7^SM#R=ILnsWJj_$RsJ#(1 zYCpHA0VT>-5OgY9w44Vj#l|*Jvdlln7q<%3@6)jXzjD&$-^dVU$rjvuP1AOjXMtml zLnMVE`9iP>!0~UeC^|dl1>?OW*4tv4f8dChDLNKuOD2UoOgoMlHswBc{pTG7VP{Rv zmw}M^Pw>~(f)gi)Lr?(3Q;vTmsiO_&_KYF=D$O}dSkiHkJVk`r_ZldPq0vdj8-Pf2 zjB!Y9+KpSdH;i4C*olAuqX@+jcqyae{{;d(XPISr9%sLAJ#XD--De$Vc~`3asTO@_ zpX9CRX12##yoySoE?!$&E2`1QSuelk$EjWd3iLuhSCgSyshS30xRE8rmZwTvAkZXg zz$SVw#af;FJsBN``SO$}f0JcgzunZQ_yzRQs56an6}V)oNFV+N)yQ0{Q>_-xNrUNz zYn2iLE9=Frl1ih#UU15ekTzeixm*VJi#hM8e*$rrcHCTKZD=Q?qf@gstCi3)tR{y2 z38iY}c24B2Y9k8(gLIh;jP%hMPKG@9^V&PBo&#bIuu=@ic4~;oU2ySpI6dv~vXg+d zXfg^N2JRo-@`G{!&s=v6ipDxs`caypvlOGf8h8ouxKx>IPPB~0*0bZ8OCvV*56vWJ z5x%O9n+BoZV=6(;5-8xj8=D_$$<9$pZ~_V*>gkOV0vq55Qn?yLP9YqH8(oJa0=-SQ z3bR$;YuQVpo%q(AC#9ea@*(*Q@?1*5T_o?cCtSCBrxF75tCz?rN31iVqSx{RFh3qA zMB?P<8f!a1?Gt{;`WBQIV+$W(%>Q$hw@hEoV37qQYuPf}hHWXBX=;pQ7#g)bjZ$t9 z6?zwu2MLg~1FB*M8-&@kE#RHUUx_uWOwoUo$lfcA3_AEjfwtxhT`j^YAG}}{T=TK7 zE-p-J$1)8mZ4e$wd!fvsxrHYBE^{%O^#JX?40fLn-=^Q+0@MZ$wjN9*S(suDf8m*G z`H!p)c>2MeQzl#hUPxmG5St`pO5GSP7fsU8Y9pXu+6|n!#czT#gII{b{km*9#@bYT z9jR>y1#B1#4QiuE+9W(=&!i*_OYg(LA?g+{&;dqKFEMW)I(KXqhDXmD!}>VL7YG&| zU2Ot{$fF7jg7X|pX0^OdpGRzQuGeg=NWqnRAlsJCp18frzLp7bKyD8RjOAMFwwGIu z#SA8qC5K&|EXV+$$;?zb(D;4*ob?KtJg@c zXNOxJW$K}FST$DG&aiX=mtEQoDSJr$A>m;Ocwv@rt$M7FD907Q92bYM(p znPnDMBXYiI$+VO}ds5~>8kZQ!YNG<$P6rSOlK}<`0q0C)^DTay-b(SXPhBmxyYfJS zUq3*%gHp$J2?k5zzaNe+KJ)u$o}x*R0iF?gzA<6&VlYSA2ngG7m8e9Op(lao`mcP& zrqstHH-3QV48gM3n{>w-RNphmE?H*WU8n5Y3dWl0G7b50DUfqr>+pOk6nbjiJqf_F z%FFXfmk69>h^Pw0U7k^S!j)BsIIc&Fl);b?J~D8CD|UXi;dnxkQ>tOVYJ_jBl~p6q zpZU4~3phpwZ(achCqu9jeO6rs9nl9uk?HV~#8_l6(ZyntA+qv7-4RlL*?id{Ie6e? zYCSj=Yg&lcv(zVEhT!)}++Y|HwLhT#=?0e6P7VUSSbuCeQf2Up`o7|}``B&DaJ%2| zZrvywBefuagKEDAHG1-4>oO^OJE zlS}4JO|<2&d~U1_E4!PVt>PK2AJ=$!Ev(mbgn_ArT2GAC1Rt)p?Itkh@d98OPCwTd zr7fq+;*to;eAF)#j>MOZyLyFM6f2M?;x?$8g36Dhdx&p)CcTmc5c>l>nj zDEB^QUIE3Km4zd8twMTI1OaH_7G~P2wNdq8tg1&rxITfdfmwM3;TFjyi!@R`ZA|&d zm$ZC9;!JrpjT~h6hJH~#=`5B**P5VC;nX_{2hSbDgoJ^lE%wq9EbLv^@sN!a>dbSO zXaWaN`8!wll(pP~wF2B?VF&&+G}K3$y(%AWv=i~GB${`Sw+zJ`9Re1-_B*#kec`T7 zq>)Z(IGe&0zjEeAG#}2$>CI6p+DO|jtaE(7-j-sbR+HH+1K@%oB1%!3sTqFO1p~_n z5f?$c#^bPd#AyiysS9<*vvija(cQDcMwHpsLS+o}IsJT4lFZTQP&WI(L2A*35dc~# zd`apiPHuGXP+=I@(MRWxp%L=8r*Q6_#S}gD?D|V+Ue0y??_LXgDlzQ$UW+3n1m*wc ztkmZha7sXmrK*x9Qb?|0vLbV$G%KpE7(5XP2?)*wzIQZ^_hKC``xA?? zVBsoe1p)FIe#^P>+TM30@k*WTv}}JSkIxVO{p=7nrxG2bPhg~H&Y5t zrn;2fXfu9OBh++wvlZ5=7(T-Flk@iQDhw6f(-nZk_zzzjBP(iH1eXh-kys}Wg5(Ad zp49O_YHejEcI@C=v}xaZP?}nd8qBs=tfesa~&!qG2Cj(Sb zwnPA!A;?pZ*+axx(|M47qqJiAy7o3n9Gszbz?4svUfK%$8r$~qTn}w&zl}@JSSpEW zYlu3nSMIN{txAyt_3R&n8MYnwFgGCD4f2&}pS`1I4S4HL`H&fY-Tz)G+q;6R=-Jsn z{&g`ta{XdUq@TL_5T~7BO_$+DToa$dJjDcXtWz6wf%()?jB&z2o5PY%&p9d0Il1vU z$}v6`Z{9HdB2dUI_f)gXx8lVtRoc#UPoO@*aUk3aduP|e7Ky`Py%PCMs=tAlm?|gE zH1RhhY9||e^9&W|siJ9iaid10oNe`^{86Si*435OJz#R$KJ4htMS9ZmYc!8GwwWHF z_c#)PlwFZXJ4B(p#wpacgpJ%ak=eJbMARPp=9w-(!sV2}Kk~;tMDWSQ26V;Ui5V*k z=|@Fk>YvsOUQ+ARp$;_z`vNTy8WjAPZ$@)%mtxIekhUol;;38TA5}3dFg{Tv%+3bm z`-yRCM^yeb4fC#20nl!NG7F;?5t=Fho)C-cgkLH6u_90<>k~}D3rXD3*Zb;(ELxJj zbiE;DrLXyK^I{JKXyh_A4D8HWgp>7iV^174RRxP~(3QnQNMH#Ay$1PzpkM3XhEMIQ z&;)~ZCQ`~udXtyg)@5XbY(EYNT`8TT>(CmnQX zY)PK}LKZTRgOAbAZ$XoI;Vj^5Y*egU)3HCXL?hL=JI(VZepgWbs|Gdo27&o+ zbeF~;+5*Lsh{c*lHgLp8Bv;I(1y*8$)mDrRsuN*avGULwc2I#J4@mGj;5dI~ppoHK z5!X)(L)W6_utg3u21|;bQk~Xi8jX~{D+ghN`-Dj9_&I*oE1ytLo%z$3rn4}ik1GlS1YmL%+x;Dqq**w z%zU@W#ctyaCHlc-pTy#3XYS|3YN08u#qMWw+Rq^rlEzOmVK-uYmVH?OV8#Br#!U|5 z;w+O|m$iz{>Q`+Szi*VXIya*EoBJ)h&|KbDrM#m*FKBCT!2r zH%?tZo#{9PbF7B#q2Fs|M&)N23G>4(mwc^TA6`;HCKp)44|sjAsp!EL$D_&>5A!4+ zU}+Y~5-nivf^=Ck%e6=cHyKGjs*GU-i04ZMu~G%QXMHY|HSq(+&pey6jU4SHiAe*= z4#+p76|mF#){5B8A;K0Fsd3SG)sxzVwI6Znz$}zZl~>Cg8?-oGi8N}wl%T#yI#+S&RiV!#+ z&ofMiuVd-h&Nc&ty4FE*1%?x2-R%X8QgGrt*^Po_l5iR}U#W9|@v1n=b{k6;n*34T z154>_$L+t0Sw&AyMl!!T!&%~lqDS)d#O6iP_~~C)lkj-F9R6a9PjF?dI8Ah0?)+?v zw2L8K2hNZ#tyc@`*SAb=ooIE0JejdY-uIV>ayDnQqM!#L)aKn!uI}bip2|Ka3a1Cm zww_%yYNS_Bz8*#q^x7U8rEvY%=evfk`A$`=tU*bgnEOmmNyaB%OF4~Q$8 z5JPewvImO>=hc@UgUFX`5wOHd?4y%}Vp?}5aIUD9v#KI+jXY~**gLfELJ*5g$T;WQ z_1_x(!pZ=Q-Ll7gF?HgVaQC$wjABYW?G1czXISRcdfgb*@X56^NyI{GyGF2ZXEALu zMg0suFs6girGwpe*tga=_Ay0~*P}Wvk{;gAZH<9=q;6;78k~FOV`dYEA5ScT*xjCa zq0B_sU5q`JTOqL8q^SH0+{y<0>;IxkPR*1Cg9iRij}G8+Am5#ge~LLUW)5!7f8|!? zQ%)6=HIasM|CWC1c*xsiUR5`hbjXLxUs|Z>R%#4^yMr;1MAlbUjgyct_|-f1AJ!Wn zfq^T>(>Mqx`d(dUkFm=uI85_jbKg($-A`sEmftuxp;vM%ZiceY7VU4!)GnPa@OaB$ zPPkG_@}J`Da4PCcE~XaJe(+cI$pr%z(2eotl~Vfcj1DUAgLJsIDre)KqNs?g%pek> z?bDVNt#QnLl)T+9xyjj6F6xw4*y;Dt^}NL8OI3s{fV_Vqq*iayJ>I5@Yp~d5$J{<-*O z(8Jx;i(I4F3_b&bMT}fhtC+YP?=E?4dE}>x)cKOS7MQDt(Whb(qnbiQc>0kA?G%nP z;+w-Q6cv+cpNEuSJjVHl~7Zi`e_QBInyG{!LwXjcaYds94fPx@&W#8dab9PKKLSa?blSa zlJ}fNLvKs}BB7K+M7K|0WdgPzWCRCE#*48_WUt z7(gJ@Q@rH<$o^fhX&u@OazPN&4Ak%Ew{5*Vf?>M~aT$|X{KiqAyp;iDkI?mB&BkM( zXQ}av9S%bEy=olpjsI4` zY-easq=%UhinD`Zp;Rp3wHr;;QS_8OAF$z!EH+VU< zGtr1xT3KrdxQrAqchT1(e2Kqu#QWx;P+2GBJyXwcX6FRMwb4vsV=1C&r4&8_RryFG z&Q~m4X-%SBC!S*s497poszLf3J~c8Z77c zOPooN@{oFNRt|#$S#~k-m2#75EYH%~!gX+we zTz_RP3?-h>q)Y`rGcgJS=`~ng?s7$agJSDItGalY?TDdR=W2C~07VRKgU*;d-iPSq zm2~Y-DH2WgqM>F}I*@U4_Oi5x8{EabTEQXnQWY~maaOIKNm9L!mT4R({i-=L{Xxujsk(!?+xWC-cCOsusjPUFe@Yv7Vb%2!eb|N5) z{c*7bCe%zY2<7<899-#yTt$tuF!#1cqDd-p*5ZNIF=50jk`ifmmY~fB7v!)=_z2YJ zM$3`$R9yrJTw8}YYUF&_Vk)8Pcr(qdYycR*N9@DP7LAKEU!lh^?peK8-IT{@>>^!z zj{_X9b@T|22wkK6%|=;d&Zo&W(aWN2}N(Xpk0 z#~9b~2E)<^65qFAEA6;;L|If}O z7&9ja%fHEt|GPh#tp1(MdU|W8Uxrx5qhdCdv6GR#S8fB%@g%{NixNd(q+7{|lgN^Y zjI8p5wledBf)c=>PK!1gGeuB{9}MUUd)zfye;j2$=Xh`V{LJ zU%n<%%~jL3-sIDSK)*1jPV**y7;dAed=AGxsrCtgTT)bhXJap|v~I^)SW2tX__?<@ z_M<`v1x>p)K;=-4s!*Ew^X(-|fVWhU@$@p%%Q9*DE6*1R72Vjp(kuIkem;EcOilxr zqb{q6<7>~n`K(@TAkk;mKeNyV4+t#}%Po4b&B`sXLqsy7u~FFB9%rG@k>Mn80>?{~ z(ZmvvaicQN;QSw-&*jmo)5(4bHD0`&~P?8t3Qy;-KPm?*P_g zv?e2?ty`Y@e$B~aC60@(;>J9y`{Sw!er{e4m0Lp?l#H`cih?OBi0KkIHyw$2;%c4X z!WX$v0Li>(g=@keX;_X-jguSW!)2Ocaa0Dt`IN5LSq?#WrR3k)K&4P^j41y|pZ-3m z#bO#cgMuhLMbt&(tW|z0B`yf0gW5S^s7ZC~Z46uK4v7_706!b0LfdqzF;9B z(P#X5@~F(Dg@c{uDZ@aRUjOYY?`}g~*H{rdPfSv&6A5c>~i#4 zW^EudMC&$l;?YP|nehU+RH;8^#Ilz2fQ(RM zwLWGR0u|UvkBB~#T-PG4^c^fib((!ZQGGjmR91@ZvJASjUoiDF(MsK!sGxzn!?;xy zIOhc0zI1LG0a><9BPx`yN=V(=1HP)uj}#l?kDz)*8dMZ0`*NAipu{PtP*EESo(VhJ z^dSK%Ka(FR=->_`v?$dKFN8vH>A`CIOlAqDQ<~&XadHu|?1UCoS?imGG4zW7X%pgh zk4gkaA*}oMqcO^b>Rw80#H378m&sQ9pQ;9`YI;p2ho?+(1>`jS)cy7``4N2RjQtrg-s-)NLZH9A!q)Fw`VP zvXw!)Q6754vU;6eq;OF;D;k#ouIeI=Xt&UE+xpnt4_W9$;Tf(VRdbIfm7%AGctq|V z{dEp&M7N|`(4GoA|M6^i)E^6f|JVT|ht>h@A=uN-1t>(HENNmsI`jzSH#)B_&I@}7 z9cv}RJ1pE&!(8MYlj>{~osiIj#=`{I*oGL-2uyf+bUdqXC@{6(WPx%6c)-y&C@dfB zbGv)*`PYEN@q|RWUxH4oF1Bb=%f9)3to9qBP++pinJ+*@v9c5!zpyP+!CQZgmOR!$ z1cibt^+%j7q}t<5i8m{}8~e?aqn!-Vt&w6%FZV&E90@^$$FLCdls%pIg6}`=%dz^k z&%F4_`f^sYxuH`LhESUUl6rK4!4z_fCW*PtFi8}W5m*Ec$~E0{_Unpz;e00P#AFPD`|GZQyW|6?mAz-`@0X`S9OLZ7G7_`_cZM#5XFzj z(a^wznIi$%K2!smm3Y>S|1)bh2snZGpdI3GIV9F)*j2khd*O)VsWnIXV2u@&lO`J) z#D~WCZ)~^p7-H7c_9-2=pWYDW3VsGYL8qHQ3UU02W(^d3g7g{6>E(AB5cda^b*z+n zv+sl7w{Do3=!yNLeEV-U{I28v?{hj(W{z(*LsKtEL0c_j6F4dv8(YZNMmsGY>LLXN*>yuA|Yh={m`4&O%GtfZRajAKmu#>k%Qm$Z%mR&iwkYEr`7mK_B5lyio__jc zr$2f&)aglCyE!3icr5b6_jmV9diw(W-*uK(d%zJN1}`b3i`b6jf@vR5%Jl7pKn_A) zjn63|U0(C$G#acf%Qa7;{kInVujQ?2+@6DU!F#F+kke6U!=8yhFq`ShfVMxW%C3(V zSSRrpgf_j80k~rVRC)S}7O6}=;4Olp`rOMo&^iV`K#Vq43{IVx8hs2Zr5p^aArF2)v}fGpToSNfFZ?}H4Kqx;(#&)rk2vmFr@d{3o@Xj$9|2G3w2fU!pxI(y89FB1X&O!?#bIf7Of ze`WzFPZD_-civjVX~{zCX9$u!QEWeTb%12bPAl*PZeAzNGW?@<>x%FBS?6;wfhATQYnRG_i@A> z^L2PLlB?j`>c5o@;YqmK!67SM-nDC>qh-hM#eSBfVe{p})3ih+(up5Drz8k7DoYY? z!c#!Prh{762mihm{d@K{u(6E(tygDZdm6NSG$|T1%?#02ZNnxcm@~aXqXsaDfEbRD zku<}IXly?gef26rn;-(VQpMSSth(tS+p|}9=%|OAKo>VT1|2>?=IOI4kpwO)=$ z2H2ZSYP_p|=#x7Os4HL6y^W;l`3=mU+6^6y(^ZeoXKeJdzK?=wg1Xl}0UM*#3Ej#` zkr(X3K$cZOr0W;s*)U^R%L9_O)_1Zr6Rl1Bh0D$U=;WTCyP(d9Bx+U>61a8uRjkTz-RnukR-r6gF}Crz?7%nuro!xXSA{`w7XrG2%Z!U}>Zu%^wc)k+hu zz+zRJ6Sd1?9%A#mDY~~lU*WRtmvV{FwqzT49F5Jn_+rKr%I{)VehP- zOviLyLQ?rSispow<0K&CN(HScd(A@7;^?kzzb~5ZoFUMUrdK`K3XP`t>VZ`8Os#2E zbFImAk!nL~gtIO0@ec(uo8fyVp>`Q0uKEvl+UwWX*+xj^>g0YIN+?)p=Zu-f}k zYKRI=J+F0g_vw50Xm!ch{06eypW+K6`pGHhPA2)fmB9n;UQ_fJi+_X zYfrr>zH;$f@tV8D#q77<+tRi{m2)+s-(r2h!Fq^sf1#~AH84^Skf9G~NIvvtTtW$j z5Oiu3giV|U=GXvPa(I*v#*SBqLmi&U_xR+7Zbr$su3IT90yHKLz7AUcW-zG~MZ`#;eYEvYMc z#?O&gN{?82!bTKWh69C(l0yRS52iZ3IcvE-^d=%se{rJaaAX;do#}c?zG&7*WPO4A zx(ILmCugO8DuBbM!UuzsgD`WZc1D3?{ZC&K)3xuxsZ(8JFZt;%> z|5iul`2RpPb8@f}{flZYR$q5m5JK{q(X8^QFZSxtG>r4unb#nSgNfM52RFl`EDzg{ zD=IrYPoHquOKO(N8-2RCn7o)wXT>GA5D9OXi@ZH~X{-p+s8>nG*e{>mk}7mjei6}F zj=UAqBO|O;$!62_csyroSSGs`s;T9?5fBNMpB56U1n1kHpF{&l!_H^?THPZLGgPTn zQ*nR9`3$5T(t!koNkj4(RujqOG)8)aNGk|seFSLY)^jNiaKEE+>Sy~B^l%F-`D&$1 z(Q2z9RFnj``b&_aJB4j0-bw}L1!1gtlmp?^Ha2o#%5%OkCzvZAgwR2d*o*h%artgq z(;7S$KHo&{7Rm!K0xbhi79$6W+-i!J6Z`aO7l0NoaENWOMeK7*(tmJx>M|me zEgO&$(2FiD(dyj}yRJyf<}Ja4a9d8SFD6OrLWY%k$ZXeDD_59V&uSJnCO`MTLRAeX z)w6vW8htbDxFLNCXvvfRNQEt=v5vfB*0T3|%wp=xhPVajkmx}P^TXaesXh@Av^?sk z!JptnF(2cTJ&I=ASp8^t1$>Af$yBiYkkKh_n^&|xEfaD~gh(LD9~R~0XT;!-R{e{g zN>ScYHOzY4=ZvzmW_PnWG`1amixJsnuGVB$qcKBUL=%TQ_Wm1C8kCk z^iqs*u_TpgK7eUbwr3L#O!^V~0+C_7h75>OlXL>SALhU3^tKj!ZGe?wR7JY}UEmqM z&=HZ^3=Uy7Zew^-E> zM^*9QKS4L|&oiK@4U0_Y#DEgZ9r9xxEv{Xy?=9u1>M&}6c_aP#=`N$f@Rt9#W;+&W zfD$G(K^hDNgpDb6w-`JVgp=w2jjrTm{{KZ+a{dQ+`CoJ;=l?=2ztNSP{|mMJMpttF zFVylIUCH^sP|I(0B`4>9(UqLv8wUSDSF-)v#2HGJSHxvT3b}r#Jxar4A^=qTz|*TM z>xQF*gYF3plz<$HB9SeAy-lx4-%#M9b}Cm`(cAyzzZckFA&hhahKjvW(`z>L%fN=t zaz>~Zqkw}dOxwND>tmmcSQwgN1jV9yog@n}7G-&qM`Bp4}>+^TI72Bhf z@MgH<0RQKiC9#4JrV0ID!B3vU8*2kS32+LQO<@3!5Zvr9)?A=Wu>Mvrtb6ErV3UBO z%vZ1iX(R~^-VmN`;mgeyc(kpF6$ahUL;$_OpPnIp{liL)d$S3}C&xMk&DVBfL6=yi zgg;GSI_+YA0*u&C!`C?9FHQs*AVkkXJ41YJKi9k%!fd%98YFI`dzu2;c*;yw`ki3- z``RuCZt7n6HWVM96m~`gb6N3;Fq2xf$~mR<<=(n)j4Qm51{s-FK;~VxQfyH+G-~jBvMn$oLt}%ZhD2vV zve2zfGnQ*rgOjc&M`I7DM1k9^4)rrB=WZ4f<1}?SefO9e>6~ zO$m7_iG~xdr+K=(UvS5D@p+oi*hs}l9tX@+>{C6@X}*BXnwMBpg)P9*fWKRo|Eb%) z!~c6EN0Qnz2}%xFtb-3?f&-t;b_}3%8pt}S$ukFP*9{Jc4EjyCF5R6CHU^GDfk;`m zV5#Vybu4nb#fcnle%3`@pny#vthd8}vm$8)GMKZ_!NT5YOTe-%2)VLK)Qkk+Dv$G(CrQ2jQYpK%&1R zq^u^H8w3lMssHP64`iRb)4R_4smR=A)Yl6=XabSEJ$K?W4YP+L=N0r38^q;Tl&aqp zdyp+bmc;E8oRiMCZ0t}P`q9@9%WKhppZb4LrT;!P=|5g@10?}er*yy%Fu{qrPlFls z4uGAi6ioqL)yu$00vQKzMJ$HHI0f4Y{o1)Ty&m!Iam_(g<^{CqCgMFHH)eLMe=1qy zG2%9#K7O1`jMJR2ZD zEsF`{)5$7f0y~5q^s<^^D&-6I7ZI9LpwAZZKy@K`m`MuV^ptj4RAHe9{=Undaj?9K zMf~Y%9pz(2@66p{}nu`kq#@$oKA9__^ z5KE8~(VH}Dol3sA?TlAR@OvwvV-XNL(JS_v-96RnqaIWp=L`j$(+grA%kzfbrz<;2 zz`;weQE2?l$Zfam@4AtQPEmE@FR%>es#mQ`!W+pKhhh8p3MsxoW}S2vSi_s{7}6c> zsM|1^7t>fEktvH?O(Qgo82qi09A?_ztwgsyOVBk;-hB9bpZsOu6j`$bD1`i|g`@^9 z`-UrYp>uCCae)=72Gek^$7$;J75NsFiOP!hKf(3yi2J{;0%rZU3(}Rc3E$5I7vydc zK-bF@MvRrWe~>R$kB0qssC!=OSxLRy_tv=O;pwEurglJ0Gr7UnKrj|D9QSyN_!^p= zgwtiuvpaNDcE!JLeBMCB$y`r%gKpnbbmSukHvqVOGU9hXYl-bKME+ zM8@6QsgPhe`6%d`!b)QCKBZ5dB+IhX8ONP`UZu1@;ct4txStkJXUT}=f)Y`qd4tJI zI96%P1slPmZp*qi{MqXmWLjUk7!c-lRhoqHhyc%#MLInc$wxHng6*;(SnYFr>zh}9{19$-{WzxSMXm|S_^8feUN@ec*CTng=nSc)zf)n#r0B4IcDcEosWe0bX zrNs*)mrY^)x-cW2hw7_;*?sgd)@8_CaM6AKWRRS@zvX!4cJb6Vr0{_u6Q+yFscd%4 zk|G(!}p(|)EIeAT+Emj@GLDeXk#wDK_ z<^dF0gG)U9>2Xa|`=o*}tWL-fHc*_8wuqsejv~fssF8Wh)AoOF+ z2v!&E{nEst!l+Sk{Gt+F&EX&>u&Hzh!_4Pbs01KE&JsJEz)Q$oeg#EZ59{AV>j=uIh zC%@0cA`}CR-4I{;G3-rF7v(0hWglDtw~qQYPJ-Zl7drt9&LHf#aS8zKPoj<2S-LRW zid4QDLVt{xqid(bl)O@v{fjqV1{X<F8N=DH733scrqgGdyIKJv=!G42-# zyt7J8zUo$Or<|aia3Oq864|uX@_3X02?me$zeN=n)4xR(Fxx+>+YYIjJ>U$W>|Fo5 zl*0w`y)pxA4od$GH*v-QLjz_0R^rUT(y!?V zAqXQ>G|1RjkuAC)H}-T7j4%-DNU9VJL)tU##<6a5Wiu923>D)s6;um>pzf$mAgO$naX{(&ca}eZIk4D>3rPsU%o%P_^~%?*li_D>DN(S7tyf z2-GI#rSqrZY{M!>K?HJ3zb8w9Nl}gGmNywI-&-u3v__LvXv>R}Rx~N!sfA*+f-}iy zDkd^8?i`5Hak#Oylk0TYGr7_e11^vm_mH@(o zjYBZrM|TzPZ9}0pIcZ(77#v2*h`jh zPa}_HZ!_A~PZ%(Op9HiFobLERF^|48K`E*3B4v#XFhja|rlR+8iqC~Y^+LxDwdgqk`JI{k34}meLOu>WHd?O!&?_)tX;{`v0vNVuVY{j@zYg7`Vv;C2|H$!?l-Ko zzT=Uk<1mJ~dY5)1Ha8(|3Ia}#x}BXpZCH1;mFP{ix!Z3*lZV!Gbv3HpL1Jw^ca7^DF- z5xM#@VL&ZFCXSeES2z9<-pKC}{+uc5bwxP9Ka}fLhAHd5`0~YSvJ3Kef%23gI}Y-2 z-CP>sU(_hgD2SE*+T1u?)`f1t!L*eqi2!F2Hj9vVuzevmg|g6wPxZj-+L!)%mgUFH z%lrGEKbqRfKqlT@u$og(_}X3?HCJa$aU|V_3r;ze=Z>DrxA}9|Gp?y5QiQIX=QacZ z4-->MK7OBq+Vo7z8Y>f_0Mf3{Kk~Xqe>2`lb{+?q9*#@7b+UmBKf2loQ#v*W&hod< z>%Duf1Uumxk^l392+Y+|w+kKtlj`gSfeOaW!T#T_QnrSy<2u(ryYeSegp*EUa|LUv zn19eNtF?f+U7zztBg}eJ+M7G9zdok}8_RpGXGxsoTte<&jxzJ*mn|vYTVDzdHa7ex zdQZmH;+A65L#teiYpN#BT7~G#YjncdDHw~W3}5|PM-;-5aDe>I+E4a;JlK7UWjvO!0HWgd>}R601nM&-uiT}W4> z`==S1Z#85e`9W?s3-^?IU{H_|^!%;D^n*%GE6zoVoxBC z-0O03y0?xM+ER#Qe8HTO%vNXXmZ>s??77JUbk&08^TK7#z4$_nrOOBl_YyA?LPJbn zOQFMaNkQH2?%DEOBNl8VY=%q{lru@n=X0q~ z;4`#kBM;$J%0t&3&G^eVG}6Lb1|La|zEUB_RT`79@CEa66)~7?%wO>iHD-cKt!yO% zYVFM9D|X7EG&*|Ev@;ZZbZhR|`DmcB>Sq4(5?TVHD zY|UxX?MY}>m5gAJ?vR>iP}{Z}iM%Sl_)#5Cv&f0ZV6hM2&BS}hsMwpbyEB}YfKek6 zk4gkV_kUx2wWa3x>MpWECr5(W5_i`C*n=a%iPnZ?lz$e0;W;_0cYcb3A&iWn_&x_7(Wp?vIhn)E61}eT?w(%{0=&eXBuvXWdzQmXB@SGU+$$u7zX%oeT!m0? z01tfheXz6LR~`2IK@zJZ2jkq@7p>>7hKK&U|NhK5{?V2=fQsgzS)e!pZ2vhWExCwd0m45AfyPm9V^)OUK7k`Qs>gp^!UUWp9cQkxQ1_HBNg%rA8 zR&LjylTJ6fH%pV;)?aUbPv2JZbUI^poKd{63y;3i43uY5@3F0Ly8Ns}bE^syj}w!)b+{Yv|r zcg6u%sN}0+f?pQ>SH$Q%6nLHk0Pq@@oIWN5CAtVfSDb_zOU-zL16%=iHXiL znWnT2t5dOGx-2dwQs%PKK6(;nhGI1{Gc^;_cjl6fOxe@bNNEoi`_^E?o6~CO-LZff zo3K)%uz>VeHeRF`CQ#Zk{=@gfX@;0|vVn?_Klx74AYw~#9rYd{jqw2z`ymN3`R`ZJ zua(Z0fq`5{h1p+1h6gnkED2N>EPvpk9J1>gGrTZi4bClMu2RWW3hJh12oly4u9gTQ zKo3sjIddJFPBTGStSfrSGv4H~v=AlVLWo6~s$y`g)u1-Bg-TXDVTZ!7%7;u}`mC0) zqs_LK5keKR*bgKD342=nV2=U=!ITy0gj~eJEsSalWlf@`+k4mRe=#Rw48C~x%$~+; zY8fG1E2)+*qW$Iy_v><+$ab=~$68qz8SG|-_My3;*y6V6%{=uEBdoWs!jfCpXky`1 z1hE|WQa5;px%&IhbGRHp1}=n)FRNLyx+=}yLB*T*>bA831*g=#dBfUv>GTKGTbmMw zMuvT|E7MQip(cy!X6O!}$qd2f1EAz&NDF{W2pfonWtKeY(wvDEQdLE+2UO{nszxPP z89-k!Pa!CaE^Z}FhPFb2hG!pd6rD;DF_S1>@S8#;E0qGhG=&}2wZw0x z==3iIzE1rToKP;IV&R&xfS?(35qG zC<9T?FJ#Oz3y0#AKg00r41pRYO4~SAc0Ys;{##)>SXwTO4K2?t! zZ!-UO?zSIm%;j@g)i;^2)m`^MqPNgkv`QY-Izyjrmt12{b}(wI*`C6u3|iSe`2?g0 z53Wj8tGH7mYrs=KVyo!+RxO&tDG)s8zoUDODSI2xzTZw%bdHc%B z3Q<){M-M`cASZz!uX?-R%Bu65ZZBWYSR?By2C zjUkobKG$V6BDU3XAa7Akc@EpET(@OSq3}05_CQfGg72nYc7}3*eQkI^sg0qT_PKu# zmul)NC_Nyt`8w&J?TH0u6yvBfD{-kv;hQzPy)Pk81wl`rTz6?zq41CuJ4EiUSV+?R zjf8UA4H$ATPYWBQN0 zhTzKL^Mq}fsg)ne*ZB{81O}aL+%x&xQ^$3tj@if^8F2Q09g`4as2#t>lq7K%2}%dB z6@E;t;K#&9vq9uOL_?A?>zIm(=oW~PzNQ^2l~D@Zl2ic|GqyyNjHz!Aff$ zy>aX=&>4ype|r3s{`1_|;vZfOgenmtUyfMQbZn;Km^kT(!$|}iGIe)t#d?25(fUQF zV^DFQten7@@)s@)x2@FaEKCN@jn3EPT}JAAS>fXd8EK>{X_MRx`k8pa%AMn#ZwUvn zC^ar1@2WVVxpudN%6m+_I@n8Zd&dh`f0-aD1X3xTU_&}vFF&}juiqmK)-^YrENKLF z%6M>Vl>1+--hOEOs%?t^xnn31I^M*Zw0CppJ%~QccfZ};2&5GYYVZKkb19)$qQ1;i zidlckg$>V0_}=84+Y#P+kLHQCid5+Y(C&K`S&Gij#keC`Jl~xpmI)|AM#j=yo1fP@ zfS+6&EIr4V+u9UMzh0~i;tX=SE z3$(qR7yfpc0#j3as%$uX`_CiCi0|IAT`Oty50?1#gcLOfQyK-0Huhr#rnR`&w=4f_ zW~K%el}h^hEc@<$I6xMTk}UT^*-79lD?ZL{8izZq*vnr3NvY#z4)Nqy19#zjgEb+3 z{*I7oH|96bM_Z437}ywnhX(5xI^Fy10NZQo&81n;=%VBJ_&2G~>a|z+XBi*O0#_d= zyO{3d4m=$4&c?k!5Y<6Fq5FcH?{9{+N5w7j^KG^7zn^{c=Jqd8P(Vej)!M*lvzuCFCT9OF%#hVZuBZ@J^30t0@^wCE@k1Ag)$ zxlA&HsA!VPQqM&)(xGG8l?LDttbu9k<7Pb-6~$NBc2=N45t-LHs~WRO9sljOTYijT z_Bgdkj0C+~%ku$i%eg!ALhc8(_w;xv4RmDy8E|R z|3&E1DLO{@xZmI*n=->Sb5Em`W3Q}~LG|*SaBaut-I<=TTg{pl6(*p%+js`Osq$v! zf?nWYV2hQ`M?K(f(^&t_lDX!BVw|gxMJ&Vq1#=sO-s#?G+vrZya~$aWNbKI}{h`z6 ze)V-{aC9-Hi&0nq_TBLl(*fl7w4-U<;+|({UO_2y$0BUGEf$k_6>aW7V2f`%f9K>u zJhd;3D@s z4}RhM73BBA4a_lefB78m)ppDE3eA&^<+3}S*4XtjTVH0v$qk#uJ>AvgiyNWIpQ_=Y zgbHX(5N#DS(!U$!fwm=JC_#p&&;cL+|J?sc2L9cU$c_{6Prb?izabF-@Sh$;tR!44 zEdL4EK7xPo?EjGI>>t5@nRSkj;J?f|=ST2gW}WLJ;QWU_XTth7r~bb(kB+SD_#wAF=7wSt~^wxqtpCs_iEmzy&e7;)=Z2^4dfJ~pi?O8cLUEzIf(Q+*&ziY+gbx0{+5!au=4(t_xy-1D z-*Oo=h07tgkKr4HMLLD6Nc`&b0>5E}Ok^OW(0EL_J(UOcxbCg(6#*lMEw6#pYxT*U z861kNuXb-C-&1E6;a`UYM5TYXmI~ubaFYJ98KlOSdd?c&tW)0sNBB(dEWI`PJ4NO% z2jLa{79-G!YIT>eO-9cUXh+LhZJIh=dbS`w*~!j30k%x|2CC#M;AlL3vnHbNkb5zk zMKquVWiTtzsV8=4l&bs~RAB~U2&Jyr+sHhhH7vDTfC_d{JUZvzUGyKeD2J7sV`ZxBSNY3@5jZe6^bILj zKs=-^l(tIOAEv14g>ZpBjKhs)Gxoy`G&XD(bay*Qx^%y>;%{HAbmxKh!#EQb2u*A| zasWQng*O%!i;Cs+6_WmRZ+8pIj9DPw11ODX+*bzBf7?X>;mRZ)1aPN0icQ$`ys$ z2qib5ts?&j`B7c$srdQ8sy(&w5t{+wEIm;95GckWu%b}JJ#Z)?F#I9p{_fNDz_$tV z)BmSc{b%$~2TG0)9-(6!iuwx=`+r+evVD}POoL$Pfwqo2T-YDBpnn{*jaV@^<4lxD z;w&tFrb|&<^Vw&h;A=bo{$5qZfK&hOKbb-#w7aDI*A2}GR+1#Ly}g~k{r)mBaykdz zBHBERI1!pZ%Xp+<0J;Q9fV^IXbqBhB?wrz9VW&I|(UW($AVOdVGpR>LjfcuSr07q#?9%+y)Lyd5U@;@OnDIsrLnN1?#762LG#h_`~(de>zZiXZ+S<}Bw_2V@{ zIB8pdJ~C?QzLsMDxPAV$l#hM*R7g3#=L zF<)um(l-&K9MP`G@c6QNk%kiggCA~OGHLs=;Jz0nbNBZ&|48bZGoWJmqgNszOo-M~ zes%{JnzsM6pNy#}x1$?{quu=pax!&_$M4Fltxf$?mRpv4kJ9h~is4&|8cWp$0as^pfjHO0$V}XAsRfvv1Nnk^k6p?WV zXxjQYGTTZcC1ssmk0Z57Pf7&Ii0}XmWbwiMb?RuxM}JJJBlv4k-P={&%=FG!WmE3B z!(!uAI#qM7qGjL2kxN5b-KX-?RK;!L>@cX+t5{}PcB6E0E^J1GcOo-I-e1v?DK2t) zOI9cDYn~Y6iDweZ`^X8Itj};<|kD;5X@&Swb|*usZ_ zVEyz#8Yo|6%40ne-x#2V9{OY7P)AFq@4%sSm8m6Cz-Pd zNU`SGh^jJIe?2TO;I|y+oC$1uOm3M;d*@hsdtzI}>4LcIB~J-ES)1j=Sz_IEA6~$b zh#}Ewn`(`$Oq=Km{f`rsk9z7Yw--!4R!=>}4~by8O|m|&?iO$^Otr$B_Qpfhw`r_V zTStvDNe{nf=__;1@zee@g|!VV`^LUG=Y#15Q={Qxo6+6kSH-21N*BGj`ikaqZSf#> z%e!C?1}{U6#iO4f--m}U6d75sN%Qj*hSbZeOs*O4N4pP$!S%o_Oha%*`l+=67D(%* z0Dm!D(>gTiZVreQ3qIOPE9>Y#r^tpkrZ?QsVBw01z8cMyg2n|%edRu@Z%2F8xjCWl zAzme1eLcv^1$C&%GUhM0w7x{MR^!hX!9-E7^M1G>s#S<(ElxfgR`tHTHBILF18jO< zz#*7#Jyf0^3oA31t_O@7v<=@9vPeinsZ-Dt1ggdXS_I*ghT#&3CUP(B? z^zFPP(ljcoqi|~BM@C~sFuEZ>RA$$@k47fAMqdu#tPq6}nCJo0gPdE&ES_X;zWWmIi?gY$V*;s#q&*tQOH2RjX zx2nfwBY}7Ajoy}n!_Z#uYd5nLydzrYzo-F@&9iRm70#EgW7F2M9=wxr- z$>vCXHxs2~txj{-s%T|YWj=Rv3-&X2wq(+v92?Nk~(@xrZFNQvuYM@9}pDBq!5r?}4gT8WY5tpLNVE*qd z1b@nu!6aHS&CeyW4`-=6O|b$!PUSd@LXt8H9^x5(nqu@bLSAZ|ga1 zLIdHyweItO7NV3N@gG)-A0Q%^a~oQ5*iY_=26uK;ZFa8gC#^N1spR`L15^wgcWuf1OHrv=*#s^k zd7*s2!a|4TPmr5ZBYBx|qLTQ;v;@&%Crdy^fU@GCtO_ZuH5=3)B43>#M#_TZ2Wg4b zVZzstmPEuf;yp=W=py@|-5W{+o55`O>@n@p^ymxmBzhBFt2)Y3`7Qj_o5I~3E=}Fb zBch@o!3|)(gc4n9`6ha6{MBrVlufxr`BipO>8}@L!+`>BH;7*o&xGH>2jn;!?AFjyJ@Y>%>@z3yxp_T^j@l_LvmDQn+@BB1ebBX# z|6(+~V^bahSKQlPGW_lgEcUlNnqOA1J9HdkH%iXZ#V<~w^ys|{p9gyC8)kGDobL(Z z-yyxCQ^o(C{!sjamWSZ_zs*B2!M;Afip(yAr=ilB{7DN7ho`!4hUhigZF{_oS@8`{kCvKeLu0*;@WlL z*L47xbj@+ySWVf(xqh6N17tfo?gqauFJ~Z&;rY-5yV53G>kmCOJm-^YYb}m%fxdIO zNB+bz;}e8S*&Ug)ovhg@M>gXFz2$-#>XzX7$1&~14ZM~Ho`b~$EBTDZy~|d@&EyGk zZ9m%FFBXYbpzGniF&@w(k@jjbBiV0DKU=O_a|PhrcgnAu6P_MEH5Mx=av*8@FOBBG zCt4ag<3Xz0Gbwuuq$9@8Pa6Afz;s3Rm8OznK_3S*@N2s><9mC`)wl1(f2}$%cI-n>XKus9RSmkk%dDo#kTM zFo(JxlQvqx439^6XraVe`YCmfLHwTC^}ASXqC(7sC0DS`}aEI;|7%EN>98DgT(FN%R= zLSYLt!Iq}OtCa+UliHntLbXCewsBwtV~G-gPXx-SLI%rN&FN~=Q`qUFU}8hC! zHx1_jH5ul@@X07Qzoj}JVyNfX2N#aBRLH}WNoWS^2}PD|V8zC|=d6Y@=uWS+ zVNx3S(JtxeRfKWLiU=knW^?DlRZZYb>EppWYiSOlqQi$7ubFjo$xAA`} z-9LW-M*HSKNM?ZOiM#WIzl$m3=Sk`TK8N)GBAJJouqHE3T1YrXZBKIEnxaK%cHubU z`74{lpuz8(sEasL8&tv}AIWMU_Umb`QW?-at{zb=;jxPfMIb)vI$#|{ezn|->h!~q zc+V_hEYeL?*W1>$y^aT(;c0Z=4b|RyBG8ltC=53K1fwBegR$vfbUE)~POgT-x-2~$ zDdu$3ZG!dan=MmC*PY$!%5z6Y+j}BxI;2r6Qq#EFmj_ly3RLpU*_&$J3R- z6yrYQl-5&vKAju|`V7!l3OnnrB!zen_P-~H%oa*ST4Mo0tt{{hNQGu$Fg%Qf2**)W za730}VClG18P#-*e(&{i25cuUZ?(<7^U3Jgs?4--(`!u5gj=3W&bXX_Q6%-eDp7kU ziH;1XK$AKV+OcKnfjaUbQG{rIxF8}RZ)fa-YOxdT-qZKDXBmHg+DU_ndK7jJ&ik8; zh~qwNr)PSbKMwAFYi9WAGP-e6oFaVwK`vf7N+$iqNa|AhBxROSXBMxl=Q zcU%*U>%(z=FY37J)k5Pl=TE~L@~mY3KCDTsi7i+&DOc)vWfXYoCSRAlNXk3&igtKmT~_W%Oq#YgVx#ggcJRi zT|cU>bi&b8pXGZJz7I_`8Nmh662!=Ey_EZF+{>FJ&fU_*vP9!xuaW=CixlLz-&bFi zH9m8oJu$uN+pDa|guCdn5JUi2&Q&dCb7B?n#q$FQq?yKj*l8tKH@Q@C6LnYv6V>ko zM;fo}1XFYgu6-4$c%JCq8HNVsy)4FKe#h=ky`JU_g_T!f+?jEG`pyK@GzMe}?)cUS zyRlQ(9>~S?3HQ&ttxY6YYD28|o(5=p~X5s9w80QG!u&z94d&EY@nM^e^e8&xBf zp^ralzAjjtJKX6y(dq-E*JKba@{*c=iK2dQ zx@Y8A_E$29Uqrps(W~l$F~?dM{xGT!ZB58&N();?ZFMM`&B`SpBS)y7>$pbw`64W_ zO+C%7WHYgIAtZ`cfs_>Fb!M~pIK<#xwa552a!T?Ux%rmi0HJ0>mPe~UghHK^*f;q^GIT zTT>K8Wc)d+kUTlt0_hA0{81QOCrqSRQUY~U!D+h%wg$jTfgYtqJ>epw`GiuPM%k>n zFKjs9uoI_qx!4vAB7cT|*@GZY;~p}=E)q5$F6n=Q$M2`wX+_?VnB-k@IPv;rP?wM_%MRbH^tu&{36toauRIqZO6GNSFY_5E-t%_3Gq?)X5Qt1Zo|{_j&ua zkAhCEPTiH;{7)Dv1=0->Y9Y&_jMZ#sVj>eGy#6F{GkoHev2jVMu90}>fFh^ks9gEm zj7>45!lIqkv!txUZE{pidvC`HOsvjzRNx;f}9EE}{AGivOLgcAJ z)!LED%{s8$d;Cu898Oq%z8A+ScH+QApiu@WWzz)HJ0bFeB@0>)E~SlbqmX4XO3?bkDK>zXiwgiL-#R!7 zICLt{uGWj^4HTv7ckyNmo})33)z77iqLs$fW#s5yCtK56OXk7Kc_WF_FE$=g=l}o@ zp|}=ydJSMWb@9ZZyK-z0n&gOvty%=fHOo$?;-K^3qeEv~oGI?_%}XSw2xir8EdKH#epk;$ksJdb(K}^=+&G`KzgM(NnAp#C0Q!HH&d47y`GY zBIQl;ZeAWuzo1I!+_dP&>UrwHG7ZbeyVCxQeM48p+$2IUU38An7`@FsA^?7ri(~{{FrgL zRqr^ra@@H8!YN1=)_+uKvXCtQN%gn^A9wyyXmKtaxoYg;!Cao?9)c!*8@>`ejsI%+ zn&mEI&A0H2X3m(d29s7tSs7Ja#7y<>XHrNNGbt&btd>}W9lUNbs;Dq#srxVXh4j1o z9QJI0>%Ld+J>dSk;iBLAJs^A8i;fJxf1Ezo++DHp)>756@wcU; zEf9Z2Wi|WEZ+R_+=nlnF%){n|Kk2@9uYY>1D+g!dU_74z0HXYoJ6;F-xa{#begF6e z2=2WA?*I=-Ss~y(S)O_84y+dVxDvbI*pfY%E&sGIgw14G+!sDJbL&R+HcWh-kZcb2 zx^NTZNk<8_F{0vB5ufuvVG59*M+Uqrl7cp4( z1zK1sV`e5tz=?7;cf4SDAuwmo&7w4by6)I!f8~+G(dVTWPd9P;-ql!Ia^c`+A1P}B z0Ri<|`1oiR#0WPc0(o$a?GA>WOO$l$UcQ)=q;aWt2WvSS2-fkYs|RblQ;mxF0Cx+K z7$^I^K9bqOD3{v~z1GAQ-0UHnIKs({8at2ZF9@IUqjt0zV^<)m6?GLGd(`oT))D`n zOed^1w1;HuLa(J0zpHqX*s6WF(aBYhZ*36bkLE7z*rkHU%BwK(*4_xMWV2sF6U8WR z4;PC&s`O=iKNt(Gz*Cr9vb1QRI2<@T?99EfMjzu2amPj z!{C>j%oj7dS-c0y9M$`?BaqP5Altz=uH<6tYVX2ShM=SW}-I zPWjZNI1LHdc;7Ncf%kK19Al4C-OGlqmMv3AT zdDv++k@W*@)Twj~j=Gcz(zd%|F~#92R+?v*p)a#{n68dKPo6%iFb~J>IV>X-4EX)c zX!1nWHyIOU6o{rcK=BN-N^FTib?VE&0wVF%6yndM3ngMXp=tUvsJZI=vIsOW*2`%- zHQNj&MIo|nE(rxzfye+-`blqbYni5Qm8l-6TgtUYLSj zmfbGEpoT@7l7V!jLXydhLn1NHwS3!ECXphVhK35GfosfPAmTUfXC<@q8p^Y_#yytK zMCTGg*dHZFS|${TMN-J~9Eu89g~JdGQ)q{oOtCyX7G^L3@K{6@9uxaxERP*qrWd8cu)U{3j&iJ{uHi=p^zBs0 zUu_}TT%G~kiP^}LmtuM0Lo3IEAN=t5*JH#(yj8~elE~M&0TZey1<=kE(RDzqIsCo_ z!g4Zpw1@K$H?00(I4lXUF~4Ks$}LXeCu0 z!OpK!z;}`Z*(6i7)K1^pd3D|L8RhN1j;n()y z&2aE5`Qzg3Bl^P5(IL~=0wI3 z*Wug~ZO!oI^wsLsgzD>XQ@zXA!UI=Ji(hF8wf+fR7z-cUbF=UB2Tz`BU6ye}58;?J zI-2YdIV4LGDLrqQpZwPr@9`UoHkB_oxhwmXWKCE6@FU(}(ASOOl+YH^5|>PW4vq+4 zL}UNNB#!s5^4K2``LC2N0|bMS6u^i6kL3W({{YO7;){O)CgXy(qoz70^yMvD^kz>V zw;wlaXOv8|vNHJ=YL=3?GbHWx5@f2Vd5+EiTjLxgV=+>(Qc0s6w1@;ITFj45tDC_V zD(?5SjU1l^gWin;_Y!~Qk0SsV5d}v1$*VB{Nb%01cY@<$NMrGs@RBe29yq3Id97_& zl06oUk^Y?wSRW4Fh-vrbRzx~z*lzb1sNEQgf_l7^6Z8Nw$Gs*rm2e_!sR&WttRXtXoTbm!zK8STBb=1a-gjuuEd*J0 zUjp>6Ko1YDKV+Sw*Q)6xqOUw!#^=Kf6s2Yng;%3ac_HIjf>9x5>ImGP#=>X(x zKfswsHo1vm{U>$_%G%-)on|Q~rsQW-mGGH7pbi1?h>DK}hjd`4$pe`vHT$c#B>M#0 z>0FDzs=9>ky`E_Y%jqcZWPUJI2hSh@B8?)bRM(4eEm!^XO;jwtlbVg~!r7=d{>oI$ zL4oXKjW4vh{I3DmE(5jQz4iuy)iP3kjv|U=NiU(Ywn(lU+mGBj!y~&sIU}J z{QaN_l@KbP8DtB{BN;ba>l$1?aV8u3-bVDAa_9$@je&JGQPL>E2u0Vu9VF&S z5*NE|M=ip4)z%5y$kx!|3cwO8-UzZ%;8!2^q-hEX8Dk#N8*Hnmhl`kY3JC+WH!Og- zSmp=vgy=15tT-u56mZcWRQN0RX1JB568V)YA2&ox%HJPJg{h=z8~hGBZ}8~u_9ELc{C zB!@a>4|<8J7ak8$2Uu~Z9*N)o9bx9XuHo|QV|YJ4fV9HV-W7iUMAXJRiT^y29b+K3 z*45Z4b~#RmkzbZIV3)CQQUfq?;+4tPCTf)S(ir6~;*si$ztJpO@v$G2?2=V~SON@Z zDhz&eBva`?%n+oXkyfNTg3ee1ewMn{C`ZM?Dn+GW;QcNkwl0#pZVee(yB=RXOfogx zB5uPnHb#Z0<&>QLjyf zr?NT*`0SwNMS{so;xrra4vm1|&Yaiv3SG1yLDW#Ta}JNBb@-Gn`usi!h%Y$QM3?N< zQF!zzU805#E2J*iZ)2>w9Pc;$c}XO|8V*j;(LxO~viieLx}+yaS0c!pL@z_sm?oL6 zb^jpwatnIG(I?$HvrZ8PPL$&q#M>M$sS$m+Pq^Kxj)t`eOm?qy7q~s+=Z!*kRR2|X9iBZctayv|w zZUo5!=ypxe>w_5Ermv{2=o6$=*)RRibr`qj5w8+gkKYc^T ze%sXO5S>w4LEwPR7rSb0I)w3)WYR*KYcy{>Rm;id#I*GL<>^2OG~UIMsF1OXXs_eo zM04XHJ2frP2jYm_I2Q z)oXuEI+%iIn_ratfQ(KfH&Z_`T*tdV58fd6(|WUwVnSN#J_}!|2gH_irX-ZI&mY8R zNSwF49ADOL8*P*paiOvM<>O9r;&jmr3+i`Eok4+$QYJ=>D4{794suoH;P#NS#Kb?r zjWVq2l#fg86ZcE*dPsOH=&o5KMx`+y#0s3*VQW{>>Y>_>026ew-A3+6MPedpmC{*_ zX!o0w$*M*N72A-aN0#?9SyG3e}G#4@xz`*KIA<_yHR4FKndt5=^QqxmyvY zIt6<06EeKbzZfYwYuRPzNbyM~v^T@CeG@BVM~|=NYO1(12KKpllwVL_k$D{cR~;8f zfd~c$lI36hKK^Gu$qM*4JPIrvxT^mP9!no#ACxVpoU8H*3QWW(AZLoX%i)WJMd75& zjzW|~V^E6_>D0x=Enl`Qf%j|E>3ojod>2_Ai#k4MejD4r9mkf>02`_K=Qny|+y`~1 zF8f(SVF?^MfUH;|KqEaAY<;PKU&r!d&KokIjgy+srB4Q^^9pn$@wqpDe{&-mW}*+$ z8wLqk9D;$><0pOIJyj23UC#w4xTw z8}Uau4H<9RYuH(a{nM8l>GSci*G2Hn8GmOAUY@(!&v0NyrJHGiNCR_6+4gzz*3xb9 zII(v{L&a^$sz}tUHNi;ipVN>RwFU3|X!1LfZ@Q(}yS(lfdkHC{)~FZ?r;U3Qt3UrV zW_@a7Zc8Rct7_U;vkg--R%h}iC8=|nzJ6xl0Xa@umRo_?;@B#L68aO}OFa~0zeSD~ z=-G}rLw5kv9&zfLt3M}505YOn?Hq=GKEyMOqTJEnPgE?t{#JKeEy~*%uIh@X^wb|c z5bU65TFO##Be1STn5KSc&Du~jj9T~2N$y$d|8 zx^B)Wa?7iPzX1r5YzdIVNnIZpKCMH~ zeLot=Je(5enwz_Lz>f#AtWzSe@R5OJAbQoF-ci$syAWjUQ7%O#G%QJs!bf56|CUhJ zbeyMfG=#Y70q7;Jca^H_%#}iPeO-U4@T$g80zyXN4W7q(k%vz&jHo`~#*(!h7I~>_bg;Aw0ix6O- zq2WAbYyrcrF#HX>@9cBjV?Je+b`#}JrmGH`IV)x7NOt+-IN!mcn|7Lf=GbVpqwBKf z;L+r8AIMwQm@auFQgL_Mt|jS6#*sY*Nu8u0r;qmedx%F2N$l;U3#if8*4f(slAaSd z@nKzm@@`AdMh548Mvtq7scN9&rk*(@TKwDIy5{n|Y9w%8yE}tUR{_#d!uPYlAr0l` z1btuW9=9|Wni0N&8CoXbT;!`i9WAPAzUroToevJgDRY#=6y!yoo?QzU&L<0EA#G*E z8jSd#-k<48s!wV-2kKK-1k8gUVx!1`8I7Uo;u{Ca6X;W#R4*JJdsy}c*1|?k9;I_f z$6hpoYKShq2f`WUe_}rGYI)(MS)Tri<;>vCaKoKXF-KY6;}n*2ksl1NQ>x{PhDSk} zUV4zzSfB9W*cAxf`iwk*wX%#{Hu}O{R_micl^4Fv3aQ1YWC3ODH!;qbAb@NR>@SsM zk=Pu=@St@6yE!J+TvlEZ|7z57XdTfo_atlT4$4&Km8fbdEb+434Gd5PAIN%14z$Lu z(XgqgUhl5+FPs*2gNA6@g4dqsxZGtr7Mg9O51a|WIdfrgC0ytNXY+CdN`mvG34w+j z?aVssergr(v}l905)fjGRztgEl!jY z)4g>u&=N=8ZzKX=I@H|sXoQ9<{8_)mBEq-_20v0zxHG87NVDkJ zxr>O?2M+DmYJ#N%xbZ8PYIyO@XS6w_KV#mN$`z)a5PO*ZdFLh1h~fdJzafm@1q>g~ zD*lZsBpfqB!KQZ|CWn5-3dK4|{b8o;vDmnXIeFlC+hC%QLC|Zh#%E~%MuLL1PAyXP zNlZ~gnLMhVE~N&NA$uljzw)P+aG7yJ9bAf%A(nc6-LkgB_H~Vm7#g<4f(jhZQFjwb zM@DIUp&}X83t1OWrHY1VYUtPjjarx|8*(D`}AVKwH579h!hhu;*WCAOy>Qedqp163E5zUzF4SK)DZw z?{?J}?5I^Ns!c{ml1ft$&GrM1%{*Y2ih{I78qTbBjNKUSnt}}ai~7@P+vajx-tk;2JsrO)m0hL2*6{p8<2dOV2}a!)lD&+~RZY3fmrcixR_6Roa~<|ORB>AX4l7 z$ukWL0oS9Bgcuxf{h5?&qGC?R14qWy(aq%kkekjzes6}Nq92-~;>77#^?6pXKW!WF zLA~wr9Z6-&$El{*jj02;o*87mm~uVl?P|fWQ#!K;NJhN7Gv5HVcE?(3x#PxTW`gez z^}guhO{R6(!LYSaFY?^Du(km}6Ll|F%5Qo}hFxQhmi6z#OzwMx&Aoti(3i|4ye-x4 zsb}_sguMR88!(C*FZzuS8)KmSUfCeR@yo{8!1E(m?#q^!_1hE6$9}q*yFVYspUlui z`~_uKKXyenk@Rt#6HiT>9^FDAa8*B#zG%wj(#J`H5-7r z8!DG~P=9aZz7=R`6L>=O8%Hl6t}F&}D~sC+eRp;zNrlO5c9uca`C>$~9DAm^sy0O+ z=@kVW${;vw3tymVw1i40z;ZsJZ5S9SqGGTMiJwk$Tm&RAz5Ugk)-YMW>*BsKav!CA zPeWg0R$64#AS?f>V3GEM>0nq>H4$|Ie(?53RVg3+R-9ba7O7^Y!^!NgRaXg=py8@8 zD)|f7hlFBfY_z_(orOCV-Az*=Oh_h+34iL|IZHxZ!;x^?(c;ZhYO|@X$@3fS+-_Qq z@#&A==`c{6pi*2q>H)sCCcuAI+>g6{+wF^uW%EkWZ>K_3g%x>6g@x0qGM@K5DdQJBIsxaXQH9f8yV&bj=7Q#jW9j9p@JYS!7MD`govXx)O4d?R; z8hJW@@^#nl$eUgTAhN}YZFeO_Gd=XN5F&o8e)I88u& z9y0};CeAl;cRL&CKk&pHkHeT8GT$K%6v4ED7rO+ne-}AO))W0@oh*mYty(rlYI^71l3}XnlS}AE}^zC(%}23 zj_UyR_EJICa@*KVyBjrYB%bDUf#2oWlj1G-d(Jtkr{g-#zC|D-WyFvr>S(aSE)Yl|}P^txEPB*ierPtJ*=FZe)CS z!Q;;VhpKZ5&!makb!^+VZA@(2$;8eZPWZ;QZ6_1k=ESz`WMb|4{(tX-ebxunT~~Eg zud065y6%TXr+B_#zB2ob$>O3xGEj7bOyNL1=7(H7OvciBf#|YL09NnQd?T9lLYvx} zY`#8*0U5Z2-S5&6enIm~C-blHQLiPmZWV^^|><7y<4yU z{4Vpc=QEmo^`4NvvEjK%o?|_!keWZ590%0A~4TN(g^&XLHn&vsJdf2${+RpZt{Covo z(w-Q2SGLdm%>=Q1z|cRc%RPAH5Gn0`%q^;ZL{e*>GnnSgP)z9`wkXg?iot#e7CSi} z5pd!sQxrHI^%9cx@xXdP?ZUCM<31TyZfE6kX=A&rA`?~N3?O8H)L>rH!a!)7I}7=2 z|8S}4{Lv9_H$~{gMOO$OyZU#fW)lMl6}0|{1lANU-rZ7gcg0vT{kkp6VCBr1nVpo6 z@ds`tF2yr~X140Slc0AD{xYS?*=5jd_e(~4@}bf&gNK;GLxohMKR%U|B(!J#Sa9mf zyauU;5Am{4tN;#~hP6$mW2-yqdwWuN!#OPDeAh`2%fKqbEl%~{-&mRiAr$g| zTa*JNYg#BF6yE>SDbofCq4;I+K|bX{?#)3?e}Qa;fh;70jOBs!RDd+qgH*JGDmP7QZMPSd`25 zL=o`V@B#Le(`Ok>bw)~aNVK5Yn1nldz*Gsaxjl`YGF2ANlyXP3$CXjX> zV97i=B1;91{Ps-SBO-cgo8h^`xNaQ|N`CBc-Q{*7`|3332;23KbJ;K(v`}D2tw4xM zj^-Js00j^5gjYLixz4bbGGD=Jb<`aVNg#Gq?gQ#AQLbebn`bHOc8^CbSJvIOrR!07 zhrmqyLP?hNQiSPcd?+n<r^5OA4QNm+!l_&Z6&{1S>E$ANJ?woaFiGwNF4R} z)13SUngmxw|HH*@%S_8dbe3Y1H_|^9^<;0VSXA@ge3&EhzVDYuP_II%?G5bXTH$47 z7ZY$&Te&G--r;a-q6K}bpumPK)?+2y;R*S0{(dj8ZDLv_%$8Up%+;qvIwF00yjEAx zG{IPZ<;Y$uv``GWm?jaM)J?Nwkh^k$r>AMB3(I+vVsvF+@Vk- z`$?whVR^YAF#7Zd%BcU&fLWsI61L~UQavzv2%JoFl9j`DQI17t5{kPLcH=Vi!$l=e zw=0vv|5sFNz+k97{5mcN4zLJmVF9y5^RIZ zQ`>NIU0_~R)XhnsAsy}Zw3SK<)(jk!I$U{Se_v>@|Hbg}R~c^ra7kmzxF@5AK?J%q zuPp(1{){hnp?|Zk#5+CzS(9}CEMdlc+g5l0onOPshRL*N^G4F{52)SPIpqUPHH3CIHmVht7POHv-HStdWWrv>p#j19iRJkv zF*tNv^B~ALSbeAfq{b9XoIpSu#MwaD;1Yyo(oFAB? z4R^KELF_?zpv?Msj`D{8(maTcV6C6E;oH-BPUP~uW3*EJ9^qI%`@pMxKKz5Eiuf@2?=>-j(6=mxwEBC{#+rw0TJ1$Rsl>K zNr%zyw%V`;BQZQWSwI!UTCOFu| z5YT+N2^v8EUH{5(Ve(K(4}80N&?Dw@iLtPJz-;ed*w#RaE+5`j8fd-w3MpMkeG{I*Bgd-K5fouR+qilLbvHNRCwbBA@RUv|5DtGmo)0!GD&nf1o8oZ zN$42`VuLkATkTNe>g74JOtcY=>FW!UdLQpP@ZQcO#;Ia zj`}TU-O);=Oio8AZDby+WR!{DLIKn9u69C8+p-P!T{LQ)F#9tXPZEu7@OVsBt3-KH z8?Yln32xYBuPrpgclkAcB|v{jI9Nh=neFWv0egH|-vK63bnFT~;De3ff3{*X&we6E zV44?0re@z}g7=Ad6$Be$KAs}}nyTa$!b?+NyG`EJ!<2^C8xH=MP5rnFC~a3@@x%BD z&4{IXk|pHG5A@M6;MvB(Hsq*{QNva{F<+qn!-}%AHyOpaf;V=qcq)5opc$XSOLfIu z3Y0L4fE+*SAFwvE?0_ zu3r(=eK56lh3W64CyM9Fb0?0ByAygv^s8sHuE?N>4Q|hRP`%1yFDX0(>ev2|r+_lH zFvIArVjx^a2h-y<8R<&Rlr8g0@H7J*7I6CHWSjADJEnbh7`v;#-IekJH`D7+_2tPdMH+{%VJ6cY{ElM5Pajw0x>4%t{AhLe z5V4DrJszcgVhpDvogzE3x{MlC$elsfLvq!|_+jbCi1L zq{2z;@Z?*m*&f2iE?r53>awYU>Kc{9=Wg)D!TR!6^XB%QiyeyHv}>)DV~&V%-iU4T zTLybf!*AyAW^49)p4lXoN=H(bec+a^0$P6@eS7|Z_tzuWOE4%mC)L_wAu9_q>h0r>k$-)|2SFy5eBfmMGsDFp&b`3sGNSFgzP> ztmNqvzJn%UK=nG#ylZdL%yWHQ@QDiqW3+VrH}TC_ApV!{)z51}r4L-;N5ChtLue>~ z$fWp=(#ml13Z+LR2$wjC2dL0lR6fe)x#4^B>T8jWWuwB_1+1c(rRrbV9 zIu~bCx0R{kJ*O3Drtdl~4uLh-Tw~mnHYDBKzd!Z#R|IvVY_l%f57FRb;V}O=C@Q_e zC(vprWZy1~R^~S`MIX<^#@$^cekQyPzW-eGC1J@z2uf|Rhn{TKGH}u<=UVOI=X|tp zybCFXChw}A-?yv~ujh;}_n#E)Ko959o$~}mujb^?KQ1Dksvc<4?0XDPCW)=l!@-90 znL^$&CIt^W{%eZ5Wci_&te&dOWMz~<#IO(j`cdl9=zcKZTfWAmNhF)W2IXN~qdJ(- zQGWD$IWy){zaZvU7LfY$8U`btLn!V;_bs>)EPs;z+geE{TO)X#v%p^;nzbfnOpYvd zv6|_jqiIIx_rtdH=Sz96W#gX5*iBJpiD71FS~7-y5d3UOD_yKt(?ENxi8p;~b;Z!1 zt<;GXeAiRS8?@$cCYqfGEnyAJ9VC+ZFLz^Em>{5$QTK~53^?wB=>O_;Yze675NzE4 z3};X;!D)acI&v>v3ljY93Goz|Mx=DtP z0e(CE!fV9zjtfY=+#d8-{w1;ilkSOV)-!w@zmQ&N0pmC`MrIaxvQo$Wg>2(}Kl;D; z$tRWZQDNV}o*q>e8nhG|gJZ~Wy;KduGdk;Mt?$l>ZkH6Q$hr8`0ildvj}Eo^_s72; zjb4E0tI(MYEHjduX(>i{Y4lo+q1%DG#W9l+%w@J&m2jA43UA(V9lXBC-+$Z8=2z^! z`)L=I-cwG5u*^YsdV50y_A8G2_#_;9;XYA?3_3=xuHaX{A~T8Vg|^xBnl=rRU*C{% zie8O>xSBDL4Fpw@Ofkwji>d`|vzgz|N74b^f4u|{h?HHSb{|a2dOOFV>ve~aCOit# zU6~dzY?tyzVa5HrBFI647%1XDc8|-7wfs80k;d)&Dmck_FmSE5@N#H8)0S&;56|V( z2jUNj6(Flrayke!C9uDrjh!eNX_}{S@ozIklFoO0#WFP<&XSMDGU;w?241YMpC|(H z|8^tiMVcYn<{xh^3VD)D`2I|pf)6w4mfH>@XMQQ4=khTB>6?R{YX6XAMW}vZ8|N2M zbMLD@W!pvbuDqa%vQT1C6Op{r4_uDYi&Pqj-@2F3?`MzsRSMRU9P@`Fj2GO4vCQGA zeipM%CI2lhgf;P<3a{;nG;kp*8c%<9;y%GuuQ{Eq* z5?F=L`}nJrfhi!dV3MZy>H9LBtU#95a##Pc_dGpQZixkB;l&v&(7BLsMus$?`R9)?V>9W zjyBe#`etE#BOI?SN$y0{q|*JXrM&W?`~?9Mf6A+`^34kVL=d5QLiJHJM|x26pG$eC}P5rB4`x`c{0`|hEF7-%9Wg&0SKR1ZS zO}!+&w9!@;>Ti&}6_a;o@>Z_E5v0Qpl6F!gn~DBJpSVrQgQj^Mg^Lo4*M38hFGp<< z=DMx%sQCoSTOyJW(`hx@RbK4uR##VKbMRdMM!5lI;S(d<4Bk2x=!! z2;OAdRfbIzLEVe_K4KKtQMZ{t7=~_HXvQ zqC)OdU8e4YozuM4)KKh0_hIr!N{9B}*mq+VF+njEW{4=<1?#9TVo?$hL0RWQnW8wI zDRxrDMo_fH0uk$+$2tTlgCwAdgvi~74~9}>SU3~hx`;pyck?`_o`)d(a~D?@mvjv%B>mnB z$~I1ZNHRv~9?k4xQ~XIoB2(`5glH5k#3eQ0sabSrhR@Af5GL>wcy3)EqIBOjpQ(#p z31bfp;jF}VkUCS`83RMQ+_#qjo-FbHXIjz-{WKOmxLF!Bx>1T3G|dsVbU;J9k&qW5 z#m4Ewpw1+YXLBt=c0@IBWVq+nhEKSW*(kq~28OxqCfE7-&NUT3gy#I0k2fl*$H&%B zh*lAsbdosG(9Y9S;NPv#l|A?`^hm@*6c%Ql@t>I@g>ua3hwgo`=jsW$l*#)ZZHU&E z_}OS=y@9;OtwR@$frs`Gw`3K>Q8P{xPkpaswjfmGgZUK*qz{4Mo0hbuvkX(rEH$%8V zWC>3$l>(3vV>m2R1)MAD0^X$8$z|A&lmM-xmZq8E9CryXV5GxvrM4Vt1>q$MS@yTu z>?o4T5MsC8R_4m+cF@iy;_5_bE_Lr&sdUqg_>4#W{=my39zQKr3V26+1oQ(kHk__m z35AX|sfmUfP@$;PaC#rSEEeQ#RdeuR?g1@?;zM|dH5wZ|V_==HZ8DR(rJ!ru;Bbjd zHt8CQWFD#(`d$p1k8#nXBf$<$0}XRQsv`E!{g7lpREQVRvg1#y-(u_jXGqb;G)yW- z19E87fgXvzH|o$)JMuq%xOxn+2s@lglvoVm0(DTd$HKsUuXFMQRN(9LQ6*lm zdqkRnTv*kr9K?er_|;3$yU_bW)vI9AilI{uLov6u9u}(i%%Vt1p-MCFqTnIBrek3S zE}4TGPlX4eAN(vfp5_FAwhPiOSUuM)e~5;O_P<*J;Gl4J_K(Fs@hm&NAiR_e#QEw} z8HS(~fNJ`SHq(}@9R42yWd(M+31eNhbj>1qfrNg4Nw(EYVDCQ~RA6Q*ng=m|ycE&X{uwybLF~TD zTcW8gc(Zzc_g^t=gVH$uI6xYFMMavI=W?=yvWHj-S3f_d8qU?vIt=4T)&j2!_xjon zq}u|WX~P#?!uFI+yXT}_wwPx`q4K&wCYp=1m&E=wn6|~C>y6tsq{p$gG4aAg&;!3X zAO!X}+eW$z>I~+a=WKzXG0dmx;$E^hmYe@SM#VY^%kb>HXPQlLjW71L$FvEQYA;)( z1wNrOvu4>YE*qJkSDvZ;cO9Z3pIlZah+X_kpMhk>AkcK46JIa#UB&E4UXq6O7bm4N zughEIuJ1Bfj*}h>TP(3{l~F@vg? z_t4VdJAYT{<(IO-06zS-Fe~DBWGH7!j5W^4vlzey2OFY=pg&rr9bCs=81SSVuPrr5 zev>2GD+SZB+)jHbC$|E_u`X$Im~OI5M_`VyL}8t8pWtw59|WZsqfJ0NRPhrTSYD>5 zf$o?XYoC*rl8c$AZq)JDm)_+5SSmMPlIR(9SILm7OaU#l_tQDWXE#r>y(H>nJAO)d z5`3UH?d`R;+RY9%r0Lw&Z~NFFtV9iav)nbvowIO`72O79PKip}IiN%g)E_%J8uh%} z!M`YF(P>Jj@;Jf%_*(O|UB>v|t4d-y8cKJXP!@J^?AxEgj_X@eR;8zwKvjgQVOb-{ z%=t*(bUfhTeY6+P7P}n7>+{y`KqqQA6C!C9at|8A)Ch6(D|h`TjD)iU%0I}6V6u?T zq~ML$9j6o=f&kVF&XK${{O%Krtm@C`t}-bRP`%7C#C=};Hm8e*ua)YV!MX8I1R;m; z!63-86Y~z-L4Vjm7=Os4_m}^;BSr}ETi?pc7V>^f^Sk`OA7sc+>NiIR<6`|^+l@GQ zNVb1YiAg8|m}#r&`rBK$*49*0>UaZO#EO(%; zL|f^ApIcZoAM-y`N~D{qZ+92>vj+f`v?n5?H-msnnqx%94VymLp{UwPluM=}db^L3 zVeZGx)A{jE>*31P)58?)UUYWXwjZyr1XG@aU#E|@oO4o$x_Uz-7jVa@Z(S1pgeLC^ z?{g>XLRiSLcjfk`)$-c}UM@rTkV4}+HdG1h+NYff`5pgt$4^P`^J-i|Z!MtZZTTHP z*1T7VhSR$BPQqC(Cu0Gf0W!HnmKh8+6^x5w3JetsED+b`kRje!XCSv-f-SjFGpNaF zF#l( zYfMW5$q9M%@4-DxDlB*=v@C#gEJ<_*TI_7PdSxOL(!Q=(YgI!LWS!ZHd%Q=vTVKcY ziKJB7tjH6#euGfmKDV?GWMeG9r0>Cn>rLlc)am_LwW1zaU}GI~w6zUo27<=k9xEhJ zylpbx1sKG)OX3^e_^XSXDR3wIR~s)DT=*|;)8X=rET(^o^mf`9^@LA~t{P>5cuiqD zbXQjW<2GZH@Hp{4!}l@LLzu3)AJL6W;2CIq6!^%vP+9!7^=&R^f;8G6h7ho>X?b== zii1VF_*=hVU{-s^ziR|%Kv<0gViR=7an4uKs6dB@2$N&i!NPJ`!wa-5nUd_(gI0M z?HVwlW0yq|sfgG^e?N32h7`MHcebixnQ_^bK;t6(p?MI?F?mSiI_`p-C;FAA_yJ($ zt1q|T#%^-*P2wP7C(`NZh0s2-oe*0{w_ovKQb`jrxEY6ErEhuzLlsk{Pc$UInI0#0 zc7TG;U1$?lI>DL4p{RECEEx8$X(xGl5KwGGfg9|`M}K42|M{xF#e-j%z5+EHEAqYG%qSx9ik@&y zMa2`p=5KwiVFH4YG0+2d-#;pDmxTSRpd-})g?B&-b2y2i*43{=mSqjrOgf8wUEuRU zU93hH6SA}It)EBn)upt+FeD{_oCje)R!0A#*3*rD>?Vzesl4Ip_yDE03W+$Zfj@9X zJ#55e3v8}p#Em~;t&KP=t!=r#V1qL()^`&I3FqIyB5Z9dqZPHEW7XZI-UO+doJDlQ z+HXWQ>lGfq))Gz*$v?v5;Y76sfLMgZgX0@}AkSI!_x!JQp04lvoAn@a55pu7$B<)75@teQZs`lPcG3EcUoxkLN+}-+@kPCVWa?_$=_F13sK*O`gb{=4yeR2oHM?I&UbJ0Ges7p_9h5 zZaD89mOPE#EC64XU#jqsj?gwqby;cSrJ#F(6y}}7$eG={&KHX1E4RlBX#OfUi6Bv; zu@}hUzCWO0##*{ny-J=;hNTe$V&4`m0xJ<<<6b!ce`1js4OC+v8z$_);+!QqUH9nX zq{#vn(;r}LVxij}s84irDF!&pJGf;}G}j{B+DD+1;{B$Yw#9Um$}W&3TjBxj@el;i zxq@n47fbnkHc5~$0Sy-RU6%XWQ$Av&Lu(Q1p+kFDvC-vxfrE1q zRp%gjbj!U}!1-nMf7S_{0J?IIsx}s`5O+aZ%S` zLU7Z2^mxA9F=!F)tuG2GYR)Z>#x8nciBqM%a`@QaEa#K&l$qOaU&H)rzb zvAa1GtP@4Br=Dl-8q>1#gy|RY&zt}Eb--Mj;9OJ+3&?8SmQKErrWJOS{=+_=F`VZ}R^Rf1AJ3#nL%YbRVGWUZ|72R0C|%&M0Jx@g4r=mh=pth=L^2@e{) z1^kO^zasGA=Xn&GkD!wwIo*i2lQ>N^c0jLPs{bvFQxIfg4~qQ}tZ)mp6HcJxvMYar zj&0=@8h5N=Kd#0d(bK}+3A?f_fuWjxS3OGo$Q{d~y_-CZVa&sjhHt&H+EzuySCJk` zZc?1bSN`e-w$k+2llx%#Ux{Qbp18L;Q&d|{jvO^YW>7dtDK~nNM~IXj0p8xPzh`fQ z%#KPCW`~E54j#U;ZB*32oE{)qn$k(=rdlza;vn3FC&N8};o|N1{AnRbZjZ|j!D2gF zg5h|dFZJPQYUu*G*KW{PsEY?%sF;@b=)7isyZ7tiW_7j-I*k?wV2M8o)D%d(EE766d zOI9P-RTb&|ot1>nFkiCrZ7ma5C;g0fzM4oxS`3tv0Ll(_|Fcm!rz^M)W*}HY(h#Hf zxz5hsOT7L)KRU-N*iT9LN0IRimgx+=Op)mfcKS@zsd6YgcQ5-Zsd`CXc;69rf~g;~kmO~~3^Ox;%e&SQ$DFmr2Z zTN%$~*JDsnIbcP%$!CC>1P30&*+C~rJyE@&QK6CksasTKqRwtjTwe7a7g7>08vc0U{{A`wEfoH^!E7{KS?ibe4L+E@yz-xlOwxL(1mz0{%8`JI z3C78m7N8472h?-i7{m0NsbzUng$qAzk0)>r5L%7A;)%%l;q0#Gs^{NgrPEB=(J_R7 zQUL5e;Sy6VJ%G^0VGCoobg-O0;`T#D!2K{%m$U1es(H5{w%IKOu++pf7Uor0W96ji;eq}`WCiIb+*3A! zJ+~g>9ez9!eQhmF+nVi!E)R+skXiOP<6Z!oB0F=L0EscT_<0)&3tn5gGq8%<Deh-E;gnYS1wHLuG7a1jI(>SaNl zTr8DmH#F3O?9xFM(XN?AzCXFQN*#0@HVT@_HG=*P?on5qMxn@A20=8UUy{iKXXvM< zSLn9W7DS|x;Eoe*ROuHSw;m~a4IPow5H;k+^C5l4w?JL3VXBAxmJ7)jhiBrkJcq(B zpphYGjA`OqK) zfVbLL^GI<9bRIRvb%CShcIo?t{%_VXdv(VIO7`-(%He%3wUb;+HeywK?Bb{~+7zig zr<})V_paT72}JI>Yz^W(MHwqD!|>GKz=Hw6?ohN$E`L3op1o^~We8-G$)D5qaw_-h z$oarle?({+)cH`~p^i+77@=ff@RrepYZb0XQjy^58|!#->hDI4YsiP_!RRjgQWwVW zbNh^wcaNVno%IgItoE&xypO7`G$J%sDI zp@pHWrHT*((R(RvZJ@#(RU7Eqa%RP_*Sz{_LfOEBDV7^DV4qjMuu+on1wNOMm*`i= z=>x71yEf$Nc49ZzIw{YQ^ZA|@QA|iCBbT!M@G1%t#2&hbBRIBR0YsIF<{684rfdVDq$UUx0(cT z#GQ9S)wm;5jGJD8X?{UcAo`CE_wioBbl;3wT|lK#fQUi)OVE92DR#%3`qHY#9JL1b z%#TpAq55Zf6SQvr86G9Ole$;udqW;ajgCsUT!62QX|88t>ow->9ct4ympJ(t@o^r5 z?o%eIZFPlcjJ)swve4JK1tAeHV9K1x zE$dL{J|O+x?*~z=qk9PaPF$-$goemTY}lRx=rc3OlAD`m%sAFdW;bO$sXo7`uqrV+ zV*b`bl2?1|7ZHY1*X;k*+V`YkW|>#u zHwVe-wtsn_q2jf4o@!Cqx(*e_4t0l8hd2#Fw4=6hJ%feGO-4}WAEvX z;+fvA)RU$_Q@u)d`5(_zhKt%VGUV}Vy@f#}M$TH(;<(fGj{%Qi{rtGLxM$QFkamsZ zn{nk6Vi(v^B(=j2X$@YQS(B+&J5QAp&*BEu64f!jGK-+9!xajhg|Vuts55xaQBDN2 z_qKjq2=?YIPWTuc_UU6q*#37nkhT0sli-`o?XN*Vg!j}?3UJ}0RCb5c2ZBKi0?NWH zfG;g8#9Xe*;sJQJ*ODYN3mH_HG?z}Dp}%~7xO~m&qbHAZv~l?{4w_})?1928PWog} zJO)H)N5etcm&=^lCuQI&awJ@uFRtZl&>-t_@~K`qvs5 z!GS5@Zy(KtEzjCq8_h)J^av!hLl;lO5G;`C;(5SOe58)b2R%<%lKuiSQOiO{uJ-U$Ct*(gzJaVOpxxOl_J<^Qp; z(0F5^FaI#*XO&l`Zqb?FVrmE4@KisX2f!;V*41#4r2?3(Xyo5p{z0(A?NwW#$&4Av z+2x*bCiDDZtnqB$4T5)BvQ{Kv{*gRfU`RJaU^+rjA{pj&Vtr&_NI|?=%Al>9q;lKooYLNLAwcIdhTzWm!tO1}VJR(^i^6=MWz2w3iL*#`IzNDDO7fGT`e5b_ zRJBJDa=+U?M9vRwO7vc*Pj0weSer1*rX@w4#xh~~oG%>pXt4;;FS{$o{;0;OA%Eqg zQ=yb$>frBGc%CsBH>FJS37}Wv%E5a&-oI^N8u=Xx^}fKil9|yQS4~r^-YXQqfFDN&Bp{QdW=}81`Lc)FQPqr07R$OY!RIyz9+m zygASbdgIY@wpX@4RqNb%I~%3DJzCUsrKHr-^_t6cT$erxCvP+C{|*35un(9S8`2Y7 z(oGD?a4Pz8(sg%lGqt^`X9~;WT-0apa2>IZB zCWRz3CMzX;`uGADUuLama)^EN0dbYatDU<^+0MRa>@CXda0xsjxNv!Sh*8au*lEGV z-zrzVlrpzG23QF*KYznrdl*hKhATM#X)j6eu-;}i0x*~I6#Uil1)|regAWs``51CB zeaIv(wDHJ1V2d6vLjD8)T_7lAa~T^cPPJr{+6|pM_Mrm%$7{lWuZy6^Cy@=n|NualVhco}3yZci7Yj@{`R-e7NhSlx0cB8UtVi>%rISOYiK6;b$;<0kW(ga#Hu3O(|P9BeH96pLSB z!eK*&3rPRgG z3a}AQSNN#B|FRGyY;>MXfOkyinvPx>TvSDA+;{*9zQECEF(Fv0?&sd0EE!eT=VA>1 z#6+~F%{v=)L4UFSL9}JM$6jzgV>;y$&Q4|DYSKS%aJ0V8U2219GoDh5DRR-BIwrjU z!NAqF`BluQ!lZjvo-E<1o|L!N^3&lE?}?n1m21y&?=&hdG5Kz85QP;hfdKg^ck5Uoz0Kk~#zDqB`V>*LA@UHupsj_PvEW$v`8VF|N?2 zoX*XJre`~ZkOj?EQsHBL`!{naFg-IZ+nikF{Aq2|Ij_;x-tvs>bBaRv;HgiB?s?FG zIqeoX?NrxfcRlPf(ev3sNiJtbv>lr%`Nv?U+^e62K*z5~AXnqj#sAbtf&F8W-l2nI zg8tw7s3~1Jmkmj@e-_5&Pj!$Q5o2Af(n~~UI_{U!5ou=ovRvN&w7QApiFAn#rWbWU z-*?VmxQl~X>Q@0n#-(MfR<6%0TkJMBK1x^JzwTY%4zo_yZ>&?&Di^&MJiJwIFnh-2 zE}ly*y}u`-*S??sVk%p&P%W$;hiNnG|e$!Z^1kHb>}l+&TJ+M^=C9EF z1{PElWDIup%5#;aB?^bj7rqDMBw*aID-Slsl{isvu{VA{{)rbuSaq9 zAJ>G-beX<7a<(9oolGCTeW-~NyTBiM<7yJpjFX?;Dq+ZW@loVd<*9J{81=`|rM7=HSYm=~*U&z%#>zzDwqRPCshS&cpzV(6zWL%=XPWbbb^|J7*HY}Ra+<-4w z#E8NY>%&(ty+5K>bU=F@cb|!Rb3Wp)gZ=&B8Jd+T^=x#B3{r#e=j?eaU&!0(<9HOaZ!!O1Sh?(zVYJKmz~e+#`L&{zHk6gF^30kc zI3B_O4~HF@i?Ij(Dx2M^i?)%zLM0#_!rQ&X+A%rhvv+5F&6UL&3Y15ZYBA26fU&}q{V$gaO5bR z4LQWOd{pMSf8(!8SY8MZk?{By?wIwpw@=JkWPB(^82=?c1U zSna)^NLwh<*d3V>pz#j6=Ls6^#x(#j*v5d7sxKE!AW)kzeIs50w-3195YpT)bstqc zPp>}gqN+F4dI1k>N$?^j`2pt$7hegR5sX61vC2aDMrxCd*4k(mi6)3}R(IddT8~OW zhA|do7T{MH_OLj^1eqo~W5hVz^m_unq_foimi#>d)#=09=N?AmFKKH_=mvbkwR>LL z)9-SEuUPRVq;UrA=3Ag&@IA?SGtIp>>V!Pc(Y%t7M%!Ez{M607P@3mA%DSw;RLc{W z2W>TI)%^BmRh@Bf%U^>KUy1$4_OMfjViM!y>u+Ul>6&Yuf~s{p&G5JmaqVV^s#&3t zl^-c0y$V*4L<_hnIkzR6-sX5Xeu={eUb8cDY$ec+g`< z6{Fz4yiDhkPB-uqCyb{XT+?YIEq@F0EA;k0trCrM>P*NB8#=Hh$oSUCiq*xPJ}MvE zJjGb3dC@kp3NRgR#O3TG1rYK}7{Pz)p=41BR3z>j+@Jic!~`{%bB4yB>z(vO!*ZX= z$VMPM8{J$3ts? zMa8+mmOL0mDTd23yXPm`6sX#5dGD!h$4H~B&=cQqI?X1d1%`gF^^w+;BNA*_1@=B8 zZu+198#=4g3QikqE`m=ao_8bzRoIxzgp0$g^B8t?;0mzHh3yW~qoh2!+~(%?%!S@$ zxIj~MC;cKjg?Mz1XG(CSeXyqU*aq$q#RUfi!Hwta`W6gNg1kWbNW?uPHvoe}o{&Vr*Um~rx zC2+^mvl56o=}UemGTTn?OU>iUtY#sYY=DSUmLxO0qSfUd( zmu~W)YScN45T__4A5GS^$n&U)quI`jRiW~13y6BuZj`M*xFK{d%(U>sW8p8@Gz1 z6lD%=Tev!1@2%Ez3-F_>)a@xgeNB0>#25lLwR$cB`s%BvQAO;u5N z)2Wzo#Dw?FBCZqasMCUJzhg@@LlSrrdgPD)Y4(UyETXgUD;o#{{fD3G5iwTPA^& z#uI_9T`&hp7NWg`oz76nYI2>7g&^yq{n}x>Ynl@CYS=f&?H=oCodWgC-{`f??CPwcWdJ`_36xo6_4aj zb0%m~Sb4wYho=Go*tMKy--<~?W9)g3pt8ladd&#UjW8}f-D{{*!UL$+`DEIlRmSaa2k_fCh>}%toMrFV&>A_U4 zcL}K2bA?Qs46oC9alE~28_kM``Q;~275n1PWVS5FFocbKgHBD_$uQ}vw#s?Upv5;r z#}&H^24axnpC01&0dn~jwnh)jp%UJB8WFJA|B~IERk|GzVH?hqMaCG8KocsPzUZi$ z%`iFxm??#(#tH)@ZBLSL97y1^h6bbkL)n}Z9w@!l)F8O79|K(G3iNi#Lh7pIlFD zZI*Dgm78*aF0GcM@&$N+$u;R~fbH+$U_lXSYO^7)*>hfXByME~4J6Gp*8?^7Xq(`aLZVX|L~OsZ2h_$?AU8CB zr3`fRh0A(EyS-cvg943I&Kzq>YW_w0arkFpo7URb62Cz@e2=8dNXue@Gr^Bpuupx)0&k!Zv#8 znAi)Cs5L~c+@o;7 z@NvtkQQ8QiE-fDm5^QzmZBacnpo5K=dWG=k<2Z6v32$8BF5Sa!`JO$>T1trc%2et6 zz;Of}sGCm{u*P36$E*wtSKpsB++ePvm#O)2|5%C5M3kqZ_amy~*NLRC>y^uQ-m~Tw z^nK_84NNXgaDkO_0r=M_CbU&rGB^K+r*r)71XzP~Y}>YNOf<1=+jf4jGZWjkZQHhO z8#{OJ-aV&3bpHe0{hq3NN_|j+#N8AA!by4if-Q3d)P zJ8jw~PgjxAjA#azMCGDkLJj31ESjs`#wRFje65t52g7?W9)iB7&tRj>8-qTC}q>nytpKAh4ReX5XaXUV^(ND1?%A?$&wxu zVQ7BZ%kF=cR`RHCV6vrf`3aT(E(*#z+5E60Z-(`uP-T&>rQBIpM?qkuZj)G|@DWYm zi=9}aIb46SQi1nriHC_b=gGP&OX@`cBp3b~lc5x+Cn^{S!d!+Y?DYTM?u$yq9SA3} zC1dZ0gF$%_KQ0tHF;)Kk4H%QezNh0t+2CJvp`f#7@8s0|@*1?BgxEbg0qrIbI&7CS zckqo7W{dBiJ^zdT_fN*0WRnb+xN7OlrZ3gNfmK~RCBLApSymBfo4`lOV2}%qtyfq4 zTc$QrpTd;+AxwfC8%d04enFPQGQY=EAM-L|9oc7*p;BIJlG8!)*>0RQ4!5j=XCEcc zFSPzQeB1qGF&;(kcgqDP8Eg&|7=ynSH)=o~68iFcTAI!qqsV?c`g!9bFqJHO+>SP?1Z0{5Xk%&-_jaZS zg)kVFk=8H<9;+8iBHv!`ICyLlQXQ(EM9e^55Y8@uFV~sUg0natucv{rDa0B5B@{Pr(654li)A z#Pori0|SeB3y>o6CBq11s&DVzq{zq*qll*>nx=oL_*n|yU5`3{wRj5_#2uzNY_yS{ zfgMpVNA}noHhrBf%pky4rXZlq*IR>Q?10Uk$qFa7j_NTe-qT-lbbVgLTs8VA@5wQl zOy>-5q`ptFuZHdz6bUeqe~zjz{`&N0vMcKXOOD626@BMQumv{{W+p5ih25>^NRHmVDLpM| zVS8;{Cewsxa+x)Gj3+X_fAE#a(RFDhTF&_Dp6Efy6Fg_qnF?T`rBVu_Ts6-%_BE%l zu^PlSb`Du(d|Jagmw`O`XS=@qbn{r`Dly9Q%jhzIN2{Mj>JCSJfila^0xM=lSa7&7 zzS35(AMqm)KdK1t+WD0~jF5r%c@M&GUd$`eSVuB@iGzZIpQ&N9w;v8jvJ*%QX4V%mr5~XIS z7(K6-OrD&yh#%$89aay&rA-GZ@V9*D*EjBjJN)ROO(-0~e>zrfy(z7FFX+TfN z>9O-$R^)cxcNaSUzz6Z;YQm#ZdR-^h!yWefxZOHL7Ak{2N1C9lgX!rkg$ z>fUXca2}wOKv~z!pRR4|kTn7qiJ{;`7M}Bd3a0*+0aujEuKDl}S(hh0SD`;9eb-=Y zS1#WLtm^Y9r0uR#f60r(!j0=GJkEW5zh)0SHIa)sO3DbCR`5o1?FiA6|66rp z=5<~yXVh4jm#bW?F1Z`S*)r%@t+*j7;PlE*I~;%oZ%Z){HlrKcW^ndp)gQ9Jd;Xkg z=cz{H6vfq)aNC7F<}c=z@|)A01tzw~1K{XI)4VyS1RCAT$>FkPZO+5d-26hzmRoI_ zhx1*8NZt7-p>a4#z0X>Wc5X(~LH%M4vIxJ1iOc`AUigIrqe;xXl22$d_%cvd?PE27 z@B%P-k{6y$s;H5N<) zDTqnai%q50@AI>#v+x$rQs)cHx7SCHl>|VX$X3#HSJ8N}(TasP@fT0!i{g7?w|#MRGW zO#AdG7EAw{J?4hkZ)x!HyTltA!OM8TjNL64d)L4?R<=_iRIe>~AH4HVICzz{FdM87 z71SRZmEtFL3L(D{HcNBk9?aZW_8|bx+C4)wlrKEUI&KTRx^CkliynveLw`de_taT;2oCsY1DXGDwSDmadt&hQP{LOe+>daRllko z-$;~6{gd|{yA9&%xE#dVdu2-EWUx|*>{>Q1ltsIw0{t(-?e>Ey@|3K@}fabIip*qhiApNV>RU{H@D%zGn`mqG?UU+u>|-SV2{s)zmM*`ovMX- z*6D2*JR{Hm!$bk04XI;+G6%ZRs~4S9;?vW!FEkU9WRV_GPUxRh-U9$7I)cLRDj<<- z;)Q_A&|cgjY|-3X-)^X!Kf+n6`23Ie=s6>j$#}JMYCNN;Sc1#IrZJ!_nkbq{d5Giu z!i5uLlA^~0gM{HO$!5CJYCQ6}Zg$!G>Hecu#y=L3XmBa|CdU!*al<<+bHwUzX!i>RTZ>XkbLe%$vK~7* zQ0&x5JX?l)QXy>xtv%}sj8RCAM*Wg$)%)OFWCMeG%_wIDu2TSux{f@D30?NuU$TPM z+l*|{^Y@9#RN7~c2eKAaU@$|3_xccAI?G(=wvf&1PDc9LJ%R_l&&my*wlN)O+t{aV z7fKBm`^Kr=7#8~B9YQk=dymjPT%J=ay?}?!>nLJo{&JkZZ%(nrh0{l^7ASCs;%aaP zSm6FqroV?+&q>t!IZ2`BXsF2CqAP~E8nl4QyuE2=86>DGx zr7Q6^waU=yoA=6(@Y zu+O`^V0YGW9dI!YN?oyI`Nz`$yY2|?p^n((Si8NG?YoQLRd}Emvs{028KD8{xPLC) zW1)fRe&NDLOC;E$AgDZ8@t{7CfJu^bD%{l6VY78Ywg12$sXSVal}aQ%&r_j!IO{vS z3t>1xu0aA~B309U2~oo6*)niJ{N;5%p|KGxeIz7mIyA9jv7Z#vwr5v=DZ~=24$aFx z-_02Y3F2`SjtdmNOMa*@Sayv0sticnR%ij&+Va1kkeMf5Z$y>SL?u8Ry5o>B!QZq{ zf&Gi=0*l&%v;H3A(w3(PN(Rc-u0*uAadBPmeDwyDvO;I{4R4%Yxro?kH#<-K=Fp&J zdqA{lc$bNI-l6vUburQtc*LkFw|~}&PGNA{+Xet`K&T5V51w=o`%y(wCwvHEU4z#| zJ6itd6MN21m*f^yn&t-EFW%j|1B~wE5-c2#@Yt9UGZlqWb7O?Z#ho#v^F!L~rO)B| zUQz?ruJTbeb7Q0aq!^J6MOu%!v0}h>XR3iKg0qQz5~;_f7{XluG5bt38sldUe3_s- zdF0aDrjpctMk6VQ=mjWyBgfywyY`yjr~}X6UdQ{#tK+5ebIU;2%<4L(VgMNoxe`Q) zGsM`cZ9IT`skp(@nuW&*7H`5riYs{Z6vhAuNh)3kQEvlts9r+FaH4r;o&sv2&;%t8 zO@`%XCa%CpE(mfF<^74HkmcW3!%Mn8@b4Y!=p2Y?SgK#FJa263p1SeTLr$ra>n)R1 zQ!}>GVBz~3csenm^~QL&V9mp+ozlT{TyeVmhSf!@J8D1F9mc%}Ypm?E&();(6r%xb zbV2Q(F;`;EptsbD^;Y8ez0Z;QNiKs7H`HiCx^VrQ5SR6{b^a_SO&+Ct8D(-0HxrqH7MV z)dO;5iQ9^Du zg^|@$NYl&SB0hon6r*9KQ)2T(8B%OMGgZj}N<+5H(< zDMcLmwe9S~FhhN^u0NzuhX-+Y3)wwW#}SQgkVS4%I{)w4RJEswSl)5fxe9j>+NW$3 zx09EARL>v{d{5f7*b`mSd&S1L8b0&$Q__Ph3Y6&55f!g3x%i>Bhn#F7w)ci zQA{V>bAFx^+pD7p>qgAK*pC3Qgqi-jP-f0xksR3|(42>Xo2uqD#kDNry40_+OJK+p zc4B{G^E+{lY=3;74F+n%Y<4(E8UN z{iDjr^({ldpL!hGui_HZ57TWWi`-T zYNRzRnDS6J(?hiv!Fjftopr8O+0(TKB9wwK=yvPJwxxfEXcS}x28DHjoG9A_c!qzg zith;LMki(-M+zzz35xT!`N@u{1;T`L;LWve@sJHzGuDS#8T7ZB{@~4sL^tlI)YOQb zKUL1*+(Mx!?`K>zn9COQ#~yYPgKNErtwH4^#EAUuA8~x@2i^o|zB2rBdA+2_w5%u$ ziE5#nOAWV97&19|gDzoe!7}^2#}^kP-pM>-CK3zNnVqaFs6Os!R*^+;1Xw)gar2XF z{0b`wmkaKN0~dyiGR+Ji2ZYEuSE_0EyP}DUgT1aj6;7kc96J>&>o616WUvuh5bpGT zWQ}lE&N*65_8kCl$M&i$F*lB>X0bc8(j#L%i$qE@Wh)=PB->s8=@6#-!N0F17kCIW zjt#b^7fjjh9sX`FyKOz-Fx+8_2J{N|r=2fHz!zi^7MIOrAM&FTp8OS5uF{&@d|1Zl z71c3{W!>kw@>{X!CD{zH^>4@^{D!1Nc0uzK&zHL>bnCml!~ z24D~Je)lS)@+AgF+hii9;)AuCRS^flboBib0 z6iUuz{RTuti_J*Jj%X~$5oH%W{d;O@6wuI^m4nih9%BtVhT(<`RiqU9eyM6;Eey#} zH+0X2Ir+zp!PXeA`48PeG{Xf$7$yUYckAJviJ{>WA#$towXx=2QV-<*k%~ai=&riA?9Y&ekxiVTytxBoeAGCHP$b{9G=N4D?G?z z5d>%pUKIn!LzDZ1ksNRpj4{Jc##JqM8UAO70?z1Pg%CWW_R4+dHy#|$oaq|dDTN#Y zBIWs!{z{U9{JANw;=+^X#@4+Et6~qP`scLVqWn;~5q6$m8OBYyn;r)0(?flw`NNSZ zW{Xy)oZqLTQ*kyvVZ$SwJgcB@6YJ!2lq?{ObEg6@ogzN5^m%ZR@HQm2q%u*j6~UUS ztvi)+fo@au-BrC|rKsB&uC3J2uB%e9q@&9*)FFj_)Lx~+Z9o3BG?f3IyZ&aRR#uU3 zb6&(_PxH^+oKII_T4Cl4SC?LssIz=k*{%RTtY*rd!H6+5HL!AVWI9(_gE7>u`UEiT zs@yKG-~Cow5!3A2*L;AJA54?a!Vi8w&D_(Qx;LNHSzJ|ar7tGG- zDI1p!o2kByDCtJr1Sj(VkiO@aO#m#HjM|o>5-4x`&P8bL7S~^AqW0_4ON~pzpR|b! zTu5aXEt|?$;CZE2OW75B46&nf&Jp*cOLiMlF!GW{4aD>e8$q_0NYKw>UguCIrn)iS zshyG6baVk*RDR+G*Bx1-Z?4@wus78inoBOE7wYH#dIIrt(yl{YL-?nop#kx%%~NC% z&Tz}iL`^*Sg^~zP;#qUW!u(bG%ODpor$&Blb~Kt8D_5O*Qm6>44j+n?zt>dh4vIvi z*v6z|GEL~m68F2#8;9PE#51^l z1uY%GIcuZWtHnxo<)TFRXbf2TJl$;u?$&Uf zvRnFUXzk__a^+!MU*decdm6vm^iA5B`%yk3NFi>GWTF0*OXLiCkwM4*iQ$1d^Kjxk z*ckHrME0)PYISJyJ3d*=ZC}{C0EM`MH6OnKr39Ck&U=jsYG*U)VO@In`)faLOQ$)Z z>l&Cc7paul7zN8Wy%%6%1LBN(l!j$jUDk6vmE_gB@7*fAQE^ps4O@HUTx}xIj>0w> z%}=TmmL^Yf#=kHKq+VZR*V*7K=j!6I8k74+YP#C0{M+$8`iHQ~ZsUS5kw<|4T{l~x z+s=xkt1?46M8gMG=H?!fji^>_ZgV7X${3)>)3U>4^jTDaT54#%ins=6Wx(p7h>sSWVGh zv+tHkOnvszX&cb3morh{t<>PUG1mL;la?>no8HD_2 zw<)nlMAp_tw^keR7FNB<0K7Y>%iCEsi<~be;65mceU`=1O5AFm7}gx9zgikvmSup(xf3U|O^4Jse{uK@Qppp3q{;c`Hl0`X$Tl2rEs5ScSRqFss| zj#|6>P-g;i>ZUhi+#6&>e&ELk%@=PGcCcGk8SpT>J^mmQ+g%HM zGYC{6lpuTx>#oW4?yUMNq-x&I$y*MxqmR<8y(`C_D-Z;O$uEhB7O0Vcyj?+Hk$-&z zB45YI_G|$AYW|LgjT2nyyYAcLZb}lwI~9=A7(iKYz*KgAvP~kKy3aHjj{!@(xXs5P^S35fL=L( z;jjRh-K?fn#t7i2khW<`o^r4EWeq5-+)7a%X+VyJ4~&=Eac-~fj}_C z2Ea5k7>t70AN`aSM_+@&=WBEr@k57?R(~D@5Z42*MqeKYzl`x9Aq&>(!>`mC##X*) zp};4K^)SOFkVD2ELX79HCp-t4=(Z4Z$t)L>Q6naTdo~!R(hc7*Jrk`lCiLg=?(qWX zt`WpKhd%9C2l4M;70y#mQO{E@CC)Lu#xYwKZM=7sdsqM8>RDCp{=ch>?aw-K4N*|g zS_p%KZoR|(rXp|FvrH;is4SbYQ!8-o-%n#U16lvTKTO!)^qMRh{*5z>;pT+jIG{%o zZh{oXFd$+LS5z*iSfi%^U-JY>UA-xSk`V;M`_(_?`xt9UEf_vPH~s)OUU>%)6*q1I zlABp?fV;>|fXiAzymbz*PBce7yz^{V;;p7``GV<_U(s-oaiFu4@P3Sr04JHrVeGSH zQp_H3{j(PsU4OY~t`P~tgN$%r4}Joo&H*R_!E57Lv!Zw^N_83_LY>FX&kPLlHv6xA zz9$?S=w@%hc5Eh%!&rwW9zTL=?&L2PX%re2 zSy);3bn?(VEbJIZA*YH!hSSA(^RYcY3t0f_dwae%5oJCD;CKs|D_<<)Zf)fWKh7P6 zRc-&=HFCI}nS9^wVb~U?KiF7`t{YDS>Qbl@nlY)!W?jQ415EF!SG8(tHfy()`p_vA zmV@GNayp_)(z{MV2$3yd)lEhgyhlR^O!#oih3F|5{j)Svj37Dv(~wS zxNXvdcmLqb4oE+EZP{6-={y;+=UGuO8W$Fs|E2WwKHSdoF~#+>pZvUF^6>BX5>^mu zJ|I0wxR*X$W+QOW|AsE$YSuIyd=!%#u?tO){`n<_K+(K%(WE#x;=*F@u%OoZu!4!Ivo& zPT?Ko%p2D5Fl?%&y%qM-k1kiIQHJNW00|pQA(@@0aFoaT!dFVt&S7@!<*DIeu z#AfLO0eIz40fqYg#&Sc@+qO8AlVnb4mTnHt082?D{|^wAWYSl+q&LCa{KDjf%Zm7? zA7?2ucs-mkvrLg(rpcauR1R01ib&tf7q@;75r7UIuVK@iYLK8dRv_X1B$MNeDae}5 zpsvpDrQ`kLoj+@Z+B;r`>IC=PVzAtwmMU*MakiVaaMLb(=aB(}d6)|r#uJRuyazS}x`g+PIWR+12S}U7U^AcW$?oxO zIzTNb+VPsoaCQ|(KPTrnkd}KNr@E=$%p?@l=kPYkE_LPJDjQcbvDxo1Z)2w+k`Leag6#U5ziKSo#glwrgE*X*DVS7Hk!yT zg(lB3;cz%rZ2_2yy3#tF1JFug3`2A)UjRkZpDC)k!+Fg4TPW2dh(X~w=9ZBKu~t9J za9E@ZL#OanL#*+P?z#S^o@xAB)y}MHDR;?6HJGqYOQ=rM(vM%-@Z^FK6x{jRRyR;;OP3l zoH7cA;Y}3Y-lH>NqJ~xE z`MoFBN)BN6z$ckbtb33P^~JhEo447EcY)S9@DTYjOErpq~&PSPQ^>c$ya za_#Xt*!Lh^1OZAq=9nxfar^ee)X>#!{+nE68K}lpjUPEef<-eLuV3 zfe#s(Pj;*}dxm^MN0W(nE`Z}QclwRy>XQq7r^>2e?)sRJ?%-F>Nw_Dw4{X2*PM~&h|1KJ2|w^@uI`qMQ8C>L5>dO*1S)ZyTP$*)W1 z6kiqQ!K~x@7sG;S*!+k71XqIAsJ19?ovnIDsu-;3>p33z7S-&T@zM0_Z1<7VS2SO> zL#Y$D1_4q9Z(S;D5zNJ(D$`7iDlGyBUR!tV;q7>~D`z9;l6=jE=nSF-e;YZre>ts) z$%d4Xeg9uLWxm0;bb!<*g{^?uoSuaHp7@5;K>!6!4HVecicX1}y6fD%TGfdG@<*Da zdvwZXXl7%671;a5r~DCC)En-n!}u<=Vs^-$1%k4lc{VnAtQ4O`g7StU3PdCF93{sm zC@at<*kYt~LQ?b@snR8<&V_|=hw$u=EUx_Wr`L`ou+SI!2H+?0OHmR|Nm9SItZ*S0 za1TXvqdGsc6v-+u>wf^Ya|J{9LUVW|N8z9+Xab^UJ>M>?SBP{!>stKMAhMUl;6Bm^ z620VDxVG(GNxB6Sp0{6LWH#kGg9v$*(~sbvyAL=g3{&%N7zQ~Oe;r|1sXcMa0@s=| zU^A{}MmrJ$1L!>jJ(Vz6?@DoXRT!{&D9B1Uu?fpN3WJz(ZYYk2xDoLmf-_HB08F82*&*iS1UcO9N@P(KM zseFYOk9AOTn>kFw<5h<|sqOdvkg>e+g%FUpSDV0k|~X^@l9xbN!E1k^`~E=z#+;gbEQ}r2@<1&UeP$IEP6V+Y5lJ#WZYJ=x>W6 zn=MMcQ7dS!gr&2Cno&x+(pZWuD$3XVeREfq$hKluC6FWkobG#N6J`;!?p1A>K$ObB-klU>`M<$?_0TxgSXUd|RP~zTS&OKr44fqGmf~$&^)By&>#i7W;{lJG!|zCQodX}*6vdAsG*^=@ zEp|*w#MKEL^v*5!g%c}Zoq6;X;}9&mk!ODWxRCy>Fs$Zv5s$l*Nk5_IJhcXkEDp?} zs!=CH5NMVWw3Lcs^eFa?y+qGv01=do<6>U4D53CYTreX06+OAdrTf9-mBiQ7m@7Sf z7Bs5I`L_dEJCJc8EcfH!Rv~YG*p`iQJ?|W{=JKxOHG^-QbAuTT-Qfm$aSyV%QXyL1C<&BU)Qz(=*LmA_4x zIFrC|%Kia#&nnK5H`rc>cMCd2Z7s^@cysbkQmnv8tU=nBl>P|XnlucWG1ORx1g_ht zBu~G#T#s$N;6hWsEja|0n`@>NZ1mKFEalU?H?L1*mo)up`u1Oh3`N>MsE)&wWv`)N z#or_ODJ7pak4oFU+`YxIfF+$RzE>lgjw4NY2%G6OgZhR;4aB#vA&WRCs~Fb~i}8gJ z#a)S|PYz3Z;m}09B|-_u!AXb=U!CSlk4hIU+{JoZ-*!#{xv@@)aK`x}Cnm%K-`Et8 zL$GQ=M_8E&uW>P^z63g&$sz+0AQ%or=m-@V~idk zfd`D;mySSc(<%WHB7rGMaWZRV8#2bJv=1a1D|4>c5EDKLN~>r`3BFo`d8mxpmjjP{R5d>78meq2#O1M8tKEyF zG}R(;_m*%ZaM5~?P%jYe%$zXL<=_#{%UHzmC(&PfJJmyU8&=v1aegg}A z#4q3c3pm)M{2>Pt;6z!}*NM!?hVl=7HjH<#<31PIqn55^ zQ7tZpmRk@9Y&dUOP8VzTot*{-&3mkg(chCFdp;U$N*eCfRsNmT^_p@~TFo*k$`kd7 z=HED~6w^GKM;zRQKb~)x`hZ$yVWoLlUB==W z;Ym*ObJJBvQc9AGDzF}8ShQF%{E(x@7lo$cDcMBTqp2iiw{Ahw0y9;))ZGS!U)8F- zFR_`0J1fj|*wPGdr*EhsKm>PS?>!O$(an1UD_7HZDo$j?!TjWa;?B<|$ehY&)1~nl z+7AK^{c~t^g?{dX?yZqHp?{y@UflDgH9!h?Ihc@=AyKu~B5^!m@t0Jzf|8{6D%k?} zzp&I`$TX~sCU2MF$3|)=5`lI^P1%eXD}vWKw|%qnIv?;m&Q(S;;M9rNuiur?@9dpK zVF1%Z0I&Z-rvHL+-1_LllU5R|GBUYs{$)u1yYW`4WkPWO&;cqN(SA2NC zCCDt|JKD4qjkTo^AYIB#VM->O0BGKJP2Ji#jTUw34~POa*v4(H9s0O|H%C8ir;%I)?%T0qozPFnp!XNx!n}$;`GpAL4TJSs!FN{V|r7e>JHXBhZNo48DpW zz7aJY=ub5f+;w(~FQ!AyBJ5lzH#B8NX2uMpl~+=V?%0G&Ml+(MqpOOW#8U^h&S zd9#3`N*A$T(?F?07C=LZlfc}oPq$QiFfV$$FfDZkcev|vtG4Y5IfhT5&h8!c* zGs1v$h2Ll+C#r+B0}w%bGh)8pQDBX`VT;N6-9qvg6RUaQXQkOjneAS3onfn1D1nB@|X)@XDe}nKudzexONlgsCN~ zWvPexHz)8~E8x{i2Pv+Pg84;}hE49XN2yIjWFhkrFkOodu}~@&hpTSTf$h=pji0E! z9uLHXWyeo4UINdjanH4ZPA}ZDqNHhb>B3OqGfV_Mz@GR&;woa!DQ1vUOY3CVsid#8 zcG$lrzMa7iOY0pWLrYCfYXw3*+tof&uGkz?e!{2emf7Nw7Fks>9GAR`Ue`&$b)sAyxlzBMxIo>*ISiQShA$w z2V2p;0E6lAntWX1@u%3`XsvmWpG=^f|Fqoy9FaNLnErR+u@Esc{o}+UV?&{XGc&WZF{e0kK~Mr>8o?DZ*Kp`z zz}(@HDjnV3-J!3SfBmCtv0WpvwRT85K%)eL!Cs@#wL+wTL1jyu9Zybj->7bB>ps2c zT_k*+6&I_|DyD#nVrHo zeDx_Hp9RhZuLBF(?EIMk4+^i!$Y?;~?If(L0rHZ}nOdjxLT0H+hFBQ1iO zfWKz3{DMD)dZ%UQP|@;lreA%M55O2)pF)KK2hN6utR){Eds-XZ8iN4?=i5P;RZ;=& zbc5jjDyzIN0fF7E*$3iUU#()i?4(Fe!+J_w%#wI-`8$*XE8W7__KpWzIPP(bOO1J?JN>X_=EOV ze{%v*4k2qne}q7Q+@b)nME+a%_K!a7A33IP8G8`5+Ce!hd-j1BY5uJuKg8Kxy!q7#+(dV6; z?XqJ5ARg1RkYHa4uQfE9}#r0scx#Y?@ zznd3>u2?_9)1qXy2ufnb){Uiqr)2Jx^UDcHwld>ry7e1#ilc3C8PREL+e{Y z4^<*cn&Q`A<2Gquc+S>+yAvO^il#Q5jtI+e_=r(GoxMM@3vW_aCDh0augN`8nwYwy z|ExEHkL-<%mU5aYsWiW`pm4PMXwm`rh4Ufe3t2aoY4EPIO_?g7vvbO9Bp;!3$rIX? zD@uqd8AtJfp!S4iV~z2ThE=J8PqFTVIp;3SZ{$O$(2m1?84hGU7(*N9lkmIVtzsrM zAaWBGM1@)dm5KmW?=L4aor+m~HRo0K=CZ++(jRr=jgFj3H^xhn&+RsDoEiZ1Q=TU@ zQn~!5b}8o-Wlk1fvK#p-0xtW?fj)AujA4c=%x^pSHb;I(O2BRpl&Rt&Nn$@=**H3u z(zwP+>VUfo`BK0_#(ZnTykId|)-~oxrtjnxUm2SIGS?9QF*mT~Z`#59w8`IfraX}8 zvM~vfCtx>Fna9vV2*Fw|dToG|Q8HW{ZnN}yMBof_^=FV1&-u6vd>E#Zh3M%FdSyI5p6o+kIJ3qDu z6(hQ$4+wS#sd;&(b?acnDFwCEbXXTpCpwrLQCij90{-||BhegT^7{*BZwU%{?3mEWds=J2bx>Kc^)l{_(CXT76B0El{i1=98;0izGaEl zE_s0YWvzw#A_`7TzGgDsBGohbxq>e*nBh}SxgB<%Wh(fD%TOkgkwh^=i1EDT^yFz_ z(Py<}y6o2PsJ6cdji55=GootJ#o^$6*WIGdCDUSELU}g0VBtZhg~OYwbF&_GU2%3g zk!2MgGleU|mJbjS34Ug7QPcK*!wb*iq5~Vim*tse)cCX7lbnKsJA2#3VuY|97n4u$ zidCTX`ko4GTpu8NY^YphR}nw@8fmG#KOC0PeTl!NLbKB`>sq_Ygy~( z=JycuO|Pp`Orr}5K$WAFOV8MQ=kH`rq?`TAsoHc&%BXsPTRhEE$xwe#6GrmHj9SLx z$Q(m=MFSA5yh>F@`76#N=o-_nM|gyIqU7oL_}y6Kr`U9DM|QUaa0WzJ34pn-N---Y zLx+GHtTdU|uiW%49~>`5$|wXlC7UpA81E{myiD)V1je8j20vFh-b(Qhqf_5zQHh1z zKTnup!||5zh|$>ld_kc+qNgGHak}7##O2gn4gsF~+VH!iM^>3$J@lTK#WAX`yY|U$ z`b%~w;ej<+W?Zy&oqgMt3Ejx)nL##~568^sq98=+@M$2XDFkCbEF@_gHM%@%P_(BZ zb22!Nmv>|h;aZGXiu9Qd4IYk$MkznYP(g=1Jx=G*1A)FYGgjXrdW8LiI=F50@sz}g zd;vRpdct5?lJ_5r69$ALup7bbrUeO@@|84^Cvla4X| zfa2~&R1ZQ_o|{C(%QG82z_6?1y@sZVu1(lAsE~hW z7AOEXHyntI9AB&JK5gH7Vz!UeQ`>oLlhnDFMAR^VV|>)Hx~h1&bzUNu(Q=y8Sz7-P z#NXv%7=rSO17(9CGoXH7Mnl(G`(IY2@6*EnWf3Ca%6Zne_r&nyfiMQn1j8CLaRMN& z(91xi+{gMwWX4=Dt^O+-*{)_zOzkNE@>< zmvlHs4AVK-9=8O%ovmTvNF4v|V5x2}uv_~LiGFirglk0L{X{E!e2GTGDR*bdiabsD z{XuC>V(b$T-snX|9NFjai(#wXNhhF*c5t(fIv>q3hmt28b7L2JqR#8&E1Io=V<+kk zf_8X6MD18uyLz^(*4Em#i@N2?RlF(G>Wvg_&aTlm%zu4M{k0;tU~e}k&k?Ta)Jv_J zCz3PYoWi{!ZMEF50ZKDMO^S0EA}Kvp^?qAOgPu5f0I89O8S)8A`hNZBygw+1t!}h zcVRqfUaqzXu!=0})1}+!2QT0mu{9B8MG&5$nS}*$Q{x7huWjcg$ot>7yi~Z(j{%eT6%+o)} zCqGvQ?>UGQ?p7rje=UrQT7YD`PD*7nzd~$Hiq4ZaF$T8bur=}Xk~LvNPh&GQqa!Fe;h+`Um>CrVx#EIhJyiRzigk6=l!ER; z+hxwmAz14_Pfs5E`2~RSK{^!QVR&2RKhkE^NsiGrl~^+|n}+#1uSJ`2S-uAOpva)I z4|6HAWiFqE$JZ0XSvG@E9UT*fCcl>}s+uAHRvjmL%)qa)%-M~SVwc_eOn%m+$X&?2T&`nXvb9Rh$QpRh-gA&87Z9>uMe zsEi2Z@9O$OyM^^d9B>8Ei{x^N0nSlj97ST8HRK6$kkd`U)WYPfoh;^kzotAMJfz6~ zm@nXqXCn!{s=d!~(CZ`LS~|W*o;}ut=*nflhO?0SCH)u2w=2y54g`$zfrB}aptW!) z?a3Df4S4k;Nd(|L;=m5BjEs*L>dEOzQCr9s29_(4lR^P*V~IeP=3X1ix2_n%#0alS zyD?&9af{MDg{=rz)`;sH3*S>aYsNQJK0uQj32%>P{miHF)M~%^SU_xYz%64IpX!RP z{~LTqRE}qWP4m$nTSUS?T z{wg(^Q%xUrE7SMwTDgreKG(CV7Xnz#=TU7(om&T+CyCS8&&*lgW?%_ilAw10zG}NbqJbQ|77Vg=+{+fV4FvOuI-qOQ!?E^-%NYW(I4Yu&(n=cc8{6dD!g?6|6TtO#wwwV^aPFQC&q=V#SopSBPVSoJjz0JAvVRn30GEpN{AW0P14bs(<*%li?AbQ z2^&yIMt)AnUaTSQF!Yf4^>k3&Zg=*70BS&$znHJBN^c<+EgVln00rZuQ3RdV%YiPK zqzrZbvIhG!K+1MYbVIJGXq}47xYdMZCvf(?1Qf(jj_Dxxjj2D%fd64B_Uoc~3EF#S zEpgh9V87AmTQ@%#BjleiEktff_*F%3b7uGnU5BTff0|#tnZo`sUFpfivxdD8)@{XZ zaVisF#8`KWWW!Td$i4r#Do+Q)?yMHlON*6Cwj^BB;g-CI0h=xa8$;y0e~5mFf7|Qt zshUNA!w%z%QaV{3RLsFPil*!%UKxstm1~ zahp!#?T@|pQJF`p)9}o;0&EfZl*46B;)%a_{1KS@$A&KpGtR7}$L0%#uaC&(`#S53 zgGvVa4XGuYl86c`8qa31K`yebVpwopnktnjZ=Qmb7bR%?mSps5GNWfN+11q=5lyF3 zf4P}Bl&^Mz5BeTg23i>XXJ6DR9jMw+&Cu}P30#BTuwX zOfnYirAG3sq2FPya8kqeLfWd*T3ZMS6e;vAbJc$?2E1+I1QZ`21HZr7F)i)qc_bL- zBb!ZOv?F`{?kkW0&ef6xiyt*5`eLjE>bMO)M*BI3DKE^05HGVx;h z$xfnL-H+7%`eh(bvE0*fxT`;t{^RFG<#%>vnLjL@O#%Zl1w^_fwi9|XxIYxzweE7| zM!ca-g%#vin>eTO2}m&Q?TLaDb5!}PqP8vg)BEjFV-IooF)|DQxWI@wNul$=e<;Hy zK0XgKFW8Y=XU(H|%~g38H?7~=I);JlNllJrrs+e1Y&ho5=(4%t3%J9(#=Ju2w%yh` zUeErfPOyO^Gt!^rg~%v?JbT=SoJ>2P-n*NiVIuMc(l=&36&)~doh|B;^Qi@2sjsS9 zhSG4kb52=?WO&B1=jS2Au1JwSf6-t95nJv)#gxK9D+ag1%g8~PunML4tKJvt{5MId z)T89_n#><9bvCQWw6!;38S)q@K0L@On_4Z?Hg!{df6|I(Q|_K0=Nxz7oMw~ARGI{!WQ{L=pEvVpO-tjz{?D~EC zv6=%V!|YOS+L`A47n!+^e>@Y`2-u%H>OZU*>5n}gc%mk&dSsP!11F3GyBte};UbTr z37Akg%2?3x__1)4m+8!D$ckD>##rV-b)p@=-n&EU?8ccS#95@2g^|YL!8R#fd4|9Z z@D8ETj$)afq=>}F{q1bgdcJG%I=llL7@Qay4DAv5iX+^@U=B4`e-+-VC76kWL=}o3 z?oq69uzd42W*&ENs49DqQrWtX14RLfj7`dv-1-+2o}X~hU{o}I?&A9X?8$JO0Hv^z zTw4wpvm%#C=1XSu(8F-6Vh!W@tu}10f=Vi!FWK@|H=C{g0DgttOQ-@8dPIC%c=`49$fPB4^ux?_;fV z$WZo{svg2@8>F5UZ?f$*Tr}ynwjyR5{Z_u>b+HJK~m0NdW89x#r z1d#_V+*Q7Mf1BFu@X=z^^_-;s$rtuWNYS^+oDQJn{51Cky#cl;BeB+C&8F;ym|(@v zsc``I3Uu6>GDN6&fxvUa36tWhFji|S;6Ot1P8lUj+$OFUUM%4Sd1PGZ$eyS82mXo{a+?2=e=xU5+VbH@v&dmOKugCIZX!A! z;(UqS^vjxxS3)uxrD*TM!Ug{j)`axBQ*WM0_jNqEYK`01c^0j68#81xljZ`Bu&CNY zR{0VAxf`FxeP(&TB;-PCvm~gAtbF3~I?!b~C(b6J%n@bIq*>Le_*Xiz2SdF~1AEiu zR~3sYf1JRB=mo2-vn9N#3XqsmT2Fd5+e}Xmm}G5zhY23Zbw)@Y76K1j{-e>E|KG1thp;ZCO^+w2Oey!g-6Et23v(5u{- zCif2?1P(B;xHp{B9t3b?Tv(C^I_Yj!v5dG?(Jwb$5&PWc+3_YTdp>HT3x7LJy8nPB zEliMNA|vzOonsXasTh5(MB~9o)F@fB9EIWz)h8Lp{f9;0VV`JNuj&i_J8AwJ9oYe$7S1yQgD_90%fBek_*k*4K|Y`GacCX z@dXjRz5AgE$dwRV> z8`i%$f_zoTWM(PiLRX{vRgA_R&dnd9QrPBlHwnz_0(q+X1f8NU^_8$Iy*4LYUPC;36o)2BTibhR=CSSgl?Ef5%jLd&|$J6>Wr2dRrgRZH} z&3RCKje%4`kby10D}mF(cv(O7=PkVwx^G-as8HjRz)rsVF|wHafGNlBXJ!N?S6poE z-ULbl16Fb0@4UV;Lg6bXztrI3c|&T=Wil7>ZzS3t+}cAwdn*l*XL7LKrmONNLi~o2 zwwp(N`0+G(f7CjU_nqya?{cc2hy>K3e(belYczoe+kUUnnOtg4kcmbF<&wUrj@f~4 zRVV~U9MY2touanrO+y~4?s}uH%e-eFNS>*R=(rP{MSY{(G-(xfU zAW;u3-%=Ef9kePvLTmayR123qO~@<_6WQ$LWP{NhTDYU{n^oMd3%u021h7C_b> zJ&dk%o$kV=|8fY~rFvnU%dg?(al$>?8R5?!41F&s%%EZ$#vO%$#1Y$%8==bL%j>AP zw=ALve~cy81I0Lwm#4$^`{WU%g01F|s{5xR)U!#AI{r1C22UK!Ay3LhtFs*Wx*LWT zyJ}-nu94}xvHnyu^VcbDkS3z}Yc<}xJUd=Z4|L*;cI_v!%|Bi;umYZuFagJBOx(yL zlU$sIrFOx`GWfgUdiyu(6LX{nCdN7Fg%gNRf6Xb)_l-DeI6}S9y|;8RjKQnfa-0|8 z8G`UvQV`U-D6}45*->82C4*FV~FYPuvH?QwO|J1q3UE+J=L9bcg!6E zgsIu$4gwDjYVifZD9c`HU5*p=C!Dg>`=knXiG~5LsZ!dD1A;vqevcZ5L$jxtglF=o zWr}yki9_7ltid9Dn0c=XbvA}p>4!3Pe@9_Kl*L4|;O%=LOtJ4ye{6_VGB|RS=2nGd zpX1(AadI;Js~=DxI_ce+LkefNCXp1zESfh7nXAFXip?$Sysde!X%HwLeNn1FT|Dke zhJ_KU>K&J)J5cZ2JFI;mx*N=}vIw9kJC!GBNW<-c>m7-gCpbghdmjBK4rjW*f3C^< zMEcjhVPG>efHw^Kz`J*C;pcV4>VuX!4E^0P;g~g;ed%5rzkIwKU*BYEUrmjoIGlaf zGe7vAjBMQe`mqy3AxnNnEU{J)XU&Ve)cx>aDYVqf_X9STlbi=m2sH$4^EMOF3$ReX4i)(+K!vCo3>@t> zBXhStkmOMr0av%1+Tz_w^@|Q}=gqB~7{TYGltR;_Ir7#5=PPbq*2mzvwVWvlz1L41 zr+d^iMS~buDjlafAgtA;q`KCo?SwZY#|H?7tI5mxA!KZ&Gp zhP=(h>Z1~b22(B9E))BzU!>EE=2MImR`h-k8mO zoS=afKk8kiEp;wn4m(l_e+Xre4q4OV<$kqxp6zd8_J&Jk?S)iX*}W_VHk*aEj5~Ga zZ7$id_TcR&LJH@iQhFh)WhDje`jkGg|G`+6k3`zlB-ie4Xj1bSx zKDiRZXZ{}t3-j8$ha5c*nP#EfPM&x0-7rmGuBYPIj=d2AO=ig#e@zW>HY>qMSDhT^ zd5jQOc=Y<^Vmg;e2r7sED+-}@?_#n91`%Cus??&iyKE6L%d$RBZ&# zz4Lwuca8U%1jDDTERKa{$F1>eXvTz30h3?)Wbux90M&wWa zpJJ!K@5jfE`CSNqe|He~vu7cED8#^pcYM3N^I|~Y^JY0_DLYobzbX#NB0alyq}s{$ ztniwfi+Mb<14;Mm91#Q9)^#vvM8b=tWqR1?!D3RElyMv)JYY33l4nEu)5!*E18E*2 z1y~%KSI*Z`Z_?J%_si6O{S~Bb2PM1(fM0{C2J+aQHf6(_f2BaqErN1A76POlV7PF5 zg`XHi=>3j1xNEV4cizyn{pfzD^Tlg#z3RsF3a~4D_(`)Cc9?}vvNU7SJK0sPx%?t9 zj)AF0jk_HBa+B`hEgz2`Nu6huIf8l8zDzDM@I55KTNUfhg@!)1-WgjI5JpCck3IzLGc%rkiCPTj=PJ&TegvDIy3oe=Nn%YoU{Fs^vki>ZW&xH=(K` zH)q1lo7cNAq!$Sm!3Ea<;48^>25@mhMdS-k43pu1S*YE+QK|jVgI@^qCVVM$`trEv zKo`dYcB-qnzAdVO@&}+bv2J06E9^_6BA4)R&Qb@3eBOz8GNubHwC!CfHCy>C>77gGFq?=) ze+U}t27D1h^(iQvn}-5~>PF0_Pop=EE%MkwaT2^8g-;dR)J}z;QatAvYDPkYlDd0R zUYy$orAS3D>scMbGZ%>w;#iioyHLKuy6qfdPBV|YTWYRqitLSSBTR<2Ea|41axsQE zt6ZS;@WDbkjcUp_f zSlD;KUFqO@{^I*?2Pg`kWjKRABWZNw<1L_8Eyxx#=XI{@{0ONCHc*08Iz_ArWiHSD zeTV^(Z;b8CBz&^@wWU$bJq-oR(fxSnrC3bh4Qb3LS(5OwAD_5Q@o8j`$CWF&f2*hS zpc|B23^BNoPzav}VIA#t%)4bt><@8ak%SiaHRjgAv&LeT?k$H!Wh7#17|ID9gocHd zms&mv?elzXo=$TKA{eT2W!#e`z>`6#_qF`cG0>3!ZG32N^^w+9UREx1=SzF0NLRBj z@w3f_=eb~6b=oDB;tEGj*ikTne<{Na3KdoqJ`~JFLY=%|M~hgu?BnyMFiucL}Qp@ zk2LqHL1e%9Yk%k++|f{Wz-jC9!Uey2G_o(OGR2h$!%he=$AiI|Dsg zV%Y6V1GW9D_zSxDepT82aib;gA@}|18?2)Y>b!$!U5EXc`vq_9W|nr~xa4}uEV*B@ zcw&cO7!u|Z?N6M~1w`Im@ENX&6W!{DFVrLwV2d1)nwQC!nom0BF?x7PhgmL03lF#?;)2e|4ZMYUpPQ*Goer z+V(BgX$KT_EwWd?hHneRDX&;vvP<;JXdLjd8uP8{vT+wM?h~rq3qy0va=A}lzjmrU z6UZXN7>HcviXSTz>}^i&;bYT?NGVaa{_w&x$R{j*y(>0fD>I;N3TDKWArfijfzC?2 z+inNBgs-iA&OsETe>jf?mL&FWNmyi}A~0ZkUj_cakhnA3wA#6E0|vVccKH)VsvdE& zPP@I|O;HO$zs&%qe22d&9+~v>CdCc~d z^<8&%{0eNVWCiV`E z@<_?A9QDHAnsNEE-drZ(mA3h00e9@h)vyHK8@oUjOP=n&)XH}^zvfW|w5Ygfla)3T zyg3@Ye_#)FEP>ZylK#eWMnszi&4g>gL9)4S$kcMvi%vdWzm&!!X{*%>zN~ie^3B^b ztX$6fu}bEW4}`^DGbh{Kd@y;lBK@XsB&r-rvqsi07}|3YWgqD?BT9Wlz20tr@u+E` zuuarVhV;oslyAK5sSCC6zrr9KU&4CK^(>Qsfq4n`b7UW3ciU@M zf8q3dgviijjrNeXmNH%)I*=+yJM0_N&0DsLp3xEc&p@N9&>j_|&pFhQt>qFLi?nmN z16RbCWqe*zg(OD#U=(%l##(4Ei@8M~0n;m3ifolka(xS7YP2N676EObi0f zq$HM>cw{7<*DX|wgX%_jJ_7ktCbq6TZ8S8ark0`xnW%5z#qygxJTWPrZoN7NlqU!V zMerNf28`an^03Q{>goqd)EJ=VE!l2lwA_D6o4qW#Ahk5bw>0D9)x~9tt7pDJ{>Y=x z&uMbb&eBS4UrP@Axg-p8c_tQnhbW!o{{SRSIh>auhXxb@IG0iS2NSpaP6ysS12;4< zmr?o$6}R$?2gV))H#9PrQThiIx6i8w)F=ZtG&Gk{`UeydH8(H{FHB`_XLM*XAU8QN zGMC`12NVQ2IW;+#5$ppfe~q^VRFrEMHcU&1goreQg!ByE-ICHMH4Fm`F*7hD4bn)d zbcck3bazQA2-1S2gfvJ>f8#mlea|`X|F89ZYu1{1uHAdz*S^;?>@0c)dlJ* zj^-2K2T1~yw6!$^KmZU(h#v$JBxYwfgrS|G|B(~38$;btFa%uke;*b~Zcs29BU1*W zF^1X*IN;GUXMlhZKtNPdKui(@0tkXY68|woxJd$(!R|0yfHpti5dscH5wj~HkX~*u zdj~Wo&wm~PoDeR6fP{n??;q{}1sA9r3<8D&w83Zxs0$_|1ndkjKtNznwAa5va7sI% z(MU-k(8I%nAMAqSe@D34%X0AoJYZ-CfIbujb#sT>0)Dp)&;h$Z|7wk&m>po~07Lz$ zFhJO$J-}{I07l>pgFxXZjLS2)Ez}Kw$qq2k&;sZpq3}PAwf;2V1^hJ}fB?V1f5ZLd z{W}m0{)aOd0ztSS!Ei4a+#X;DbA|$RRkirho@ibG7;gL9e-P}9LSX#C?qHZR*al

?h*Jq3}(zKa9bsWiwhKvMiKw6PZ{P0g<$6H z1^nx9o!|%$xX*u(9Sm-3_qz?-XGowC9On8As-gUs4Ms%#x6B@j28e*f#e~E~08m!| z)Dz+W{GGtie+voyLn-iEjOoG82Z=xe>@aOW{a|)b%nz{-3hWLApxvH9{e1qZ_%|XJ z5CGW1AZUOM)E)*W{yREG47K|cW6s|V<_UNT!q86u0Q&v;?~?_FVYUdkv)A9|e;hGT zPf=S{SDWXrmjA0%R77|JeE5U}0epfYAbEelMIe3=Y*ppkTkR7>pMP^glYx zmO-2_*9Z#3%|BIO6lUene;nj*48<(fzw1zeLlCyVw@*++6aaQ}1A7r;4j+R=06qek zHMNC${=selkROgfV_X23p8WuJ2sh&22Q486e*l7ii~dC70HDo(k%$BUh*@u!zx4tb zZUUkIfIo$eM;wRCLPQFQ&4H9vLevr^4m<0X8; z$E><;&usI0^5z74Jl^M&;lU(X1J)apBt<64cWymcdW=!z9Yk`r_B{sdWWEopy1vGH za8MPhiaiQ#(P`_bD9q5JW{@!6{;rR1f3)d0TopN@d6@m(9#ZC zIFEAj#lBDdvlQKool$|0+=Ev4*}~8iDpLIn%EGtGjc0T3TxP#>-%6wOu~EA#zh&D~ zyug+8t4im0y#^inBta`w-#ypEi0TEx4us{fBqC@ z+?5Jb|1e+rozdx0c&Mh=%y`$x;^i~VWR0i8y&;^u+Ho@1B`e#0wd}1i?sj1~O%E%Q zC{|Q>s57`W{p^Bsd)b^bXHPgo$4vyEil;cV7|On#l?Z7HFr^R@B6WwQlo{Ot*8b?b zRbY1F<1Kuhd)U^d$T8zu9tDq~e~k8Cte*8v%YSpeCjk6d14*rnR`T|6Svv78_gu2j z>T^L9_?3FMxQOg5PEov!Mo0pfvx+v`BST_n_?sdglzv<8esRoHg7exIt57GW`1~l4 z`g1R$1xtw7y)ZEPe9QZ43nvYrLyWBP#(%GOPs(_--CndWw}%8rcAEF*f2V9yvTnd5 zVHX9^;FZ2m-DD#-RliRgPcC)S3Z(?aQYh@%>acdgimp7FgCF1BZ=5elgQy<&2XDl^ zif&O@JZj92;u@=@AK$s%QYaHh$S+a!BV+4G$ey0=Fhk}J^V2pf<{O(Egw?%$MF!*i zh4bE^fm?M4CJ&f*XmM8Re=^fQEP@X<5KmN5y0Kg#3Hj`Mwt@$Z9??3@rDN){$I5=k z`-I;zV|&aJ@3idHR0=pgDGRnQyhG3~VAIx^@#sC+*SgIi=AF%Ru1*iL;62Dw4v=HQ zew8VfnAb0R;=3shvJR#iSK+552Rnf4A|*;W}LZaIu}y zDh4WFUd@f;?VA*vXaeJ}ml))F*_WEG7-sNt0VOp2>XlOtuVu7NT-v_(KqDcTWFS=7 zN|9iB^DaK+>_zX;xxgOJ$PC38HrcjGZI$3BBqBrkBBMA0x6M=bliieh_4!j~3qzHY zPRBI;{z!iy&jrcKe+u;`AT5P_&RKlL+jKS#S#^}1o~h1ad^#IwTgOrtAZ06@n`c-+ zHZ3B_yL|3W#R}1U#XUPI#p6AE@0MwFIi)wL6X1USJyp8SWg{sgZ&`XMzulLIK0dO& z(~Y|chS{2pST4%9*UAwhZ^MM77^(sldm0rOp%)iSe|<;Cw{b@vlxjjY)~tw$k7O@D zUK%1sWb(Ng?lycJIQPmAm}W)ZC16W68e#yy4X6sp6nQN8L`ae*v!1?R+DvW4N8r~D z#8`v20#}sN%1!&huLyLgZ9qWq+q9blO{aLcH9y-I4Ahe*EtXB;MYy$bKLSZmFTT12x-7^LvdeS|sSF z84TLBGOnN=YL343Z4mfQw8MWO0(9V{826(0+Id&jBW~VwQ~JBQSxiMQo~Juy>s`Nq z6;(dA^atrSR4kme-cHJo^=;5n)qQt19P=Bxe>0s(R&;1jgiFT=gwIls7OH1&92L0J zt?fs{AHT1xBI11eLuw$f;xVZabBGesX8p6g)UPfjED7iahw4+#DB|49(6iTPZ{>z) zV%)GIcT!rKtd7b;Z&8fe=@OdEf5BNU%r1O(2AVsKjd23Tm2s4d9N+#pT6k7F@s%jB zf0A!O%gp%6dz)=x)2ye1pl5@Q)Uqn$PE`R_ET%mjTCdyce!c07HA=2HNnZwNhdZAe zsx$P%N3XroyP-zCy*HPrPzAi%j7UVN5c80)1v8-BB6Ver-n%hyPN$3`kuE`P!bm{A zHjE!Afg{y7qJh0Rtm>V>_O>m+=qpeYe+Soj%1+eNbuovL$cwnj!>=+(T^(4>&3^vP zCebOr{L-w_3*q^t&arGa*)y-opfv^8+^d!geqXLB%XmHLQ5`MQ837M-KfMz;H0LR` z_w1luN@W0L@|2wh_uNxES{WQ|N@}4oF7}q6MP=miuJ6n0bUE?c&H0|NMAI^5e{DVL zB39>OaC?06x|Z>=$VuLFLmC?`KcMW8h!f>D&8f^|k6hhrF=~;^7p(M>iBr6}Mpo2X zdUCO~eV^{@D3z@lXrD~L_9jSYWE1aD^bViCSiQaaQ;2)Yk=3LtQ+ZQ-J&NsM%=aLR zrCv31;nQGb!!_N;j8?pq)pKIof5%CWG}!8zlDf?LIlt0SyXd>F-#yKH^pqyf-)@o| zSAsXGShV{4eJg?UyfJ?d07>&!5rY9QzMQ}-xnM$)!%2nlo&_&w&Y)#2?8aeqa}G1^ zp3!T{-R*Uv4%%WD`B84_i`Rx!60g6hG~yJP=-TtmBu#cklU1m1AlGQ_e}L5qm>v+C zd*@$$AP0_)F@>l&8sqz93PcO3oD%x&@ z8q1VfyK*lw9SG0h+YH z2qLomS!qc=DhSadLf+@8rTn>FAtkcz`Bjh3TgET?gaECmoeaBP&ox11&ZMA0cuv*K zWX3nxY|Q)Nk)mi6Y1ed1Oz`(}ez$qT3S6c`m8V5~xFuKh+sAj-e}fAHy<3x3U%8dx zlWi`JvK1)4C(*(IM!x$449ms;s{Wq&l(*PcqR(6Kv@S8#~mg5rZvTLR#FRR zFTCU2*jKE&;-heCX>QEwO>!z&I=X1qWGZusnv{ike5v#O5`%G z{1mq|rXvRFKd@i6e;+|YU^*MGa3$`d;5~i(BLmO2O-Jg%N~%8%!AM+T8*-uLB>7$g zt&-Z?I|1=JU$QG+b}#r!0W@#zijsFIES!$_SzE8PIZ6UaaPDJ0seTo3b8YmNo%>vr z-u&xfo#NV3{7kI$r_EKpXf`CSV`)G~?}kSgI+O=iI1ek!e}D9TJgfe49MEc1&@-@) zS~qLTbPV;&B{u(nR#)1*wfE_aDHTYpytE2v{zT*hgLW_9PQ@*eep620wmE#|`+?sl z8B!7M{0sCl?*->O>Y3+Cgeg3`pBWk$rk^xiBiR5PKgq%G$ZFVB-p8q`$GBi|E!>~a z7s4+=FX*qrf6|@VKU9Z-|$wk&q|y2<5K zp56!}61N$P{UwjGy#LEn?S?-#jF4=LM)x!tPp1*p=zsEsebN-tiPX zw^s9SbsvI0-h98%ha#-qHGSe4lFMB_LXoY^qSem}^`it*~$rSqwe^B0_%$iE9e0fs>|DK@bqFZ%=WHu+G zx@-w_eMWC^u6kUm{Pb(z)SD1L7n6()zMr+)OA15YC~K6B-|MXC>8mbOgFKa0JgNnV zlbp4z`l3CAOgB#DLvgb6Zq&Skxtg^N_6e9tJU^|TjdT8>-_l6vZntQfhJ?DN?H1;(-ZH4*+olw_MlAH9^iU|DEum7H^XO%;TSa^)IJ zui*5E1xlFrZIcfz@bW8=4%ioKBOR6)KYQhL0bKQ2AlMj5K%@`l+zom4xfgeNyFgrx zogxA`)beqww_f|FLgh zFSPYPRl5eMq$~{)Vdc-u3#`-?#CXCgB8}g>-i{q?C@)SZC7(q8u+P`Q_VBV;m^9Cr zMG3mP>|V=d@qaDQtUEuE@;N>>Y!nV659`SO@ammT?%p{0B0-ze)?&BqqGzXugNjUF z@VgfU7I7AC0K>QrKhmaMji8I2e{{h)+A>zsf^fXl>&>30L>0-z6LFAdS6+=q`hiZ1 zxUVs9B(lIWob8fFktdkC?GMrsY zIR!1|+EaC)^qHFL37~4me{H#Uqey|I)1M@kF)0UksP0i%>-ihpFw1Dp-N>n&t!dJy zr*l;E97g21nX?HJu|1uN26|2+h1#&TzMC1nDRB+PujR(E&T#`2ohp!&#v6(E_VplM zb0bl=8O}&m3QnAz-r2~*AM+dzO6kW&!?WTd1igmC%Be%9>-?CSf7dlPzG@m92D286 z(=A13B_dTM+dMvHW{ps*KqyJ#v%5;VWHgR9BHlgfv)=!hV9KTMxi=rT@fx?=lSpKQ zf5xGZm9R-!opv*O8q|*isrBz{q<&OzdpWFOFR(z;6_2{N-@!|BuI5@*O4{ru`%CDh zf0yq;as2|(IK|x~e?y>K`_)_K@;!xao6VNKyqb#@=Mr{j?;}te*dO&HURh5`M_ALm zZth@BPKEWSK78~v@xkVj6YE^|QJlO%&bjX+60Vygb#pv=0xCb=gi&`RIb~_CfPthv zi~RJf{N2Rn{kW3v^BZbk`Y1Z%@v-XD4B6H-v`?12qOj7$f2oCm+Kj#x_3f@zRT&P^gk5y&-<`}G4hNmx&yp$8B+i$s; zytQJDvxx%VB>8%9QD|5%l9hJBz+hi+FRGP~|52tPfAWD)zk)o+nwNa*4)yo3ARqxj zC{9Nb0(kNYJwiv~Ns@SY%ri*MEmpUuK5X^tvGfs*fICVzpfT7GyZW2cgh;ZjDe62) z{bgmat&<5-Y}M6Ee^k1UX~~mB=7=xD{RQ5o+Ax)0g<0dSHzalZsC<}@3x@O4#dvao zvqUQ`e`(poKB@@&+HFV%SV`8-A)spLr(*H5Md;1KI|FtG4($f7m^?G3`~)JKv*m`j zF+4;H9``i!h)-Jq45hIdM0qNrD~!%2ZsF1d!q>mFKUvXP7`{mLf0*K4&UhjB*o)Y4 zG4xAT{!+s8PLdQ~z6yvIIbO(7jUgUI_O7>oe?J8UPR0$r);V36*myyy8qbrCfo#R- z%7w`GO_>HlZjQk~L$P9U|L5VaSyM(Ad)u88KZrBmt7ER}X>D+LbW#7dGlwCU(=q18 zW|f;-#upt;&kGri2T}Hboa^^0YCfZ+jhj5MO8Nvw6HebSb2GnEr42qv<}2*_R((eC zf7}}(65L-fTU>2hf=jOQJpSSnM;BGlSAyPm`M;iPHS)`WYUcPRmS+sCm2P7NiMc`z z)AFg`N!t{VAd%>*`uWicw4sTvbr{3wzA(S8C!1%S=MRdXFMpXf(B_m3*o$kbQCaPR zs#yXj5l7x{O@j9a2VjzY1STy!2YAVvf7LDRmy?Eg9zTjuuHl9yZ4N=AZZn$)TJao` zCT5Y{Evd19W&8K}wOY2a(GlS|nzzzPy%o^38iHN(#$_9e{NGF zX;_SB>O1btVgP9$BUl}fBZeF7-3P1KeEK<@vIX2j=R~80+x>zfFC1Fd_nLe)_(h&| zT0hpyJDm*Tvi@0={P=NS-3I67tyAH65t_tm%Xh)^Awg8y3+0FxjD#X;+}3iR^;NYC zNf)~Ut|tcVmtSroJM3l_aNxMte^_lIy#i#mCHal9JDS~FluxyYy&7DJY7{%_*t+B& z(!Tg4y1`9+fBkqk}^H<7H$T9Ja-YsEYyIF^zqAT3$4dtpgrzib(%Yg zQ#|$ZvQU1$8N~asSL@eGv}Lqpcu8-11DrpR@AZzD`_t<4++%a^LL?~{q|VrfEf8LN z1)go@tb-XC#ZXnfT-1>Lf8Bdd0(W@w^7%Wjelqp97N!4+VWsYqr-tNd72k%@J9bFv zR0OXtDk=JM=5~B6|6Z@rFs3D3S_$eq?``3jrGJ%oO=YuosNC67O+gv;e9tgTsJ*s< z?ow$Z4@+9+wsMVqyWdawR=Hxe@%GT0x2$VpHN!`rK~u! zsn_DqnGRcL0W%YHzTuoSnk_Wp6oskE5lstCRbTqA6-Meleo`8BXW-1ccjqBb%lT4E zs_yCoY1Sm_sw&CAu)?E0XAgU_R?SH(%N>0O_Gnz+4YVFprj~hmp}_?CIgaHWYV;aU zx6XOTB5>Yeie<)R3N{XEw1lw?DiR*#2HNS&VsflTFcKVlsnCq}4JM12Dj)EV)iuwD829^x1@@ zb|ys^?*0th%Q?nl?&SxBGEXwd(0p74gMiXBbPKvq&sFC_FLIZ-ZI2~>NrjQmmfTWGPlKyy zdOwkd^UlW1e{#h)4FKR-`}I=E>jdzBHTU zdEgO)>L|$dVEy{>bBa6sqFD#xk*=6}!K0E{+to$C6G)e=;CyF_vem!CM(fA(!A=N5 zr*)vVKiMlQQb^Y*fIJ{zx7b-L3k&wb$+n6x()8wujv%uJ$~NLKJlRtIRf6l05^bdPKWj z5wd-L0oq+u_POIxD^PhuLD^Bt&YI>saRcF!KTw?Cm*S}F0~nsQ=&VG%F^$vu2TaMEK|t%v9+M~&NMGO zlV9liq^8;Om_n-y15Rfc_(*5wl;p<4skV zuO`PILep4J&GXxbDH zJ!I32V4|o3`_2^sT&K{>oo2p!97w>H>|_8|n@TYc0m=IB;XLF3Fo*imOwly=a7Ey^ zvJC-hSiz zC*?Xn@XS_kXngXGV6CLKt05k?U=!<}H|(2d=WG&(<@71((@#$PZWFx#q3x*g;Td~j z!i<$`73qqP)ZVi=P@axK_N!Em#J&Sr4dVS|mUYRtEhhQHv`$2p&)pssf3WUwNSP|J zIZ0aqY~a)F<%+X9*8rxVSDvnpceD~RS%lb{g3_37iq@Gv^%Lz*66f6!1FGQ`EGQoN zJXE*m>Fv%gqw}jxx^`hv_=IB*GOkiMFmG?y=osR~D@?L(B64p&9U6vq=iUqbvhwh? z*nF|UV0RPt=j(>OR)3Vbe||{*P;qlNwenr#L#GzK3MylqL?s5=vqHf=gKV)l1b>6l zp|YMOjNFlySTK!V$!XX0_U5SX0CRl3u6S*rn7MGQ(Sp8Q_{`n#af;x^ z8YfznVLSunrM}YxV6%Hg3r}?Zp?(YyH%8ZoIxrWs9|!MUZ#w%5f5FS){A%vF&8rUc zMh0RLGLhD%5s*45=yhbhjQG(nJen*lbIBdO$j|*gX{c$PXvLYE;yD9iFrbpEG_Skk z^1DOk&tmD`W14*f=QtNa+-^5Jv9tI}!``(v4a z!O>KjNF`WjAHIBAPVuI@yHOg$wy9ONtzk6EqdkjNet1od{mbWN4sSu@XOcTakbxa+ zaI-wsp<%Lw>cn%>RN`U*32(g?F@<5i4G9v0cJHNFZ3kn*)}xew z^a+VnWo_RzrHNY#t-Wv=-MzFAMuqY3?_I>&-5UdSY>4=7f4UQ0@VTYS{mER>%4PT6AXSYN>{--aNF$>JCyQW!SovwsepLlg};k;S0as zh-REbf8$!Mf8&08;YS)fSs(FQI2SKye4{~a`No^pSN5Xvb<_(N5;LnpwM_n?cOGF6 z9LmqG6>9dS0Zel*l5;z1`p?-t+D%?jQ3r4$XqGz$Y@dGTv6ezC%`_DGHY)kBdwl8)G$Meo&aKw_SYG*#Pq=XI>nyZtrg*=za34&v zv%jR<&+*d#c1Wyo@Gw{!IvY7JjAhKT`Zed!;dIV|XE5~ufOl59obZ9al zGBY_eH<#e62NVS|H8eRlmvJ}-Cx49h1yEcK)&+`UAxLmSa2j`acXxNEad!mzVqgF;F)(oc$I#A^3m{_XYGDG9r2|OW*#e#5NQLd}Jsd5}%$-4L{__z) zVN3~NQMB1~ND=~Znl99)2sBL6l4iQwL3WhR z=Kq>6y^_3=vXUzGKNbI1CnRL&4)CI7VF1uFu`&P{|123h=zoLH|AtX8wD>oTzkDTa zP3-`j|F8wM=|7BJ|GfYd|2_^%!2iaQvjc4{5J2&^fIS|G$L9T&%7CrKb3o{{Kg9Xk%gR@oxjr(z-Z<4nWopbOyHn*Hj(&&(X>PO)Ok& z{;yWj*${LHf`7JV*8gpzg_D?tJJ3YI!r9pTpKbX^uKMT7tSxMT3U*Ewe;z9UTF_|z zj}CNN##W$5hZATe|D^(gF6Vz!irO06nfy61CRR3pp`)Xr2OMa~AjAsrVg%iy3DEst zOAMf=v$b;uxd1@r`2bAq9O3>QGx z5eJBi!W#l{QG7!nE=q3*#6|fHfw-u=ArKeUHw5CM_J%-Q)ZY+@i^dxQanbxQ;skNg zdP5*T!+-xGMv$eUF=*A^^q@#aZwSQP$kEW)3J7{HnL5A8ng1*Q=eGW<00lIDL!feu zEgX$qY)q{|EASsND=3Gtoi%9P{v%=eV+eYX{>2Qmhx8_IkO{=w!~*CDbOJrJ-;6-~ z-nfI7j~@6xbQqaHL_zy%=w$x3HGg~@TtHX;7Jmjb64N&bGBEuMviyM-f5imhH^|ET z2fDtoVEiL^3(oq-+|I@EuWTR#v%erH{u?_M(9F#}?9G9;e_4QJf8}QYm1+4G1a-#h zF9>Sx-=%>DV)IvNpsv0}W(B$0f=>J|J*alOH!>h+yZ=#xim`uFgM!(E-rTnTvl}do z|9_VL?`ANAa@hkNL2u~4OxgY=v#@&`Gz+LX_SPhfD!R*;hu(8l6__muUI4bb(k;jn^&I9a^yHVdds zC+olWjgb*lxbUi+!6R!SDF48SbtcX{I$uT&N{o<{S^Zg*5xk<8nWwO5JcAP zuXrGD_rD-0qsLzm)IrZTzMx2+K*xVl{m)a~*u@bvMCX6KcR?rgAN;RxQ9z(O&=_uE z*3Ovwi{;NR&3Bc8IBv8%qr6k3JL*Z4v|bC2EiMm;@4r)4qy?-w-Uxml=tN%L5`U$* z7FZ;B_Bv{)ejm~pr`Yu9{iGkGG_lhJH$RTiJC=JS_^SsG7ME5=Hu~!1A7-~45<9|rzJTI+vrl|WF&y`XYCZ~2R5-crDJjiR*;R!M? zq5TA@Pjg;{dCR@??hQP71epjliOaOebo)%H6&6b2wsl=quEPthjN4h?0bj^*deOGy zs*38+^&N@iKUOs9hod=HFWrQ8}A0{2DdtV)C>4~_;JnL7^y_fwZoQ|xKi;F-xWKLK(wm^CDD;-7G1w{ zJiRjv;_(XjE%boRxS7kqpvCd73w#*&lr_uyca%?A&AD9b`ZUU%wHvB#;A1T`l9R=8jG~|^i(zM+rNMik# zfjYI{FW5pj$|`q1h&|WAy5=TFwg`tNwn)6J%XA7@y6XYltP#nO)-hy+b6b8XmutKf z5n(T;VfCx3Zi3Z|xPMit)4jm8#Kf%nay!qsJ0}+#>R{Rhc|$^w-(A49iW= zbb$wg=z89AC!L?G=+`_!CA>QAn&2dBBK)DMhp^+jPi)^aC*8%f6>a-Ua8kmswk;}1 zXX7&n=O*A}n7}?&X=kLMTy5p~QW8d(25AJq81JKm!+>q^%zu!h_y3Ge_^pJdz$}_g z_=T3M3WXeH{bgJbQ3|Po?%hK7PhqK1Fe&F^FtRQ$cIazc{pMS4reOW6z@PnzTsjN)pAHXu;zhqB1!hqKw&U&Z)z4emAU4e&%aR%>_ z*Lmzt0&EcpftxM4q29+)+J*6e`7uRR80?)z3P!brbJ@>{DW4k;O#IoG&&K$qI%^JR zemaefyDai7u7ujD`nQseKPu0DRXly{-8Bx!W>~{3xqpzAx|hw=WGEeucJ*Rg!;GKR zx$6QzIA%Hwq!2M<@!mqlr>jhAO4BV%V;&ArcF=qhAC&eo3>DTA63=LX!^iVjqCgCj zCJ*jnl4qDwODt!TR3IU-emr<6t*4Fqw65Kbc{B|i83@1Y!-m0<9-bp(A~lOQs_+cC zl^VuMrhjp<`z%|^)Q#!6>Jpor6*A36mC{6&kI>VJP%`n#**03Z#^0SI9gks5i8(ZO zyPqJ0(tHTXWj|ZU=2Q&&6ITr*BFub(2~1JH2it8x`{F79PYj{(3#N=MtMY{# zSrP5gRAohmxaVCV=OHF;4;kIGMi;1Qig@3cT^7sDM#c|H|f9l2xVGa%#Bl z&v=wgpAIJWn;^cH*#nM_`ZawQ;M$~ z2sN)J6s<0V>$9gu`3-7Wxqrz5izfTbf3@ID2^FIrBd?`NEV2)>xJ4nIZofqaYJbAd z{SGQ0Q(t>jLoF2LQ{Uy4Vpau$io)x)Yhczr9B_~(*8XJ?JJ8YBJno@B&IRN5w=WR} z)Ye|uFts-NqE`A1n(Seuo1q%RsxmPM`**6$b3B+LxyY&_o^vM4fssvQ0d zDB;|6P|K??7|sL4I_X)R{BSafO0`AwjOl=RUzW3%%QcHDli`V_8-;RS?0*O!miMDL z&{a;PHC$Glos&dT#^Qx6PH^4~XT#Q-%>z6{3N~6yzUG2Q8g}PMVI)^RN>C7U&^0*D zcXiVG2iacDo#xJ5HTl9QnEK-QP9RzP4+ei}zzf#L6nYC-i-k!SSfdQ=PGLBXzHU!J ziM*LiW^gzUx(n?r_lC_CXMaXPm$KsY5B7p%blNOA%JWP{Zt>K{0Oz9IQ#l|ura7+2 z@Xr8r5BwJzVl}pVRny4pdK&!r3)dtYx`F_J1=ZIR&81_G6){ zGm5#7olfs3t^v$mY+@p7+|@j;YZg%^(us75JwIb67TCL9bReYT1;-@c%=Z-1ISO%p zCHZJuuPIz6%=pmlTx=phd|Su?Hf|*sG94~m37$2@ozs3e9cS$?_CNs*KOZ4?8I}*u zy}prW@>1IQcmf_j(s(O#?rq>EWrb1d%bZxSMS)Qy$%S=i(wyzrgtT_;LQ3{-veu_8^HJOM4^>zP zn&i_;Sz9{S zvr08u?X*4ojB#QhQ<0+!vt*d7nhoC2biL*vwybSxch@8?us#q#23xcM)392 z={tI7@8yN9)*2*LauaXS?V!{3xm_x{k$_i>$1FSyp|CJPxYs9R7&>)dt=i)={9T4m zw(#pl_CaP&3P~m;v=cfv6{FLsWHLPVZBVH7NY-c{yc<5JR&(q>&<%7a)KZ&{_IY{~ zx&{(#=zsqd@+=<=$qc??;&+{HhJl7SyllRn3+d7TH_RdAz?#XmflF8l!8CJQAbbx8 zbH~0T2CFIHp{)D)cTgroa@oDLozUz-X&fv;Lb@vn!s7C(ugtdHJ@@715GxtGN8XVH;0(TJFZu76k;E}wSqHY9Y2rb@J-?-&zClTVCU zk{`5$MOdZ#+zj%9AQKNwbT4JqW4{6w0GPkT5 z*0+C#Zb1|Gx4^vZ#}dZ`Wc{mS+hzYw7=%{&r!8kexp%ZO2(}5~iTv;MS26VTgO<$D zV{1NVVDv_FM;hQ)j-M1t3x%wnvjMwnJAal5Dq`y4_~Q4i+Z63%g0V|}_)bx_2iKVc#v=cq*Kp0X?F=PVWhroo?q2dMP?Ul}e2{yz@#B)B14L9&U%ZtHdBb@I zGIs{gHwPTs`E3_}XKNW@#yE{`0P6^LBAi3aT(`qyo))c>|YDi%48-W9gJM=d8a;(zq;Ofu7FmKAPrTJCEFo}e+ zb*+Y}3Ko}0RiCz=?!B(7LTGeVzy#q?g68+*L$G_oX>duHY0aI&sZ1zLPCrvPn@E`$Jzu1+v2?}huyR;ZS`1G#8x&aXh%a9+)TTUMxKL)rjC&2B} zb@DBa&sJ1qk69&_FIsnVlO7Ty;lEs+bG^ZT3@uO)ZTDk58i*? z{7Bvk(jdL=Gjo;m#ff+%m z5?VT>(W)n|^X*{x2!G?ATY8x5Lqd@rq<(TLoAtHKyqM)0N9+8}Ht15ZcNnx&%wKNt zEVkw(1eU+~n~nuOTh%Mk9i?7ByOR6$+0dbqk0@_-iZl%`SbIDXV6w+y{P>`GJuCdH ztjFe{Mz6L)+=W$NJoWuu%>rZoBa$dZq^d$T(`cXCa7&mM~5@PLS|FJfHh4N3zt&A z&Xw)1iYtoFQZh0=%;u<~M>mJMPNn>ogsIm540Jh`w zfg8Bo2|!?bLzNyAe4Jel-LFpR)mvo4Ii=6=n0PcWJb%wd&F@DD)22$;Hcz0}wkMkkhbJyZR#_-E0R3FR1%Dn-B$DFVopB`U#q}?<1+IR1hcWgtPqLLVF zsSv#oE3EmoiaamFzYn5@^VRQhoDkU4%1n^;=R885NgMD145|xBI}H+koUXJDI2H3Z zx(+q<>VN4xeWi;1IQy(PxKSu)WEw=h)ZDuICBu#gHgcgr6}H)P;?nHUY{YOo>YGco zSk^=`PZyfTgU3pJap^e3164CbzuC@FZs#|aLWz`-AE|Q}DS92_4l+QypJ_>| zCzqc;+~&o`jyy;xe-|)O#3CA4|nwa@Y0f>x&$!yPyVcz|A9ikV^0GweU+lzlan$ z3w&U{1)`>;X70<*3-DP|@~E2|Y|A!>^1aUe70bD;(6Oo$nPHb${K=PR=5Oxgf z|J19}FWA2iZ;JU0k(3h~ui(S|E{tfZS%3Owig+|eC94im?Dj!t5BvF6Z#oVV4;vDm zI~lg!`%noWqSGCKAmK92_Nv(PJf1|JuE^YP#O+R8A4R}?NuWH;RcvX5aqKB)?CjH?+$;QhWsQy%hAN&?$|m!5^;u~D78cV6 zF_Aj+=q0-OUn&rkddlx6Xz_z=-F%*D;6t_K>P0_}3xi)(+`?dbp-hUAp2&tOk9Zs$hThkZvMO*+$brzXT5?+TO`ODt& z4c`;jfSUdz!~ow+J|~qVgAc}3jSi95{70BGX4OT{-roZ*+hS>^ki!ePLDvdvg@cr$ z-$FT%0-jX!(jMF>%g@cw!iGLNOmnT9SQ6(oR4jiCHGl8W2cMVL>|}>P+kYSs3nvXJ z&K|u*ROYQX9s;pj{ICPiy6|-J4hrze>6V$PRz?fUk)t5!3JK9?2Ve$hp;$aiTX>aL@52>|=TRMEMKW zu#{^jjWmHbIP67r(Aami)qii_(%_PwQWd{cYcH?zsCH;@;X(y0F7X>v-;l0X!Jf!9 zEek2Lat7Z8cKia5EY2Y6n(a>eWwx(!(7dv(XH=x#0|UK8@so-K5a9-0dbdycF)Si( zD@YGrj?l5JOVcxtCoh=nlVc*cpoyd~jsbnBzYiR(B>I1Yt#RgDC zWn<=aQ~HIzZf268g6aNdBIQwR*xmfR3u&&mfxmSyd|c_9(?`76=Vg*_?>6xT56t4h z;*h`&hSM=_MAggIbAQN(O;{Pfkg@AnaD&F85hoo~@Y9 zq<5nXTYzwHE^%K_54-%+o-};m7H}ES$H;ZS5BK1$xq?+RVh4=9i zB1mv0EYE(NpVrI_Puw6)Ep|aC;DoDe30T(W(i&JRMksKox8Xqx_PIvXB|&KYq8e^# zsmW(btUN_4?|qngb&6K+#`wd7zloI&aoFB{OVjKN$} zL{S99GEzAP=@1jquo1&&m(h3>eRI9cbrDPt^9`wPW44@!)=7|qP4 zHRWDl7_1j*eZk8Oj~>B_1I81$$wyQ{UIc3(Kz}}v<(FE2{>FTmc#DTVHC-QRnvX4B zmG?E?Y7@OQv<-boq!YPK1Rd9r62A!p!)Gt0_K5JoMjK(UIMiMb;&6i~5QvsSB|27E zU8Sp|sotz(g%#@*D%3@DBB_Y*b5!e?QF47HUXdIp*-j;+n&pcjYO?|C_(tThTELAM zEPs#n<*l!oL@oCR8)epHBGLy5{b6Bi&A@7+siH9=52e_eV4>hI>xg_|{!?h)60Z)PU`u9g%OCAn?CU^x+6fov}cOm`10cG^)rf0XgP*~^eVv=3_imVTM7^+qyFjF2!DZ% zJ|G{d!EWa(!D7?e(F| zRAx!6W~nlgkOeP;9a#9M%4)b;BkIHK`;3`~$@|h1g7a2_IYheA!Ezq>`8H{TuoDdT zFZA4T7*G6WFWunk!V-;;;fFZCNq?(hc3607_ZqmJWN`EOjg+Mc!Q12CK}`~dNHdU_ zSDmqdP2BQ+eV6vWilWpn_Jdecc0TlNTY_3SOVL5lo}eJ@6@d)7f-Bij*7BEjcuMir z_Y`D@2>2Bd?_M|;R9^%ObSChj& zl%?xj!!V*E_8Ja%R)K&Q88#qp}spg-U_+ z>jmE^^s|w%{9KZd6|EZR`O->b@yQ76z$z$5k-~Lq7@u;_h<38>ne4tO zdQpW~#Ksuv{qxx~RHf7}v2=yC0!4au%paQY6XWyWvwhmO5Kc|<@P8;JvZl4hN`*ze zSeKoXJX8H(eYR@}3^Bu?*JciFHqF+m{3~#kUj6V3YLV(P{n1wE+iXjsaggU90^^|Pgc!2zT$IK^d9&XOF&sH6ht9UpTF9)q-dKKNSC6L$;ID?wUC5+?Ru}^-J ztPm5&*m&-ld}&OW@PFzROnktMwzPIaq!Q4AW(H5vn668Y^WuZrj#5*ie z`GrEWaiR{E)~a0j)zs|RH`3ubw(vKlnPdUIS3Ga?=u`u+et%u0Cpll6cmgJLo|4{v z8rdtw4?kUI(3p15C-Je>e18V6pyTmW@uwP7CC>S6J)&#`*4* zCZAu=_<%*QIDenu4lAyqI!I_5cdN#;l2$zAd72P^L#p{EJ?2UFfhkYTs5W$e>W3vj zdKT7A!s9qa=PTP7+ljRN-Ud4|{Ta?Y;oe4BCe#EKNv|{|D@Nj~3dP^UIn2cf;aURi zI3GKb42SE1#J!)Yh)XVj6R+;Tg`tYUx|*5lgJk_Kz<*D>ZyCChW&BLDuZQ5*us-T> z0c>i8oL;J;g&MMsRW}sOYn$vOdKxSNj+XhWn|c+FllBdMckgYwfbM0)-T?k1}|Y=2zi>zz~=Bf8x~Qg6u@*#AV;Nzf8? zyN4>OLxBcwQUEgzCXGAD+eCX(0_|A()%2yvowR{w1l2xRU(Rtv|7sZ2 zyI$JMIk^etpa@#OkjCv&o@tin4W?9+%#zW?$KS4TlSe_xjg*!_3kd_A=9I=_#w?d$ zet(AY2;L4fXl5{WFC|fkT2aJbe}Db0G@78vL}xE`)gYD6ICt_oaK92TjLHx8$S)Io zwK99qpVNkGgZq6qz`*y)aqT^QktBQy193&frKu#^+ek_em>j_9o1S)i-v9hwGPGOD4PUer(;quYQ!HZnb$QZ}>`TOXwtJ{+*>`~xpNd~>fMmOP z!2rC#F}xp>?@VwjxD(v2uvejqQlKTtW84mhT6xKFd6n8~Db&E+)M+1o~EM0xHVCw~uBvx!kZ+8L@l5?%tih$f&v#ko5Yn%NbtX;Tj` z6+YOdCSqJSE+vNA&1f6aNLGo@Bx%HfY6;K8XB+x)k;NMk*m|rYMNgLYt|2bKKm|5W zVEuv}#(&}D(v_25h(%bd7x!+rsn_1GG{;@x7Xpq9j{aOXfwrM2K`8mjgMR~rT1Qx7 zf|`cKrOSA78q`p10vLKi>Rj|pxWVzL&;W$%Y1|Ul9BGmg)p!9qxX@{=_s%YC^XeG~ zVU!i=zCd2QsC&=g$VEH`SaR_CBBQEHAe@O6m}kn)ugT&0@cejid=WFQ#9ChkkI5~?)p;_#! zDA0a7u)WZdW;-w%Tp3)T7=}Kzt()MKAEy4uGQyH$D1UqewoZDY)+vFu z{$*X=fTTpF+$%kql zyCayic;pd0%119hx68I^G7DgSnv;PVjr>@nVHz=$Lq=%0&@SINMYZ=SBE`1IPZ}Ml zQR>K+&Ct`Rs=@Z56n|Ebm3Oa(0kAew;8L%V687lnXkJ7Vd{sDqB|UDEPr3dw)}}TV zN_a#649#{(@wE1IuF}cKkJeG>E3PK+(38q=i0td z>+4uV4ucRq+Y9^R5b~V}T0yT=Ognd&rluH9A6WLbkzC{H-8b2RPCr_9paQ$GQ`k# zV4dc;n8d@bfki*pKI-!cnbv(?I~QAfUHHP@n}6cS%!yeFKgyCN=QzMqG-;t|70qeL zHAyN$s!Ak4t0&85O08)|_Iy(V7ty=n?uOxu2s8X~z-YqFk|;=f#V1B4rhdBClA^@z zdeWDM2s5w4f%%y_oS}`Gt(#|wPqocPQR~F)*-L@B8hn;66fb-fS?607gUxbxxv6i@ z27eq=adcd&fB*p=zZ3v-W7xdS(pZQ^t~~JTm4)fo+N;89+|t}{GBzOv+)Gz#s>jl` z&cv`qw6*>8J+W?E)Smm2jrNHv*cZRpSqi9k^P7=3`p?m2<}u@Rlm&YEeVvARQmvq!u)lc z8)c+>-c0gCGN1+Ul!0^>N*k4s-g3xbe8KF5cA&ivPAMz9L$w{OA7BndpEvzciQE^^n6fvtb$I!E^lw&KP0Q|c3BO9ySjpQM>ep4Q^(Dk1n15Z(83+Kd zfElXgEi;Wq909iN$p}ZD67^23_6-jbN+D(c_r&iN$UC|nhf+7rtqUfF8>7duzpWP( zc-Lh_!VraTdjjR?y}E3m<(^JW=)%ts_c*%KqR8e~X?CB$lO>a7=-|I7<7wMF78LZj z=8nB1X1CCIA5Fd~vD8R$aeo8fk2^i~E8?r!=W`@>Ke@s5Qgr#7r+x}9Z6WydrkiYK>k-I$ z&1rf9SS6JRK=3^(WENvsQeY;!k6lQ{0?qqQkvtyy+|SNU@WgPN6o1w%yCwK_KZO~& zH5y6AVWKRcvaVCEt|KnKO-jsb!M-ONpii45i7^x%#Wylmqgwme-I%EtcqJDQv3B}n z@X3N&2ijD8FfxS1JCe|KH05Gm6`MHVE~fTZai-;gJ6i@5HCGBROQ;uZG8vC~ z*=$R;o(I_#;W&QNCOylWS4t~{k%Kd$0QOu@;*}?fCoy*ydg(#2i7OgdyjF;`;|7){ z-{$x_j^q3*3^g-NqnRt5en-qOg)jZ!_ z_1=`cV|Rh&K!5N1^00$P8i{%B9~sUJ_nvVJcPKn9N1rNS-4tS?)YEq`13_1ov`g_o zabVLgq4KrvQI8E)iBv@E&wVwU=$Y(CXQwkP% zrIzQ2_^gnwn<(>P9jl+wL0~hN#~BR{P?*6{H6Rn+meeQTgKv2e*F|fwZPT@Q58?Bl zB9)C4WPffcl?9VRX1Chwae?_?&A7o{4CER3+%%eOo0Cg@BKhI3-DMbHQ|eAb?_E?D z^Hk~g@o970ir>Br0{+e@uSY;1GkAx?fmhZTD>{%z1yUQFpogiNdW+TzgK-7_KAWXI zuVdh=R!GD4S9}94a+kF{n|)i~ z)1RWWB$dU4r9xF)+#KauJzwi6;NHz-Q(YsgpSFq;Z`1uCF#^{J<$_#{N!jKZyrV`X-(r)ipV-y|O1xFjmq{d~JJneZi~3NZ`iIQ3HAp=_~|@lmwdx zanqvLzutNeY4BF8PMb+3n$Ql$Qhy{~xDcvB_Z|5Qn=7laBgTZx`T0-FwAwPukVo~d z1i4oWbTm5KFo(p7Uu_WK99A!Vi=;JD()xmn{xqDS*Jpy1`7pEwm2=*?T#Rnqo)~EQ zs^-JLyFEN6S+p=)l_7JZ{nQa`+RHg~6$Kz#JmU7gMJOZV$P z=)?PY?<+ha*NIP1fm;bBV z1<{_Mvy2d9(OMPHpzihqLVvNAg3WXB8_c=3%%|M;{lZwp;%3?gj4#WyiXQi7cZ-?q z)7tAxXGWM)CE1r0pw0allT3R4$oC;L$~t9r@i#$e|QqHk~V58R9knV+y0;r5WUkg+C0NxD4UW7-!o2{!V=}Vep;4+MI;8B}tnzL4RHSx0m{aLH%ESD{*LB zS=a{SA2JXU!bZo@Mv#nO9$5oW5uTWoe)*7~4{IcS-uEJ9FOfj|UbI$tU{9^ud`)cb ze5>-+*OA8ry5$=8s?BcjSB1r%slY6;V)x>~Au~%W*!l?3MnN^e5i@Q@=U7??JYtMC zC2g37_nmhlf`8tf4^pTdB_7F00gPs#QQxS9UOLMf$G@ftme{ZL zLV6NKMXMPI-oeHgbek2N4SmS7bbe_k1D}qwZz(^_Z4u~0HdN=L>9}j^IvjovzXpTa z`88uR+wv(K-b>Y|#N&hNf=MHhr*i)A$V8E)ynhZW5o$`JvW(o`rsv*?f*CVEP;=13d5mrjrgj#T zv486Y9grqA2ODr}urToX^p9eNNq~}yETVM}Q*hykC|eY3r9w@t&$I~^hzlVp@EUOM zc-49DuC%(mSbKRtgZ@$Fr6RJ0Y3ca6y$MuD-i$t=il^g8$bj)C!uR_0=3isl&d5(??HxUK zWu3jbc5(z5FR5f6b?dQdoa4#v@(n?_)8E6tjd^_s_GWW}YQ5(?L^M1Sm| zh9XPb32mw6zjc(dj^<{!z&{9UR zn%5RH@b#c>IC24Vy99EM$WAT1FlYC(<_vA4USDDE>a3JMms-D5C^M@+P$YdX6VW-7 z@;cSE(nYFgeg_8Andh~Dpk{@kaDToEn2(^F{Djr}^XbY)MW}&CB%7OX=gZIhIlP3V z37j5w>oOsm{d?3iQO!F05T0Cu#%tmo%08V`dl=RCP|IzOikM$DG2Q{bf~V@vo0tcD zq*DIfQR4)xaV0dBrSIpjOjrHCA>)6$Bt5XoY{Y9YRv3u_kUbGq=^liA4x9> zt^ru2`ZtI0ag@PC!v&+C+kci>dLpPBbS5umXQ4I3HcroR>e5l;mCr(RT^$0Xy+->m^hk^eRA2-nHt}S}qkFGm%9b-DFt3 zJ-0|ZSc-p*pG-T~Zht*A@JmBzRbRH;s83$`G`b$)Y6!emG<`i+fJcKpjvp{{SQ2f$ z-afhC&p*GIUCsf&AX)`EYjwG@G#t30`K>1VjSer;2utbijh~dLU2<5K8O5 z`-%$R@*!!38lS(RIkZkmE}YN#ilX$ahlcj*u>ZGO=Wq;_1TYX{C{Ir%} z2>&vnnxw&h_R2vCNql^%#)8rAY-su}&EO8o61n1nQGfe$-6{8k)eVmi9(!MIP)5B0 zg3j*_w_~z2l2|r)HA~H*cio8yoxSgs;5oj>`PH7^Kq=ZuMBx4C!@WhZbRTKJ=To7^ zlfa0`+|ZoKIhI3qjZk2t)5hRqhO6x$iF1X+RUda*Z_Up8XuvwX25xCJ39zTM`K1SW?(7uUw0 zJfrxFP11yy>pQrOhhXaMPpU6H)LgSWz{eOm&Z5hNg^6{H`z=sbuduGjbqZ}!s+6$TiBBY-FJ$e(rI->l{NUTq4 zx>_7a*7LBF%qiB_1fwlqwK|ho`l7;PGIgVbf|q65QX{a79upO@X@=`5 zXiEkhxk1i*X+pRDhNgS${fJQo&p$0he@m)Bag99S-sFNrc=ih_#%Eb{oes+fcoqA~ z;C5G*>kfA9!~LsCRrb1M@YcO6Rl3%{l7HWncKI146W(t~?t(-=5HW_7t9Ig@6y^A< zgZ0O>de;dz|E#V1PhyYU;2vz9Lv$`ou&!gU-cd(HDz(oF%U1Io!GWcV0+UIwjoaxJO+T||b+Y2_IN{f_X@ZNX zY->f#$nX#09mL5DNJ2{91>oFl*ZP1a>a629=bhtQU8qJ*IN{d@)#mLNfQ9>AB7J}( zy*@`^rR*UUC;ytq8&OivS3;2-MtAKhRW-M==0!<_l7RnpRnu03{^JzQd-J_tW?M3A z<7m`U?LmqXWqW{eT$FVG=Q(ba*~>Lrv3#ZJ$^?E>K_F#{X67AZdu)3}CS8@?Rgm`& zMifE!1=nF?@$On?G6=UXpgRZFvms&?{=Jm!*v&#gKU~D?;ViJ|Y{|&1;Xoeuv2ajy z)>fWd#l$(Qm-_KafbXkuZh8e72$G2DK*0zyra1i3Ejue>hQb z?r`#mnK$|dQyZfl30@wo`0$2!N1SiB+`^pyl2B#6Sa?oZ#{r8G0A)3^1m>R>z%e4y z{+}2ffwIM@yv4ab+94)p3}(pAwUij({hdy2_h?P3!8olZ1}MK}FTVczDpXI4-iehm zj|F{cE8UqDL?MSA4)M#UiEH^GGm1h~Hv(J1$}>;5#4QuDr3pm>i}>I5kXPWpvbiCV zg0?U%zfS1KD4v;80P|-(o5(}@>FEunG9JA4fFt^A5I)!8zZm8zL-HQ;ERVs7Wp^3j z2E`4@hYFs{1YOLAIM?pSVzGw+DCe8U;!Q1N{y(9}lJ|lDmCI04x};zqm0sN{Jz`{r zolgKr6S3^hImuu~Rt}8A(h#{sjmP@*riITB#7}L-jMlO!pcIL?Y>J}o_l|eT`fw6l zKMUN?pe&8dR~MV>6sS0D*Sq`mh8qM;r(e#V@_jDdQer%0?ac*1bL^G7a>}#6;l)E(DMXnXsBoJnsmhZgg~Ne`n-=V} zQL-{Rl0X)?m41!nzfz)QfFyh%zrh#ZGq3!-38w%8qqJga++;h8XUatch$~>%)mp?jlRf2O_Mfp{f0WNYN%`VGhP1Ic7s8mt1#m7O>+lZ zChyY`iaBB7_tdeLrLhLaqDj{w{$h{{*n1%ZThNA$xGaxpp@l&I<=U-AJrQBuJ3T)X z_w};_M9N`EnD1q(Y$0Mb>A`=&QewtG9xH~Sa7OdtaiP&7CHgn{6N}}zH#jgvI#uQH(^si zBZu~)W*?T3ZT^|(pN$S#R-emiCDCQHBJu~D65))6R^wn$kB8E5_s<4^R2xFJsbClU zPjs4cDhP;{S5v+hi*CxZG;%DumUkUL?NjC`AdH4i!ffO;g-$I-Sag7)z08^8h+QO? z%&PS}+Y(ay=|QF@o>H=q!1W?2f;SI<HZPRmv_Z!UTIDhBYHX)O(UUG6UrW)78>>1wQMqj%Ue#MVNoNj#b?he3 zDljkf$K2&p3&kL2&X0F(yO)Vw)(W(nV6cs}1tWzfh*c1gK>g29cS436kA~FrpKegj zWA{qF>IygwyF=YPW%lgVL-P{@HwdQzb>+01N=s_9-T}H9Ca`SPtKIv4BG{oq$U?$Y zc2nMJ*eHhUr0bwNjeQ$Bxd=XhTTK4~0V;((br9$FuWylYy2X(iu6(V1{rS0Q%juR5 zavdV_KKtRn1uaSez9j)(4my1tAtfo&#T#8op)XV^YF4YLfHKoDz{SrtAVgFWnWad z_poTw{Ay5*hszjs_YTxz%dRvr8~rG&L>txJRGC)aGa(`G02>IyTLglKcp|_b<_HA` zeUoS$ro9a<+nAD#zpOG#`G{u8o4dUXWdcFUI|+|AN`gwI$GAD>jxSS7fyGrv8-5R? z@@k=zP8dX9&l<#cT?y-ecqA_c(QKbO83s?GSf3X*Ac&5W1M-B{Db-oq-IDA(vHDmP z_6_*SqjUXOvz%xaCsW@r{t=b&Fz&_l)8jc8AP5{8=RYLnL<*DH0icGXC2vNC)HYIo zf!749jr{9~E{L|9+^M~*icPXaAqD$_`q?KC+75>8;+vorETiZEJ)$m`7Lt*Cy5gq_ z$&|7+5QG`5J9Srw60?9UQs-rACf-8P)+M{b{h6V$Xp{0)MKR}H)dU1}pU>;MS4$ zzxVRj)O3}r)`M4o=HG%bp3gah`*v(SQhVEilH}V-hp7gqNu^QfMhQ)sp6|>t=>9Hf z`2W~aBXbNLgiF0A*$h{PKm=M&T79s2X2fRX`0U?-5+WugnXgE}-7pF^4G0fVkC)Sd z2TcbNXjx9Tzvk@`Yvs(^Ar5Enr)IaCq?oTzfZ(afGol&5v56tbTR(|j&6fz95scmR z`LJKPe6<$Ve~x%BB_<(-(ujwoY}CZ83{DfcXvl|V`eX(=!xPKNW4^<()Ef@2burBY zYKd&T5GVchU1fF!pEE3rdM*jb+8Nd~u8si?U%AtF(=SX-ixy!g_%ko#z^5~r1&i%= zhFnOYUPn=Y*Z*22sA~((8~`35gexazb&OA?nwH_c?o*DvVii9Zh3Ht^jTXVmTZU4r zyaTOdV85$%SXB)_ZC-uzI#j)LJL?x1Y3)84E!*Bs&$u@4 z?X^SL+cYxfgc++1mx2SOs4-)RCY5>4NQLg{;>2*;K z%jp&XdG-*7W@V8ee^J7yTk)kRUg_(WcTY-gun2h8wxm%rw>Ks(%u0={4tOH4(ubRGZog)?*+1H3mIF z3u1o|_Yu3;nKJ(2p5y+|ggoJt?MS^-y7D+Lfk%~=8Trop%8c7$juETc?1c6dTb)}iOu(aTOK73muId>o z7m7-@S^YsJN)6N_qPzY-!e9KBGm;u6H4RZG5<|87=z*4oI>0|P^voA9(yy3|vw2m^ zSohR(%EF_!Di1O_Nf$DwgJ)Z2#{Tj+{oPxZ3!l(>MNu>&IbfyxA}uMzxL#r{586(p z1!kwd;+ZkW4qWGNXLM`woL$C%Q6==DCHicEY5D!8<|b6$7y7XTil^4aFF1=sH2b74~CiL^DyV1(p3W zZ9W$&qRT+>%4LsD;|!O~xQzl=!^?9JxjZwVg{LD>^tjV|{rv?rWDK(2e;=EQ%p{I> z5{q`J0zuw~n%uaCqoZ&?4|=EIuxU_4h~*Q0IAu&Mvh3lT5C;=uDI|bF|>9V&vhS`@=*uG%v1sDxq=c)UrEt(@K&rpX0Bua!rQv5e#(W+_;KN_^}H$6ZVnDUEH zm@V*_$04I`B1TRHTe<2c8Q4%6NMnG#1Q@3Ssd{^L;dv0;Cm!+!OYf>N*!eY>PzKYE zh59c*uIR)hN6IB!*+pxKQ9OTWen8|X$Q>6v{rgU-)a$-#Vx-%mxkgnE6wyroz2t-m z?@k~9Pe(~0n>#V}R3==^8D%>@R(RW(0MOJhY{23?YttqDCL}d9KI91GSC+s z#k2p&cD;%0HU9{X)AjNRy+e9L5IEOV+R_A&MnfByVk~{&S6i8U@~5O_Y5{ffrlu?Z zBOAPBZWU;48a>^AZlswB2S!zR>G-wX4YBX11bbovN4)Bp{&2&ZZP_gWrUP~D(G%+} zgbphEng)NjNw5)-dWs-b6(xaA)UH)V=9fL{x|cF9k#eL&siE199;4kCtuih>D*z6l zm7)!7&t3L#2F*pKnX}Oke}COFi2Z;Q;5daZ7n(!CZG4;#|DNn{cSc%sR?e|od$Qd$ zb_snhuE8KEo|iFL*4JH#>j3TlbJDI?y#r;F>c;`1>8ax<6IYxtAEj#ed25#CcLxvl zScpW}0sNq!LIPDNTb9q|cFPkOlG?O0N@r_t&l`~uL_YX5(j(Cfcgga@@cq0< zwt%>aUT`@=0TI{q`-N7W1<&o}l#n+Ow?+T?FT~%1=R*twi$b&ZPh#fGV?-f9<)cIq zGM~OL9-Qk~hGrj~OgnsdP_+6j!d<7X)k{$ODPCFPlIw4z_Tf9RuCD zJNnNOfBo&+%*3Al#nPy!3HP05CV`CA(w-f{uWq!mq}3fMUNrO|T~257`)Z~;y+3s+ zjbSvmAcBVqGP@E`8^hI~re+7gJSAEQHBP|c!$0vgUlxvzTo)=GA~+G*f4P>L^a`=E zU@2oC&;;1!I2YE04G==6xg9LD7oamM73aYrhA%aTHc&MMjFc{iwb5GW(mywVU4+yb zwBK*G*@GlcvXYDO4h7*X42nmSQ{s3fls1^DBEI*9(klKEGM)X26bu)DOo-88_tYg6 z@w1WgW8qskI*ykQ1a~Mp^97L_wUv(`&1usApU!(nH@*N!hnAD~2X0Ck?%a`~ zW~S)?pCx#U}ok054w(bU?t+v z(YFxcp99xxukP@ z{fC~u^|d>iNIeGd*U$_5ev!{v0x>#3I+ug`F-S=w0x?lp`CytS(LVZiL|w{aA&~eo zf)Ck8*pY0t>FnW-LN6ZSpOa^9dSHJ@93FifKq4JsIpK)kEY7Wa)O3h`rm9Tv*OHH5)mhA1;D&TEu4wB#l0IShvxdOVc* zO$($Djkwz@2D-=@YdJ;nWjyMT61_k|nZF$Y_@A9=WH8WADKv2{LI%~1istNQLM$xZ z4;NOWqz5kGkak2v&Y``!Bk^O_rBud!9?k~;>%H`9+) z7A0MXx;VDq*QdO^JSR_Hs8RmYYB5Wi|K%cX(~b(bGyACGgS=Iopk&hy+H++!a7&5~ zBfG#Nwo9&fMkj{xyPn;hEx?CI$sFvpA@~kV2sT$>_6L8_cJNl)G90;jr2L&{98hU& zqT86m|HzS(ips349D+KqTg$H_+>v=M&MxFE-M^Ul#FzBG)?_?=Lo<18U9nSC_Fwyg z%})Z1;~N!b-tS{Jm?V!ik|Hy7^u4e0qaQAL9}~I>iNldP`Q^7XlJCuubRt&J!WOS5 z0D((B?QoEj2x`Tc!8e<6PY{z)Vq?oP{z9cC1+29e2InO>_yKWVoj=&5Y@3vXf}sKz z*+BCu_#IPpF;ekcnq9s1N*NIeVN&Dk-AoJ+cK|Xnjbx} zxIu!<)Yw$-TcDWq4fOpm1IoOZvwC*NGo~*<(slP#GI`RT>KG%MsEO(R-a@>Fbe>gi z=mGjS`pEMLXYf~rQF!Jv_uo9Od1-F{fVZUzN;@D1#^zub;hF~BEtxQ|_q{+@C zvsm3+y4g$!dArHOzRV)n*$QEydiC<&YuY9*?WEa9iid9o;ea`x#v z8^&laz?E>Eq?-#!?IY^pM3C1N*{x~GY)S2G-IJ%K)JAaahko&x1D4!v`2uan6@J$J zP&j1okm_KtIw`x43W-!fi5eJWgU%FVR1q!6I{&QLT~ zJduBkzUk4E7{?tU9ny8fu zy=$G-`B~E^sg&$MBKrbV6b096foNT@T%JM)XocG5iI1~xwS~0G)@A?&P|l>p)b;K) zBFW!{r zfC|Zyj${X0vkWI7UIM@nn)3H=EHwVAw4He2wALhU1+%rvJo$JbE}FW`P(zI#Xn(OFq==_uyL^I2*dwPCv;DpGW@sm=Qo74tUn5^;oY)kBmjy zx@i=^%;J8ZccPO>^mwBQE=^afzdA3Wk8Tcno(H=qvVcx0c z;a-Y;uj?wP$d}Q(rY5ldHD@kYnu9pV!184B0O6!S2Lv!z1wN5RhRVnyNT{H%w*$Wo zE=#|Re2g^eR-KOD3JaC1gW#P`3Ys(}!p3AQVaXAnPOx7PWEI?_u4hbz%o z@TNEGA1kQ11asd_pZ40Kib)83rD!p~^QKDgj2naMDoYT(vg_?yxYD{3oOU#rbD{^V zt>Yg7;|mzjjXk|MhN`@Ou`y8}G}C7c#^hi`4;-L^ZNKG&aWp_F5pc`>OGBYeeVA7q zw%fn^Qng5DLuaTzK|f*PNERcMVKHN3@>EooSvREqF5DC zjJQ{7YK;2;QT|xbPg;2iWZ-077YHxiHtJ0pVGs{x8q46SA-3C4;KBuvDTl93v%^*57mm&=V2$W&@-1n!q`2AUZZEN5Y{ z5Ul4+MHkI&wgmP2RDrw}kbw0mFIyYj<_K`~&5zM43mvY~;@hO{vVBoDnB)AWHIh^W zU#q69Ax$2&X}qiPt;gl??4)#d_m44|4aMzaou^9bX1(t&(Y47T5>!`u zn{N0eIz+RiwEyz>@8@PA9(1{w$vy)eaxj~L4Cf9=->+!yQkjcr;?A)&*M%Hpj{*=1 zFd9=)>rk&nbiN|*J^?XYhFo|bGok4GXzwAD0yE&yUEy$-x-U!H@U3y#J|+gvk=j8A z-Bjzr+Y;X*=xuq?MFZR0*x%&+JJpGWQK56Hz#_SZkP+|OnuE(z7x2E$5rR5=IedV^ zjUjmsEia(^o7J3gPVg?H1Er>LGy}LEH0slh)MM|r3&lZS){aKBoC!bFt|sQ&%98jl ziUfMS(}I|HGPuR*fHITX7c5(^_gJ3yt%g1C#lgW~$B}-7`LsG>1onq|BVW|JdT;V%Sj)*>pEASK*PP=7LR|Ke-MK&KQlbK|h9(ICBShb} zKA;2l%y=N&w5)5e6~2Je{l?q^0~#uX4L0`+U7l9IyvNRX&^V+?E5NOQ!b3Crfm)>J z&QnGxXX$A$2^4!Lyq+&bOz=Uq+P+d@r2&cVK<*X=&`%j+G*;=sfRL+#fK&+Au!Ft4 zOf()vGM>ad-$IbA)+qFWpxU%+JF=Hv%;8MB`7!ob@Fzl-u_h>`e8^EIob>5K6!upY zThI=K>w~OmM4F2lI3R*=Z!{GhDNwg1X-gi?Sb5m3+K-2dMKJH&am2&xjbg>@?b9J9 zoNkqQ!d^(At2#viH`z-Nf)TsuHi*I3_VL^`_mXwi!s2z&fgJV7+nz$-+?sCKjbnLl zN^m1Mtw67H}2s=96PyemQ%I!iR#u zaBMo1hM_tDZ3IbV${88S`I(X@s82Iz8i+Hvo3tOJs2A35D4WBKZ-g08XC#e}!InlT zUM`j9nGm9r5p$ zbSg>c-&=O)VN)uG<&Ma|YZJXM?wNchkWbww)3r<1+$w%PRSTgD{g#X1>oW<6gp_35rLN1Zsh@WIM7kAZI9VE zpV!#YAEU&_YabIEGYKUU5=x5Gv=tP6(2;H0P$U(Es66!p=m*L}YJd1;fV&!z0fbt1ZOshMi}T5fh2d=w~t; z!HUM{mcZ;rjX=si6Mpg$OD>x)2xj{e1m&*$^V_)#xYc^Caj$;$o!yz;F~4ohJ%zf< z(3Z03pNnF@$c*IbCXKxV*VK(IWCffpKpEWu4J#Z8Hjh(N&tf563zQ-C{=?7`p+NT8WS3ionU zL3k{p88P;+0!eKiYxVt^RKbYgL8v2tVIVEq2kW7zz`&UT_H_1cg7Nc_$ofFw#rqt4 zex-=3E&>M)lZ%Ude0<1=ZvZ)o7FCTM;aq)?0nWhw`#BC(oOlp#>r8!U2ME908DuFy zeGB0G@0)f3ZUgT^bq+xN{$fJ=REi?Om@p8dAW)g0bZ{}iW>{PooNI5KLU?zlHb4uB z3%?4llpp3KU>6!pU@+pH#za*5U=!GT!9oMTW*;MoTzm?6K*IfgAOK=1MK)rB19)K~ z;#jc0+6iJH-)dYSsQiAxYfq7SD7YDw1=N7A^@62YwqTVN(SDb?n;QqTnWPJI-wr4` zJQ*%G{qo%;SJ8Z5r5^x>0uF(&=h8PQaeCQL3-_QSkS7sR8B`)@ zA3u---;(%+xtFXy+LNetx&_ zXVFud*nmR)4}fAF%)ufrZm+;tx6WVxG5OGeAhyAiIVj=)eZ7Bu&c+HcV&Q~5edGRE z_b9hWaGZo1)bHZ|ZkATZjDUWTn;Zk7uskdP#V4W&{ULyY`U03^I71yMqWATR!v9gQ z06KH84_5im=-&aHFn&qmlK2BoSj17KU?HJ?r|)@%Jc*L;5`O@!-x_1T5_h?pKa8=z z=E5I(Qb&xfcZ|Qk4x(GgQoDjcAm-f;U_Y68&@@8dlus02t!DCjOu=cN055n)q(Fow zrI|4RA?bh$As`@Za0;2LY|_5rt3~@84I3pOFtd;!SEnCo{-84+B3(2qt=LJt_I--x(JRR|=((;<-lf~ zRa2WU*@}L-JXF>06k5Vv?y#Fj_S~PM2-VY221CT=g-|`FF5P8Rfw`>1@BQEU;sBBg z_GgLBSO3M%@xA(Ij*5&`o9{wIL96B0dnLWZx}`0sw|!&-`~!&%ku6IyrL+BFF>}2s zpQA-b4o-6 z%5)fsSJ_yxx0l4=d`!_2KbZ%gKIGuRbO zVEh@6!4>9)gC8NBcIvBEI~aihyXy^Dm}nYUXMmGbLIc)LoVHbW>cLMsV~7sJ?WJ}h zdDni8QB+}Rn(xbdUZTDeP`l7f+zm ze+A@opZ(0lASoPRX@a)4-rPW)Udrl1NB**b+l&W;qx(yYmv) zbWfnR8m8jem*Z;ixHUVD>P|TI-uzFv5(21=JDJ1Udf})XftqdlG?E=&M-+g0+IcI9 z{x{+IaCli=U;?cFR4y_&`|WK6)WogAz`-NK_9zW-Bw=e zf{ok^oikDUW2A92Dk_mZex>vdoV7^(u-R13Cm7}+G<7pL^zkTSnM;w5b~0lP4_kk7UdGJExM6|l0gS7r)&pt#NHlYFHhU`7o;XL;Y_~2yS+<2%AE1R5WH@8b<)`f;*?Z;b> zo56sbsB2ngUe#Mo9rlU1%2gVqaX^YDvTJ8rV%IgbEJvK0T-*jC%kH%Y!aJz?R8Z~o z9!RDAbmvqcXy3~U%Sn+^zDvY$7Nx7ql%+A(pfi2fnFI7M%&xY%9ut6iwAlc*Ko*^5 z&(6AFR(-NRm+dV+?m*rwH+GzhHPd!Rwn6Og3r&er7T+VbJUq(01u^|{dz)&y-St?3 z#qSI~1_Zi9@Dsmbn@>n+$SW^n=G*$xMVtT;HR+!LZ}7#p zN{eq{Fv0B9T_+#;pewbeNe{GeqYm)T-rRvl$WG{28ukWuL}X3qZX1=zT45k0t0_yH z-;(HGBLza(xh!)lnu$NwIjHr1@u&*zOAO@_U*Jpwm+^u(B*6ero*tS|#YcjA=;*4& zHo++iCi2O$$-*pTt6#eMdW!-gY1KM$J4n_b->^7A-Gs0tJqEcZ_xxGM#jW}_-mJQr z%AlrJNbL@@2e^vxsO2GP-UX$MpBU z@K2<6JSY3=U+U(>psWC!P)-7en{tMVcypgu1f}mPA!bXa&XS0z*Sd&k6@6@(A*==s zm?bFvmUY0Zf_Hm97S%QoiG=!pt17vbFcc?hRHj|&Id*9q(?qGh4LVVBJH3^C9hAv4 z@Au7D+K`-oWdE5Ds3~4Oc%Pvk^|PY1c7{CD2wGy_k9*N1^@}%03U<+Bmk>Ad(-JN{ z+A!fqIotd9FzhBIX9=p8MtC9f*JoDx6|&0Ih42D=f%Kby9NUHwQ>0h1@y|dPF{T=(IKJ`wdX=YputHk=S zFJuQ4QPtG>uHr~fHnUk31X0Dx_rr`(6+e%I^ltRM6u(Pt;F5eI+4jtPr0+zhgFB;z#p#YBv+3n=uqh zXn|_3KVM(G#AOKWfTK$H6l|f1gx<8BM9=_C)6bW$qvsMKM^LwLk(hew!Wa+GMpl7K zDE@mbXxZCT9V$`;3+)gaR3@!u7mFdxozqVa#as`oS$}Wg#8Nj87mh)6kM_r-R^7ors16$|Uo!nFx#{2O-o}|}W4=zY3jRb4sp=j=ymzjxTVHCY5X|JqRTeWC z?+PNAuDKADb6`q!$3gOi_@2q#4qz%+ef#@eP5P|!XsP;m!L|_LF;rv$f@8zD#V|3g zGqTcoLr2AB@(3s-ZPgALiq>>(>0$uql02{o@_mymH0PX)_{WA*#sy|x$q&*^&FZNP zlv%J+w+c$Que!(whte~Ke$fgFZoiTQ5cP)pW=UWZoY&4LyVbX)>4uSd2zz-v2-Ws5 zyTHyUUYCKD$w#?63n(|J8!{2891|rN`TG@SeoXl0<(o%l|JeV*>cgg_*KSX2? ztGex)Ii~@u_ULL+>;olB*ZEN^ujn5c{$4HLT=)HibW>R*ipXI_cSQc!_P)nr0eq47 z$ebe5+COqGg@Sybz0ntQN7pN#3I*1FH9J~Rr{>Ij>Axns(f_Ukm3C<~%=u|(+&VjS zjtl2w&NVH$2&J#z>jSN)j1K`Bw+EfH@D{meA%>Sv5#kGyMY2y={d`sa)&l_ZMl4<~L1o_0|rqugf zgFZn8f!r`o1!*xb^`Mwsg)Ek8gF}%QIR=QplI?F_*sJlqO9i-r6`MlYmwx=Av+oRi z7l2UHVBB=KjdY_XqlM$KYQAecwXCEF#!*=;+TOTNis28J8Y`?pxD4rDsoJg&EgS&f zoq@ZaU1val|XiFk}v* z4e&E^U{1GVtwjZz&6`(tZGl1IpH>xLtknTgrGfpr^i|1bL=s>S~x9 z6V%cL){O9TZi`7eu>+Z8*lrstu2&j6Ct>GWes$SKTqW4Ij@JNNEd+USuypC3RnT`X zt@tLd7fY2l@@hF#M2upK&!{9!6a z=1^?(Yq>0PbG`uVA0l$rl17rw~ep#~&VRt5x zl!tlPD!O2JuDklU+lIx?RF1+;{ZXoYQbbfhA$Augv&(!Y?0Rug@7B;}ne2Q{Us(lcsQt>3M zg>iQC(3Ss9O3}SZeDbR%f%W|rIDuIFFfZS%C&N$UvMUXR$PDP-i(6M*UWNCPESLQb zSE$gP{;dFDxOcw`90B8c)J(|M_XoYGY$jClCepBQXt(&^UTY{IagcwJYVZX*MfLYr zYDCm3p*)NY$Ygr?-?vBzu=*OxgKj0=Q%U3w18c4SS(%Oanv8#)sBU+* zIN(t<{(ZayH-gbF;*!9ib|_EBCgkNH(WvnFZCuco&m9c1E`*3W-oEMnBXBvk_}s5k zuH?8~KHuS>Ed4&w58`vWUJdKgoOmcuGI<3Mbg@d>Uh&PKy`lHg?d%}0663V0_=|W~ zqMyiO(kpje)sO+h=6LXsEB#)>I4PQeLuj_|3?Bl$R1NM|!u8&E>s<%o=!J!$lUhk$ zneU!X1Z_q&S3;^sqK!CC6cnoko_?Jg(fs%Rj?-Xt$*j;Qv7+sdv>?vTBrjZPN;o5+ zR!ZKld9-58+95!1#-~`LH^G|6)*vbm%Ddx6s7QlklQiMj|rB`(^gFPtrns- z6CbtM512c35WaFHl%jH3p+J^=ABOQSb!h%=<;j&At5g32=?;p3TEQj76Kll$sQ39o z{(rzd3#30d*2DP{8w$uFS)PQReQ{C%Dx@Fl)f$emb|qLzBuppX*KO-kWiJNi6sB+D z0^?xuM=7p56@*dCDTqu{hIaC1X$iYFPgC2nK@KyPj1V%9`qkl9eImID6 zOf_X1Ro>y}1-3>(mOYrjo36Q?7E{}9aeWp#-@%OIK#}@)>vKY%%eNkCk$X`MJpnHIl7meQ)eJjo7nQer}^MjdW}VGc;z`A%ROBg=^p z%m#r8Dw*r#8VZI}G#a#ZictBhzi0`IUO-0V`zYm65FEEAm)R5ns%RW?3LqaaURk%- z5P(kbJDOhIs4aAf*4dmS#ZWG1Q{v|{!S^B#c6Y^)DDY8m!xuy6XAY@UF^a^O@Tmvw9!nJKd%bE3>J%5mVfn5;bdH-;=Pc(H zTmZ5FHe(O=?&nFkMx}|2p#4a-@WlO0|Akq0>;6D%3~?fb)j|QTT$G%*=TJv;65&fv zC%$J1W@nk;q+7TYga42qb1ibuDUE?1`Z^z1k%Ho~g##)e!wiq5L^v@8vw7w1`E7nb zB;<%PaJ!yC4^k`;Kdm{O^d0*^kR-^VKT)8fW=(=rS-M`};~xym!I9xeD}_s+FO&nw7)JN1?F?)p*aY{j!EHuCSR|2EEMelq2cYE(xUO8r31H#6HP|E!=X=Ud>B> ze8N6~aXqDDpVJ+maLvg2PSs^rYSqc&_~U!9DC~eFL!~_U*=x_btBHuz=yXf0tHTuc z9kCCXWG_hS)>iWKC_VYLfB!-+oF1!sJZ3Kk`f)T+`X-;AKfGPv?>`eC239o`{!%Ut zQ;`e~9c}pRTBO!9Bp#D@t@u42ztHcTyg))GD|(XQDPSKi7!i=&Mw30J2&6)JI03bA z^wR*82HvoB4|~<#RQuauv}g|nVPU0sqzG;n4my;}@wnU|f(jlv|FaDFvvA1(OSbniQ6u{p8Ul~ z*?IYQnys8|yD!_Z_GG&zvRRkL+{op6)()yo+)Fgp-9c0`Wf?<&=P?AoyTz{n*Czv@ zPe)lymnk0ACEVNS8(DFccjJk->r#EkLOxykS^48@E^|;^Q7yvxJ&CS`c;;D<R5p8 zqNc-D!q-OMP%a1O(SpXB^ZiFvGvoPluYjc6xPe-0Idsg=SGHf70{St)V`mEhnXOu0 zwlibS6t$Y87vZOv(s-x`HOeoBLDK?2%W~f+W?NVCr~|CJQ;QS!_@^iT-peGOBnP+l z&=|SU(FoL(y=_Y#GXJY1Ph1!@RB#D|JDP#~zS^{2bHpxT?!1aoUFW*JRrtO!&J*bD zQf9mmh)a9*SE+YO*-I5(_4lW{I}>xpL~o_dfV#x{FXZ^&TV!rd*8eZJF|)C;rUA!5qJuNDbFi_cIdVZz z0%BUgm9jP$%3?x=9NZz`XZCJTwt{%!Zf?NR_HO?7C^*355UrWsI`<^`_Wag%w3b+% z_-?4bYHeHVIT?A9^9YmTg2t!EG?U_!^2fsx(G`wC1A%0m2MAGcIXRiKO3Kdl7C9Wz z5P3tg;#$`L&O)w*KVolLkbOtpSpnF%*8W9FL3a1q1?^fI!aw zF|@Jg0*L54n;8LQ=mC;8)*uHsa$y@=H+wS^Q^)sJ{^t=uX-EZNR$mlzof~?+GG}N~QDA^dAfgIibD+DEvsiUJU7XyQfiwnKJl>@zvy$K%`Ex^Tp z%+VB}2yy_~JA;e>f0YZ6)wcruT^KzaIY8Oe%;8^RB^zT$7kzsW;9X#8W(cx&cz1BJ zHUilL-q!{wNk{|aY(dulGM4_A0WIL)TLWOEXZ%mNe|!HG$jth0XMICM8!KCVYd14% z6M(UqB?urVCQa|?>PQRFw>J9AP~Xyj!RFmx-&xlzn1{=GBXCf|G;@T=sSY|j`mILO zkcpW!+@I+0VvzB_`2FPV&0GPR!1u;u1OWf~{O3vgy|;{PtS#OCnE!pa43Z+MGHPNp ze;528m5`8)E5L(}nG-4S^wScTZ!}l-7|K~RU+vWeSyZ?&ve`)alosXE4 zrRCpR%D>eAAFaNXnWfvm_3!=ac7u` zptXtRe>Ty~LCnk*WF&9qXlVMcR{59w>t6$AX=V+Qw{bB0>p}t0F#>`AOZPrih8FLa zhQoV5{;dMN56yosDQazKWAxW_F|o1%^zH5S-QeDj_?=h*9*pl}XasWoyMqA?^wu_x z?=FD%^1J}XHui9TohKWAD}X`pFVVk<4ZtAuhu8rO!heVZz##I6H~|cz{~>lp0K=C* z#PZ&Vf9O4?^dEYUDf1uV00J1~{}2;^LGceU0~nP4(0fefKlGmU*FW^Wg32FyUqSUh z#Q9G8f9SmigFp0MgW-RO^<8IZWBGm%|07{x`AcGD^~e9O6Jjub`Uk$(4f+SZ7hwDk ze6PUdANU^5^iS~j5=`A}O+nWGuy~i5{R7`Kvit|Wr)%{G8Q)X+hwuB1!eH|cd{4#p z&tmU8Wcxm9Hvh<-@jV^;f8hJt4u3Y|eJux&mDzu92qWWr9LIm)`{p|R1K;y?{s+Ft zaQTBw?+Lp7jsMkuz=lrt_V3sC-xth#cmEIn{a*nBxq=MgRu*jxxdY5=1KJ;|1@T^q!(Bf=z5Ri*oF+5Zwu80ke`-xa026IdmB@i=X%hY4pN zBx7jcyv`z+3p_NonIaz-+&9XVnH;)t$Z#~U2_WxK$7d)!gpLv=v6eiF3)TmhU0Zkx zh_Vo9lh)~fkm-+@(yGi9zgsu;)i{r@bTaN`CjbII;|-$i#nqHFq3ha{$YE5q>3&CZ ztYNkQ5gc_I3vIOjQt2@^Pr_g2oUMBUEU**D)gWYsx=d`SadpEp=Vawey8N8TKaWY4 zkisZBP z?{p?PUl7hOHnGZf`Y4%!Wfuy@2bq5*q?0#ejQE*{sWZkPW-k_Bsp5k>W(E`=LyN`M zfquu2kxuyv67#EG-`I6*$89f7)KyZ9Bu(3cn0;L&-A^#*SGi~c$LcWWE9k$7M_Qg7 zrcow;iJZLGm!KFzIAlOX!b~(CF^jI)NMzLYXPyO5>p>P&8BNt9zt4t7;F*PI!{Wof zCU&6Ve-LpQd%Qf3V~Flcm$N^f%pYyH8*2!tah zkyK#+m2Obd`wj2A>z&CtOOTY>MFm2XN6(5+dEI?nY63#Q0sAr8YyJLJvYdd-Q)wQ* zR@&GHN<3-wP^cdzk*Ocu$+MsqaHV7=pX2oU%kpQv7FF4rbNjQg1wTf};~}%j9*3uY zk%Y19wY9IuIW5S-Qe-37-y}v(d>bwg42R0bVx9~B&)zMJ;zzjgP(;stM{`9L02yz#fp2C#VK0f5F5(XzJj=GJhDyVOuONuJCB93Wmg10PWp~E@zX}i(i&e=%jWLu5Mjh}V}MCO zeBNB3IWHeRV~tO|gcav+V!bB~ikz=GC>ToW=?Y^uAe8;#TG*woBY5DKLO#GtO&UMf z8U58k9^hEkXfQ`R5u@wjYbv z*&`{en1V^QQL!o&X7q4u2C$m@15#7T+e%#?pr|K!8Tm~P9$3|Be zfDP@|VIyeyA~e-VC5&yHK4=-*ZPLs;op%ft{!-WPQ(n*tH)KnHiYA&JcpBN*8kJ~6 zZf5qH(4owM|FPGZmYl+aZJ9)ZMBa(-t0nwwp>>+q&&=bjA49vPc_%PSPA4Ca%%OAE z-7=BHb#(kP)1vvmkwkC^*7OB|M}~K9O1+|*>eg4Q&ac}YO!Tf(s+OFXrGhu(vSdL! zMCopJW|WbXGn|Zn7|`y{tYemb*gbMj%{<*d=v49RHOaGVc)(fX&~1VKQfzKMlMTVF$F*{z@jJ*{9!}>eLJ`UtBU8>PTEg#tHc~Sz)8F{J)Id7u^FC3S zg#DsFyZJ)n8$Vc?Da>Jy@nuTC6{n{WKA*#dj@%SOc~TxjCNc4C;}KIw`1;~~#tkDs z3d@o8g~vi%WXFe?&$Z=4LAX}6N9$sQ=@=z8xsp<897*ElL+6P-3T6!j&lo(FA%52Z@u zblYJz(N-bts&jo;UjPv?&HSQT`Hd6ztgY_&iHw;l)^NA zgL~RKZh_yfc581zTbpHmq@Y^fWnCtb`)Zr&Ibv0pBU$^N{;8otyzrV452wx)taby& z%X&1-Ce|j$g4gWD$s1JJeSU{sezTK8Sk`R#iamq5^wd>*IGteiW(B`Mz+|f3lNo1Y1)q&sEee-iBscBGF}Gg_Y0VX zvX3&QcXQ+V0Kw_@G%gr}(R{l~?v>0J%b@GNF6SX4m5|`Dy{MP^;vEExxWkE` ztfuI5b5Pl2>d337=SthYUzQ;J0S-!kPu{p9Xa$uF{-lyk>)|p3uIV%b6g&8(M#|sl zb7D+01|u!<-ErYJT>4vRvf(BZ;1D@sTyR-aY}YKCTqC);zUAn1`a99X4whRH!~FRQ9^*cCq0| zWit-y3SfwP=(+kHB1gI2RKw|-)S74wNHjTyUJ#|#M1BupwdyC&0^#3#0?)hgp%{Fa zmCf8;&Y=AF&Y|tRMiUEU?+zY+i(a?uUcoc*B?LiV?h$xDN2&=!K+hmfaukQ3+JtNK zFFgB{hJ*H9l2RzGsQEPeAHHxN)mLBi7}6WCi$bpiLU>hlR-h|gQ;OyKR!3G*DlLR43HIHhn=kB}lRnqN>e84?$$dAdXe?KbbF_eBIK2vw`P_;e4zN zEi&~u6pOIn6heWMW~{tFt#GrURKWfcLk3J)s@7(%AU%Pgm6I&w@buofAPN6DbA=6du)c+2EGzU>2xmpmcp6B31uWji)7cJ%&pH$oG+k2>G)NtHHPn={{` za^qOrwrnh^jT1=u@otpKW1Y&( zo-u-hA<-yZePCFfQP70v>*G@VxaslNSFhV$`{qVS9Xc5cny|?YFoWGXTW`y?7#CK0 z#I71nePrd1k5K}PRQY9ngs9oXoyfXIpE+R{YT2>%5wv{R-$;3XxkN*Mas(4s)V%Qx z*|muKJ2_M?`Fv*P?j1Uhj)@)1R6LM%q_mFagF_-IMT8%P!<9pNYt=&)O!{tv;%YGR zQxxHLHZ@I-QZjO*`L@<^1m;UZz>ulFp#h!Q%_+D`y%OhCqjoAKKWS43^eO2?`Q-x` z`@~uhn#h7FEiGw(Ixcu(^G^B5(wQ3X+BLR7xM5s4QDxxdr#KdVJHEZ|#Ao~LX`h7^ zr{HV3f@}%3C~xe)sMJV6jW(DvPxu#AAu%TC`mbQ;JirAnbj=0a4l9H@7&y1VIpT%P z^tJbs@wr{_p3?pBJPa!r=^|Pe7b<&g9VcuWV_@jr+t~MigEpD5T6~1!SL(+Sdz8aN z7VWWSpBvab`??hw_BHOrcx5AZsa+HiY1;Lr!m6OHuQO0!$n zr)QFSf%gFd;zyP9R>Bt3OYGUu(e_>8lH&=%#32ag;QPn+Ej7<5Cv)FO0PAEjb(2S2 z(v_oY(=d{M0mfp_tR+Moh00<9fOtTm@HqU+z4iu^kC-aORWm(rbAy#)xBHbqi&Own zzsMvS@|g^kb&|{!#(#=?k;v-XSJx6M&y<0E#X?(SSZZ~#u>^56v~CAPTsnU^B5J?G z&H-N9crq4f_tSkuivRxHJOiV}%lslUzVo-*Rgb??_hI+DFjL=zhQ9D))^d>4 zllT-n&T@)AC}C3#qs|Tp?YVN<=IP%N{PUJ_m(D`DE+yWd`_3~ zsY@Aua;K7w6VlCAEUfvupKlxaQfuq8uq&juv}_yI?7&Zz1L&yD_`ABFgmS1 zCwF!->z1pq?^qy=cvL%}%<64*po3VSL8f+pPl&EqbS*CM<>OWB`nQ+2d>`yrgWk|e zSBr`C)2^vQ5YZLUX^f7#X%h*p%OU;FxidB zL}ikMNMhRF<89TUwwsMd^-3?fa-K4OQhqWI@Nm_U7x;L$^mBezZNo%~8vk5eE%Dx< z7eogE`q+GmUH3u3>e!3Q$`rN@t^OFTwZp8D(SCaa z+&UJ5h$>wn^QTWJ%5r^Icim*?sS88A$4T2we$4ZB4MvCrGxrorq{+^$18vRp!G0)q zMo&u5+PRT}1N=d|MF_!Jpso^)zuR~4*y_T-D?X29yh|dH0(2>6$N@Tke0Q|Z!hzA} zA+#|D#P}tM;>POIf-~MPcK(NpSiti&dgIhcChI#x8`|%cCL63(_9LW0&A&X41FWim zteY@n`?EHZx3>M_;a9@-O*T#*pUw$T^Izo|y)YMH&9Ety5iELgi7zcQK<(qB)@EVA z$8R?rf%;Zn~qp@CS<2(@;~<{MsB~yppO#BQT{!s=iDqQBw)X za)h!{2@kH(reV9>Jv7Xwvha4yvysTSJ4R+|uL(+0X-mhYf3Wunr0;*>?*0l?l)8k|nH#p6Lj^zDC&F>Y zOoHyNo9QU7{4dazMpcJ);KjGHWRfPO(7&yv9riqg?9@lK&2bVMj9=A-R_#wC!_EWl zuFX2J$8K0_ZI!}*Ka}+d*VSg*al!dPa&o$Y3_oZq&Uv^C*FXlh^1I2kDlO6<&}=J{ z>OD(R*ZWZySIXH>EOC8eq^oWJqK`}>re@F%!<0NACq}~q*d>sUvPB@&v$CSkxQd&2-~IC zMb3#@cIW}o0?gGgEF}G-O3Wk2)1u)h)s@*bb!@6w_PE;fUzTOt*NZ6k1 zs``ciWqr)+hH3*IBqoOkVT_gV*X1VeCI!b=yxk1V$}o+sQdAW4 z!PI(Bbw~vkqTvH!9t;i8)ba(E&EEcR5ns-KK$lZQlH)RqOKyqnfye0;h7}7oPu9AH zJVioK0xE0=v|#y{Dj2RKgC zf;gy>SZyEl+-}>;$?)IQkm*#j-GDNwDx2c8iOQKoAm`e1u(3vF zLUgkt@o;|32Ouqe78Uqn&D~aF!T-R3mMgZD<2e2xW7eT5zXs{% zPY&~AUO2{8bCr3FN-ti4{QIVZ-Uma2H8L@sr;5!){tj5N;3jK2SZ%@+U#@WDA1g~i`ya_jkv4q>eJgps}qTl?m}_4>Qr&K z$wU?2A#kWnd4Wp(4IR8p5sH$3)~w=WFjWOrnbshEjA!4JKBeaJ8=1-^H&4lzcI!bFum)wGBXp=sFE1?yo-IGBy zZ+RQVjL_L~Jhb&$v#@?#u?zY%F!tPNki036xoKRtEK@4xc0GYu{?zJcb+Mm5CA*g3 zxB{+V46HpT!DEULtnqQjYd-0A*sJasr;d6VEhyg%Kr^Puh5Fofjpg^l*rt;WHI+>; zjii#8B%+yA4#v!-))%vXxeKv!+my^z+?N9ECg+ME^%eM0X_h-VHQG@$>{SdK@|;*+ zEG8Nw!aM3kof=0VrhTa{Nu^BsVOpz5a%lCY)qPgQB^GDJNk(G7A&_!XLL>aa|aYmV)~H6-^4_ zDD%svtW^@k2z45GTS18P*(CE?Ozk|4N|nM?SDrx zP}vhvft)fmI?M5OlO^f$Oaq?p0U^>E@nuYe|6i`v8rIW*-=EZiu zzvd?j>ZB7p5j{=&(YwRvwnt}D<5ExSLz!%k)hCwKVlhE~v&=CXWMuusa4u4eOPfx& zE8=q{u-|v1=DFR+gn1=c|UP7$)*g8Q-!=thVV}$!hz5**wXK$5n9u+tZ zhOwpGFC;I2t+;fsmlh_GJ8-bkvT(Lm2j-M{yG}KdD5xL6Mi0!;=8B`6#(9g5X{j3S zz$g{lkjl${7cr*ey(I#H0|Yt8QyR~9z|We;13-<(1s6N^lox8H2F2(YMPo(Z+( z=R)@o5R#7Dvgi3K)8H$+UZhnleh-RTDu`KUT%mXSF&;b@5az%3!Z^5V@%EXw$9dL( z>ZsIzj_86XDhgxCPID?nz z9Yo8=IvmwoY{h<>E;iH(v2%*z1&m3YvNG`1DPK~ zNJ4OP-RbD6aMyoy;Ac>Y=0Q-JaZxS!dN!G3KATA}-Nem~gc!L#rOd*FR$!7K`R29Z z_M~B*EN3#dY&!lBNVP_cxf4*k>Q)^W-DAInFI8(yaE0Swq-A9~2AJE&T@6=k#OXVK zXt6WDh(Gp-UhgqaCA-yR0UX`b^uKhln`Hx&`n8`8%U_esw zdN_Sn6lS1*S@C`eb;z_PX%PKl)~(P%uzp+K8NOXT+pq*3Y$7?(9&5Wd{#LNiN{ZnF7@#{CJP29SAba-v#zWQ~^?J+;7>QSJv%X0)W+4T6&`)Usbo-j)_ z2*uE)L{qs{$EZh&Si?cy0*@$nT#|$Pm}%ugm^=#N=i_=CI7g2hbh!D}kNrKmC$A;A z)p*B&Ye2E_>DKGRuUu0%D9mJkQ5$$r7L4QpsA^+5w7Igi>0*$|dNQjnIq3s{bJlkJ z083@`Y`Fem&8McC8nqd(g7)&hRquTt-Eq3|N`Rj&cmj8)6a#qow32M_(6~aS3~~7O zKKf2&IB#k;sqMF>{9}T$ZI}5Fli(;vbjWEsJ32=bbhV7I_+*GgL5YQbPb4OIS`;BC9meJt5DK%ruH31QajoOshrTsh7oKIR`F=j!xXEZg;#-dPg(kDOLPwx+%0+fjsU=U1 zTTPp$auU~?bY8P8cC8wJ7qK9%m5CZgR1$N6Z;^0qg`r(*%}x>x4A; z=lk(nFj~-#A3Uw|&J;Llf6<{0KWybQ01ah7xdgxqJr6`Xx%8!fifhR!KH2d^-Om`s z?i2h-dGceSt{0Y=nOMd!u^;Q&*#1_v6aLsVYQ#_YJ7@Fvno*s3lIB^ab&3A0OBApe zH#>{(bH&ixtZBc7r;B66p;(QOWuE0%!!h1}t5*21`?FN;?8z1uUoT{J zGkLEf9~i?9-SDKnC6@Gw5h2n8G(y;XYhp$g{P&kSaJ^=JSYsNh(+vYw>dw22pYz$5 zi>D(Ks`lyQXWtW+ksK1R5ma?2=Riu?y1YuPmVxE=c0{EZ)3J-Rd{}vq&+2@+*x2P~HqnvETEJsr1~Rj3p&{p}6^PFSX_f+p}$=qtCn4{>UR7KQw>S8A%buhNf3g){LjtqVe{=xinGFC6qz(W8401{JGG@-kRex z4PUT|>YU43L`#qZ4N648LfRf?H&I3Delx&8{QNV^jE?$?zGH=)aVIv1 z*>N<$tq zY0=)AJpS3!iW=KSz63`tAUtU{ogy$MA%ZAx!kJcakguqeu7Rx2L&ybjjZ&vL;gJi_ zc7i9l&dD=yvU8Trb{eaDo}7d;3Z|cVAz(Je2tLH~>C4K%UKwpK;NCpomG4MDlZIq} z$XP@Yy*MKL$?Se)jTw^^PqOjOSEM#3i}mK_u(-zhjDS4aKB^YkWKSiZAoI&Adhq-R z?Oh5gmdA-bK=AzJI@(pBa1=I$NAS1@E6Jb;oR?!?-8&2`CJ8zNw z?hxb0?a!c(9lVW(mWYCwE)Z)fDAPqdY#3UqL8CI~?)#$hEwIH|gbJA>BBGWM7YE#o9jt^`_N5BB7uv(6K=tyG z>tm<)5bIhb?Zf;?tTvUQb)T?GrN{lbdALBfBo0X?o+mg2AR*InkvQzzg9PrFaI|}< zy^wXZZHAPv?2{^e2xOqGQ^yy7r;s2s6L>Kku3DY|oR4|NHe@U#a$5l z31%o}rVpL?j%X`f%j6Em`|TGOi|bG-+mQqgK0t#h(ph-@s8TJeQHp0H@9cfcy!99>n!pxLK9();|Hu{wU0qonFA)CD@>QZ-^MH;mP4a|Ux$}p6eMeIa@oFh z`Q-T-$%YkJ$*W*4_iMS>9`Hh0s(d*dFBM8CA6-@>U5EU_!-91CrI*#|X2LH_d~ltxSx=iYFDFJNXk_}Y6toJpyQv!$XAEP1w|y`Ez;G9EKkkq zFsl$<=1Vf>E9Jp|)VFI^>?KG71}9fEIce+ls1TGj-$?b_7+s)G-{b|>qfCG)0vE4D z;_FHj3kyJO14CNa1}^Vs`1quWPFLbntWi5KndP6=A-Z_;J9#tZO|ee0W`UUNkFz#65D{FdkMU`XnG2e7yqe{)Ke+ zqD0(KjYGRKSGA@bem`xi|CQLuC6hsY{ssoIp42u@0TC|G_-MM`4tUc(!(x<%@Z76p z{&5m4qz2LNpPvo{@i}}KI!3FDQXshY)iFs-(=SoZk4RbCc<@ajxs(T|dZEBB&Jgs} zg8W{8$iJ9xWzS~X>v5hN7F=q2HV|tDS!PI=Xb8>WLt}q7WkAGnp8D2 z-0VZS5>aOerY({hpTF}pU5W)GKr|2U7SFO<^x^KBvb2KZx6T1CUsZVJ}T3TfgHiH7r++OU4hSC1KT z4UscKnfMJ&Jn4|xh1dF~TSExiTxB5}`*oo{L*>8@T9C0ty!waodR|3`Bve>^`Am z-D;&V*97){YRv`v7Bq#l?07 z_?C7iW|CzUSr|Am_Ol5iY~FI3|0fgZe%}KGA=R!<8$`-2LUlEfug^vLKE7-&uyx3x z?lXAL<9Z>VI8Rx##*S6J;`vDTe;>Fm(NU? zDH3$bgHO*^O-c_|KGxc5*>Xr8!S-p)7sQ5#YodFI1ZHpa#LTcdB_~cJsSX_OWyFPr zAzI&yPmd`IUD=)%y*FEb5$4o)=a@`F4FLFzPAe*2Q(bJPQf{-_-#4iQ7ZkJgEXNkZ z+8n;*4o3#ui)atuYlKcdkN$E*gi~R!QzmoCFJhxXeujJ*eDe@GsC) z5`lm@X4@Q!Pio3bYn)=#Bf|Ry)XM^rmm1TA+rL_q?*AHQ zPV@9~i1!e0kN71p|44(UdX zkf%npRMf$Gqfx5Xub`~NpUFs0KH!aF5YraL0j&G@&D{Ba>_{J0-{L}L^LcU4KoAf} z*Qv+@ckRW8e2qbx%=$U0sU(98p+-%L?ONmpEH=T95-YD1p+&+}&cT^6njo1nCu1>V zGWM7E1ZOhu?Ve!QwV{9^46#{gdGmD_**I$|daQM2gua}CcpQOL1HXf%h8u-<}zu-Y?a^S7Ga@8@s~_qg@bKS zBRO`&Q>loxexn+MMkr?*c^X=|H~Vcxt0upyc24KwXolY7Nn5p));Xvsd!XIRlOe5; zY+pH=$AQ0sL4nJz>&iJ$Oq#3uvI6G*Es{jWD=p!F@R{3m^R77fK}H9Gx0(oW5<4~z zB<|S1%!(oZ$yAmeVr5lTG{Jlcr+WA?3~Cc8++ke6RYOVw%Q?OGjQJj4e2I#5bhG_#M! zL|0*dM67(hyVt;H?nimQ#Pz{UU{GkgDI`7{ArbTjoKhZX8-A22g9cRL^|en^zbDS69weZN3J>T z7TPK^+rtY{_`!Q%x&mKf$LUQ1YirBL!y{UMpll7uU6VFa+*$h~xD1I^D_B{`zLygX z)f7j*+)>`)ICQ*4DQQJco|89fmR%jtQX)i|nQ17%w{QZxIqG1_)yQ(GoN<#}YQ<4o@L6_6ro9!XT!9 zFny41=BfXcV&pMNXp5qKgzvaNAxutavq5KXk1CZ zmR2WW0C;3?_~_$^98d4I61qOs?ZN_9$WGMjYo+<~&13JI_7>B%I4x3e3g2%y;)tYi zr~5HOJNWe;4UW%c8RyL^-w~~yqQrWCB89oox?-ZRF&;l5LH6}Hs|gI^ji13mYdI(; zfpK_i9cGiGL03lf*e?cac^qo%l@GUDS{P{|HQ5bDipLynJ_d6$!q&OK=Y0M&fkG&a z0OKb51zLB!V?{(-`$TYga$U(H9c;{I5qw()K{LQ+sJjx$Jw~z}s!_pGAhNI7~nxJmRI$|@wCI%DH&pi0{rIe*wE6yp%>gsC0nt#xJvzLz&dES53 zh0!gyLHC{Z8(vQ}9DmX$3qTRd7pjaNEG*^HG|8?A?)OsU?oE@feO3H_y*|jO1og0n zL8|{F{gf6cM{;uwL9s*d0lNm8MAptcVUj=@9Q0HoSfzYd6uZLC`}n&u0t$7?`f2MM zWrvJ)_l^1LLjWy}m!R}V=pAq9C-%J0bS%`D-XmVK%a3Z?x)GUvv#EG^gb+ z$(r@D-j8l}kzT12NCu#PI;UHLD$|(VCI~55aI|=HXqyX9TU-!RAIHS4 zyU!3M{MI-+#wU;6q@CsR5SK|= zU6Ku-Z6ZTX)F`-mrM0jN%owKzprgKs_S%|w6ApHNt-!T}QF*cG4#&g6x`^tf)_=2y z3@1Z)<}J#?u1t`noz3{ zpLjP#0zr5`J<_&C&g6q)L7eFAPbptVy!iB$V^(8*Nk2IbK^?(Ljf3{FH{PFp^mCiI z z4&g(s>8>jCnc}xxYQwn@N0c#^`gvFe&ejOD8GEz$eM>-j1e5+arOs4MO8SVP<#-tU zZN;UR#XwY(bSbfnSS4|bg>?nRP?K#T9@cU0P!(%0KSNn3NaMGI)t?bjM0hW~ zWRoHO3Z`p!DWpxNruD({!YlqVDZ1_?FdQP(!U23qXI&EQu1}r$Rgb_t*d1{W$%;2J zRp4jXSB%l5fjpH-xu6OCtYgUUecsYXi|Hh80N zxP?C+GhbybXpLzRgGAx3FLvp*bnwIiH*URwc+u(tlS`~| zzer`)p_205nl(tZ|7akvLGm~qWXG4HU~-{WJAsK7Au z)N+WmQ)hG`&d2T@oYKg*`^V^_?Nyy6)S~HcbiV6FADQz~aPjX&n2~Ue0_m!MfC`O5 zDN--7(K`NDJJIc(*+Pszm3_8iIE6Zr;K=E5sw=N_SmTh3=U$+eGVY~#TQj6ShV-kJ zXY(&~&vYr(s~J*8A6cB1B~aKk((7k%=ZX2(G5K@KaWB z0*#0^^XR-S5J>VlqBCR&wp?(zRfGES2);;*E*jfkUvfj{ zFRU=BIh}ug-|9#oZyRktYdtLZ83WN=;&i7ib@$9yIXCncmiwdR0XJVNKAndrdSb8u z1muXUL$q}HKN6FKhYDqGWOHH3K&`FqcvK z2NM!9G&3^_FHB`_XLM*XATc#JF_#e;2o(h}H!(9cmvJ}-Cx480b99_*-)?N%nIw&E z+qRR26WdN>G`5Y#R^ujV)YxfkHdY&)*?Yg=Zr|^mKhB!9W}fS3-uE+6qUMgK zKq*HDH%3+_7CwNuqPl{H1}h7Ig@v7og@p})l2XIk%?|jRgFvYTbaAzIbm03#M%)Ex z;s&Bgn7DyN6@MKa0CMhj09JMYD;FOtHy;ZNfQ^NP_kRQ(UHAYJCLY%207WK%oTCHK z6@gOR(aGDz+S1Am~vJ=qZA7O=m z1Q-DSDGz{^iS=J||Iz--khR0#$|h!Jj`mI_4&K%dmH-QDJ0L(=N`cAE%Z&kG;$Z%l zpoyKUBY#NW#KXke&cqZX@OSVg04Y&bfC(ta|77TD=3?#S=E~%1ZTDA}%zv2y6JXnU$&T-qZh!Jk&Okw$i~hN;9}(l zaDVY~0sQ~lkcx@*e**bKRo2165y1OTyr4q;Ps|?w83daD>{)Tjl>X?0;GL|Csmx9g~#1o!#FOH2)*=e@U3wTibd6M+P*)?rxxU zP;>+>g~R_A)dv2v#fm_4Yj^wq&6RaC0e`K9sDmYFv>ADsIJo{HTf0hGdjZW=tliA4 z{vMNm06##$zsQ<@t(~2<15m}$)%vfa2EfS5!t%d(pj|Vw1sx}@py~XZ1q7|qe|wd5 zFmp8jYv5z0^bg_&4b*SM17Mc;jd%gflK&!ZRsgf~Zv@g){Ea|*O8+7r767x#Z^Q;* zR{M?E0nF;Z5eI--<2M3DruiFzJZS%mctObIU&PAB1%j?-)}T7tnFD`If#NXzLki@~ z)WyUMXa}@#`-9B+AM!sV`>!(VKV(~=+kbz_^RoX-|6dBA`k4KTI6%384n+Guq6Br9 z+5BI~4vN{_(az4qY@txg4+wxllR8NcFDxlyjtUdm)@fX|C z{ZAht5z9XyC{C;2QG;q>WDkTECFW#;f-?crek5B0yB16oFACs2nS ze>Z>wR0t#T zI}=x{KNLWL{0<9rFfzMYxd8uY7$|=$H&4eu6hJoIe-9bR-`^LutC^$ApQQoS-Qy1k zDxK#aBMvg^^#=rb_5K5b%IfnwW{@QxpvylI{MS)q=I#QjtJ~kV0_Yt2AN==S4+MGv z%@9@=9nJVdZ5l(n9_vL3JsI~W*$22Z@88@>IS^v)z;~G-z9`|@RuFO&-tm8==1QoH zTM5ezZq| z#U?l+9Z{~gb>@j$P7ks5>;ZqXrW?a}VR29lnKN)cTpGcx@zRt~w}P9mXQG$wr&YJS zNDp`)LuqsY_2CUe>FC3X0%TGE#XCV4ZdIJlXn^~dlrjma#q$Za&v&x1%7#v!LR0f} zp?r%-88+}1oLQ`z}I! zSvz{l@>nKJw4?0||NhM`abJon@x4(qKD_SpLdvk)77D`12Msx`szda1AOj+_3#76{ z|D3VBiGDHBxgY&BP1g94{PrSNfTWa$K2H~Ghly9pF7- zqsW$>0S(sE^9!V*IT>*O4!wS`w@T4(G~QiMLbZry>D37I!Us>YQ0CvMsmzx3u?z^z z*I%NJ(R4!6d5kBgSab+0R#_I%?_)DpJe@aBnJ94ls1wh>gO`6w)j?hqPT^F;A)~Y~ z;X%V5%wi&YpJ@1z)mM8Nq z8IR$@!n_re6DRgW7&p_NtrYi8beLYmHx*VU`oOuVv4U~2dmvlbxH>21>CgMoU<_er zJQ_>czG=9vXKjBPTI)7P1?8z$*6_qpFcrjgI%N!2)Q%mS;vXTHAV z@QnH`67#09sab<1Hk6lv?mUc?P)^0f>EeT6`f7g>93F%#r49p(#c}e7(|gs>>zRA- zG6Ki8;9jPyL)&dNf;?EC_u@=X)jcfy+D-hvUR{jE^fdMPNpX3`3* z+P8n??~BUuVPNlBDNYPfh0$#vr59}bw8 z2VO<7BIoxQLIR^OTNC3x;%P72^oI8woVi{+`qbxjU#d3l1Ajp^u(%@KeEH#L=hQRDeunA#xd$oYM10u3t$rD=cJ z%Y>0cTEmGrxrc}<#$A^8h*8=7S_bu)EE$WVG2a2~6;%tY8oH{tc(u&FXX++7Gj3s} zhW@b`SQnakqJ~+FMR44Q2$FmsrmDOmSrJ*N_cM(>MxCGBS_n8Aq`TQ%NxPnRLKI%b zQLA~APfj_Q&dY0l9GR7f_P^tu$ya|%^oUMem+bC0hZ|AB2ny-K$uDl}Lb00T5+EK6 zFxg>op0F_r!*3CKdWIPC!*LR2i>4$naL7~*m!dw@=1Ab|R57E*#=8Z}C;3T`qYyQ5 zfPKvLy)M9)53FWIxBNO1XC#aSq4(SzV`LK;jg^yxAv!7&oNRK5IbGT#LZ}KRc zFBZOv*&Q&Y_b<@5ar8hst^&Xd(JeVLJY-ZF0G2s~p)rb=-!mFsg*> zZz!4WmvU0_%`YQM^0`ApW&D3V+ag`&^qJAY99ofhmlGv+y=lpm$t&L1chpTDUN34- z?8tQtSkOF1nrLl%4k8UKfsqT(?hcYjWZ~D1-ZM^HUj4G1ej^kqseh3d}5SCgL6H^J5I}X6v%8BC={3Qz3tUIOoS{=n72* z4n1t1II@T7gZXCRV_v*&6BPX1Hov1euOG~Wd!eo^n#PlF0p%1%BuDs}uH>f&_rDS2X zWzgkH)$l;4#b?GRrx1U9S$P^SgLH^4Z=xqBH||D=JY`cxLa>&t_sEo2lVMphh-|x0 zxe7+@>OB$f+EEf_Th{{ZNLk=Dnd@ii0zZA0&1d>F=5SZRhNzVj$?aCuLNp!pPanof zijpnXqnaRcqK_$-^Mqrxxw6^k-hYZMJXah9GK5q>IZTGd;-P;q7~-ju;~yneqV>W| z2{ExZQ$ySkZoq7#UWoX0T%a3oj`6n|6SRp5&1ITjGwa_@U1&sQ|%Xw(0w?%m`f8C1w#x2~GS-r!*u*U4x(G-AOzkcS(IDCZvea}$c zjlsNUIbqFXS3G|qn|#DW&eC%1cGGIhKTmsCF7KGQgkOD%{b``pfCjehgLzVf|6H|! z*dBK1{AK*1FPgCf&)yWJxz}|Nn;!!tiS$`N?apT$nN=aLibGArJ`Sc9KRwQO`o`A0 zFQmeDOxTE<(6hd{n7J|wy{T#_Ply-ew6SLa(~P-y2=;%0_9375PEhERWQ;f_s=6VE z8yT#Y@d(GgBu#MMKI&V&H!^UhHmxe@P3w?Kxm_nr*c+YVNI;i^l_o=8b*b2^Qpq%} zP-D9b=0DA?>Tl1bhvy*8TPciQSl35q4QIx)d8krxK&3C<`8m6+s$-_;hL~OeDV%6Y z$i~giji7(co(Pe_V!1zdSbm25cpYUvCn#xDl_|y%l`=QU6{I%=Ubf}F!*Si zMGM8%vRM{BlaiCF)lXoQ=M4ewLL$$es)zOkc=zTmD+JR%8AIMKX}%E-Mr20^XJ$NG zZ#WkB_zj`~@mP>-tu8aY$bVjDSfaB^h9n~Sl;9&)>Mlbl|0jR9Vdl*>AM_uwt}-6v z@CtuOPK7j1x}KaNW-GoX8*~TeUEEh&>`nCW0AzUWH`cN*XnPV70pJu~hgikip-|Jo zrk~(O)+I9^oPmCBkFlcoL&?H3*U3dGy<7G$V;+?AUsk|68aG_1m4CeAsFHS@5Wgx- z<@`E6^4*&2{rpUr0Xs`egi)+&D#;*=1jT<8&hF=w!@#jCtWo_lylH%tm_;O9Q{HlSs!JIz*`BqFNg-h_B&LUDJw?=^p% z2#$JGMMja~q+vLP~jFx}p+m#Uf=Q?BqC_%liU5M^zhtu80aPAv@{t_P`)zMD%Wr#v7{Q@Kj1Cd_}Bc+tBC z4_OxRDYio~@9D4@FG|0vU>niu-tzyU?7bYT&KKyiZ&tBYE-dqQgB8JuIS)k`83)e( z3xm&QH{Ij~O4lawe80E^AZAQbpyeJUwQR9io5mc4W$}WCeb__m`W}BE@{}Of8otxz3S(mZrmZ;XhN{i+^P{`<>eNwP8x}Wv z_GWk00#~%|k5=&gBzATn5%gepI7w9!(Uz5d=>#_X02|kO`$o#7!r~Q#uGv!z(`^-_ z0h{^u)!a_GXT}yS2Lg(Q7L=w%20bP2-{d=fs;|c{_hV7}kZwwNX6T zKrj9|#-vT9I412=qYuShmnp!4cW1AW@j_eJpf6 z01A(q@pFKKY@gO(so*kAJp0UglbW`%gZ4iDs{0G7c(nT+82RZg>FNrC0L@mVTcS3OZ+*n}E_%3fvO@$WD3(AKW*pdkb(B53iSgwCPNoQLtW4;E?m-CZJ zF5--O4IX(sM<`dx^n{A{5FYZekM_bhqIT|?m77UbBn>UGF3vQHsy$#M$ACS-JU*x$ zx7j852s%dDI3?Xr2uYc?P}Ludq4p`(jjo3WI^OE&@Q4B3W@=JE zEl)y4o@1Evwl{yp3k(>Gxan=@AxF3M>c$-7Hy*rlr#%&cjbyj6cb|vNFq4W!=Rh^# zVyNu(yYYVrogN5KadPF%nBzu&!rc9@Z4ho+ z&!oH9!fn0DNKGKX*C-VmLw=0zzB_wxqre_zWhI)crW%+uxFyWQW!=>IISxdSbrND7 zv|_m2y_{DPei>nX&eEh49guscq*VT)`xepfh2T(a&1q@18OEftX~F)En}t9LpBw@P zVW+gqL|lJT-abn@E2>9oP!BNS75)o1bkT$7l)u5t7A?~^_Bd(FYMQp8&A=t6=g0d^?4bE?5^a9J?3Ou^6P&~g+(`+@xczDRED(|^jTp2@=P#Z zTI*4G*xVthqp#XZd2Scn%lJ_hBx zmQ7H-IpDHC?=aMM^l-MGD|HmgLCr3cwO5`dZ=Mi1)X&UbO71T{f@2>yiguq8N>KqN z;;(<&(~>vVT$A9|;Ob;dy04XV=R&uEq4p@l&5rDaQeolG2zjJ9H*XDu+js6${55Zc zzc|R{KWy?=LC{ubRew2NJ@SbZ2h{J{f;p4BP3JoJ3NKgG;md;~wIza$Eiyz;IPyS2 z3Fjk_W{@*VCez!9k;&c@a8`a3=_CGnJXwDNJZa`V*%o`iWnkbh@ZTVQ z-*l0a!<~?3OnbAb0k-WQzh7vNqWh*9SbB0-We)o6Pf0t<21zwczNec?VnXCtL0BBo z+Bk?p)wI({6({qYk9WkulWE+qKkoP07~;ueO^Zr{NGFLl876BBYilVpUun3>mx5fI*TDb5KuzRO-QMO$4tmk5qyG&!M&W zb#6)|)?m)}qfs7TgLV{*oi)GZf!;2wLR1eX1apZzFIB-mY}PU!_G02i!S;W+8FRM; zc=6D_P@Pe5a-l@vn&XfNcaR%Z5tDmjZ;3KMb{p5pHr%di?Gdq%*v>!_!%A7EKT3zG zDvBXG_Pki^3sb?Dr9M7M@hccg2iCBv7?S?HUi^k^V^4W)3*Qb%qF!MzH&GzMszWDI zOgZ-dyi(I3N#9wA*gcE9P#%Bjg^oVNL0$M2K*_~MnWlrQ*Ezw?na2J-E_o&+;Zt-;j&u&4j+J*Q{-;CwF!a-H;WqCx8 z+F-$Zf)Z|fhbBgF){2Xo(oeOKx4ZMLG{E6yECHPj3NYNCr}h?1jdOp>9a&TI(YZu0 zWRf|w{}eKB#>(cK^z2r&1c6Uvj~I6kX#tXyXiW`F(YX69#6v0B{wlwN=avr3%erT( zlmY||qq*um8$)jV{BDss3m=}Xga}ln-9G+4(uqdWZHQ2=HeqYJ%{aP`#hTC5%X9M1 z>rX};3yL^bM(Zb`>zIE)l|aJ;8>TaZVr2Dx;h4j#h%6~O!Tqk`Ihr$%e0c@lq&m1$ z;CBTXBD9aJ^*Ot~_0C|1UD8Ow!A95gaZKBM6O|vxUJzifE#OM9I62`FhDCiwOZ(uT zQH2)Pk#E%Z5O|WJATH?Hn^coHCh#py8+;++tJD{-7(@|VzfOPu)PhI)RU!URy?Q+j z390n-UE+Xm#6)xWL28XiX|?v*>BOF^HYIyP^XG-(kEoIMx~`b)1ddbhlvO_aY5sR{ z0q)>kdm%_Aj?;~kr#-AV{GZfBth=O{HQgJy&Md!#PHW4mhRayLni*G8`MCSnbQ-*d z$PO8K%cq($Jbi!2oH{nQ+1c3)+VOgqTB~QiyL1D?-U-dG3xOE0wwi8!=g^-VeBExd zs5(Xu{yi!)g|u?VOjkA?MY}aJcLw%8cx6q?h-YD3S^^B~2Dx0o4*7F(lU~eplcvSS z7J0Rif_By!MrQWsdI)b^wM;44^~zx^*3E?^4hX(DR}+6dwn7|{mFLv@euArcvQkPa zo;`)_!aQ2&rhy6^Qmbc{=m|Ki#cjl2{n*m8OyoO-nosX?k$i_j;&jZHN5Xds;E--N z$0qkxh3bS7_^WulQ6gvq=?BwZKIWa^W2huvN3G3UI;EnZbGuQ!<^QCVj*pl&K3zV6 zzX8yI4oy(^WqH4`TT7RHtj)IRn|X7cal0bQidpZ1+I2?vB#-7F?+ zY~Sl6%n{V~ak)DLu~dgw$naQ}?v6+QOn9Xi!ufwT|A2*QTFj_`?H~F*{LGBbuxAl^ z>?Iza*F*=p1qEzTR`Ks*5S2gH`M(RKJvZBNRB2nJD%gG zBCcEDL?@hOrJv^w(rZ_9)4`e5eIv&&TPmO2$l~#bqO3It&CB;sp>3#=QQCTyl^m6$ zm=b?c1{$5X6bE#grFb3|$18$kWKX_g?`vJOrMCrylo9t>8fb>E=V!Z`4bV_YzU~J4 zgyzWSMpRy4jz33_f6M7{CxPUOD^&ArXXsHzrS_6_wO-GmfVRmmA78>tE5PEJ%%+?$ z-?7!EbgYwKaz~5G6Cvu15daHunDXeQ>lc5yWbIEWK^)LGa7z@t>$s(4ddyaD8`v^? z!9JzdszV7w(IBTd<~TQkz*1XfN0kwx^IM<({v~=Mig|_grlJ&Z4RZlb8KfR{oBw2G zYe2$TZ8xlYhq*r9BzHuice{)Tlg<*Qa*dyW$!uD+&$9z)4q`B7Um`Y=<8VPSFhqaq z?AaIbYrXc$P^Yk1V29&79P35Uwt7jBq~s70`eiVP&8pQw+Hl0v5d2~-gMNBlW2CH9 z8<~V)tnWqbJWtI2HZkXEP}P7k`08>$X2+z1FKO<4^-_X0{n~r?odVc@gF@3TLP@P6 zqqnS^k8xM?4SGdd;W#2|Xc^Qt_ild}fFtXMR4KQkPE6$SmY|(JV%n`m0m zvPKMe3@o1#)X`r2+{u5U3^e{wcnDx>c=-rT*9o3=FPJwUAQ=3)G~ATw;HqTKauz?A zwSy~A*zA*0mY^FBLl`Xz;R#!!*oX*X-K8u1y6bnJE_>@3T9_*QXcOg1Qv9}uVXQk- z**??A&W)z1hdT}n@oyWPcn^P_ZRHuEM3dS2hoAJQ3E3(~NScf;8!lz4wRks`5s2lc z7fSj(?knZx)dixitTBNIV{-)^YrkrCWGLIm12~2%xRz=JIU?RlY(EM6<0CPlk0*$^*hOn+yru71Z41OGWn98}ejLIh0n2}L?(q!^hkj!m z>Rd45l!_CxU?@!|m74f*1tmQIS$vJMa~W1|`^rar)W}acA=`)AIIKv;dRj2!EQLq62?HV+yQ>Q*#^1@cZWGC>`o8R7kRHbkYUPIzd%?{1l!Vmn4=` z-qFdWB31UO&#Y#Dxxa0hQddh>K2SKjmSew9oI#aZ8Pz+o02tcCWJCOuz7Q`CTsi;( zFlwBwo%oXl4NP2+2|YsMvHl0T32{}3AS`}(!JZV6h|Xs}THk+eS@7<9elOoSdt3|; zi_62GxMmBf$|JCbaMD8D54FqQ{bBVdw_82$yhU(5FDj_|+ug4%g8@e}HqhzTn7^1S zb2GzfNduh)uD|ZS#dFJU;f5b>P-SVMrU*NIKykSVwQIvnYPj^f-3=Q>;K&Ci$%8B8 z?n;!JeK~9Z9K(Ndbis4%N(Jww$b$PI{Rp`B?E6?$v*&Z4n+x%b^wjOeV-QV=m)5zI zQC#RI|1)^F{zDf;bx$bvTikA5aVrrlyM%C(tenUm2_#%GcC%3z8wR4n#v~b!W{aH9 zds(0WwZskH<4m07Ou@Qkn}!ls62~@EUe6QY;0m052seM7igPALr{qbMd+W>T&k`4^ zFYXN4823scHQ^7I!ZZ!EF<9u@bzCmHyU2SdR0k+BEnc?@es01nN8N$Ax`ot(!y!%m z(euLUtUpPb!E)X#o79^n zUOD#)McsdEz?mi$kVCr^B(n){Zp5wBTRpOaEspbCC|ux1re!~6@1It#?Ov!H=g&%X z#uSK({LnJ#NVxaWX?}UXUsVo(z1z=~lO!4k$H{{h)wQs|B=&VLfv3qumAe-xPGFEW zRTIyZs9(>|rMsK@U%WpH*{TH@qJ)ik_Zv8@M9fpn#SU?NVh6({bZ%hCMcN zHECadR_^54JoQK`7m)af>JWa-wgT(IA)_Fyo+(4pfaPL$mu{{(wwe$Vy--JLrj)6Np`~e1_xQ2=pawfsY$&vdUnG)S z5vjWsuILZ19(Y7zrp)g!Cu|+Z@saTZ$5!qb=~~DvLtna~sUlP{SgX>RHv5{Np#KD4 zl}!07hAwSBK+F@r^T>2N*5^nC+aW79)y&Fx9-!)@Cso$KQ zd__KZJ)Fbzb^~^kaEWrM21{&Kwyx`5wTEOTTKB?E0#6o|cyc`~Z^1|}nZ%Qm#7+&L z?8n~S#~Rf*2i)&crz|-4r?>7Oqw0fyrNB~O;@=D*QGKV-lweUg1TeD-^3bzN6>@(B zggY z{8KZAihQRKPL*C>>md5}>`_6k!goin)C5q$V zYfS8=Yjx={;hjcXgwqK@<%fS&wyjzLwU{VFQadf(6Jvwoy?;5 z3cA0#P7b_e4=YC%W0F&U%U=CXvtKf{VwGn>a(G^DRh;AVF~7FmjcG!5mml(AOIf2T z2dBAO<~s3vD+_H9GEs^4`j_E&H%Zmpp)RCkDg?Eb=@UVv$)`D&6vThxv!6tC(-IYH zyjmRWsCOMcu4=n0FCiJ2v7c*u;47!zTAaM8(3-mwv=^<2%yEsT8=u&cR$ZzOAGr#2 zz~*NSl)mu-xFcSBvlw9d(CH|c9&GI*8aApy$oNT{Tg4-q}Q7b znBrN8Qu%eT&as{KPR@VbC~`>;5?!|4IS*gPk#6~`*<>$h zj}~rZ4L*rS;&v7?E;zZ}PL7jI5l=n496mKxt!pyO65e>!8caP&?k6vD97FcZ#wyZE zVRo8QGQTqGDlki(eqT{p2~$*3cGsHlec4d>fp)JnF0ZegA}m2XGs=U38jD7{V%Ji~ z<8ZNW4UN&eG)aFlbV^YSGp{7NdxR(N=Ylbl^sr~E2>a)Vrkhi~g~C@!be|fXPI>iF zy08!vjd?<6U3c=CO3JQ#tLIiBk^S4_5RMozq9h9swF|V5)0ex?>&;?C#2S?KlyUXQ zuUx8QkOJ|GQXbiqQFBa{4r@Hckc3CUwWGpx|<0S z-!b@=yehz27ir#%)c}u6mJqPhpcJT)4xx&)UzbJ}@oz9W zFpol&aLRxD1K-OQ43{#Ix6*ZD4$_o-E)`dUO1Z2nY(-kHTg{nOq)p{9h7Cj_g$}$7 zEDra`OL;{bQ5f$wj73}g@qKVAT4T@n^Ref!kU9}?Ozg1*4?xKorT<`)SY#7raYu}R zZ0Q75fGGdk$oE9ExBtah$}AQtq*^oU=9jd+VsA_K1jw839Z&zn!bBU zubLAgiKL&mkNiR|hsNA?PxJb)bNDfo+7gcf8F{~6)kg@PLBhFkFQO46;oPU%(0xbG zW~uYhK#w}EB6geOL!}1I_t^8pP@>9M%f<%{Ey{Ro_*y1^dGRV$C;~z+=Z;G4{1Wf& zCf9$6`&~-a8oGF{jc;;&e)CsS&*S!(d0J3m@jHYjIOfO-4|A5Z^$7c0{m2C^%TI&v zieRUM0(=Eg*0@r1R+|@94_k?A&DN03+WCD<7>c$O8pj1Z=Zxu{(FAO? zpRTvm_MA3$Dmix~W|@Td5Zb>%hr>>vt)CrokK;Cft}ko>TMUak|9S}PQavA{Q!dpz zW`6oJu-vgw%SkHQ5Z^yntk2?%MDY6S%=@mTpd^19 zh26>?Ij9D_Iy6s{h9~g7Hn*Awjc2ty#7T<}h29BD7o(K-7UIQxm3pLl4+gd21QzSF zqrvbr>-8q!Q4y|T3oG;f!Tr4$z8NCF@R6PfBem;Z*?2yYpCr=e$8n7Zj1b_gc6^&U z5A)^E(RMHK?d0=A|A-+z)F7Rt3AdchxxUnBTg&DWMv3u0DzfaQ zh@;BwpCvS6>%A04aN|F5+E=5n@WEC4%pa2lSOm)5yj01A5@0(KMsp$koOw4uzHjP$ zR{dGw-l~7!c;tCor^Cza7i)h%Y8v_w)AN1#GxI5OY`nxx^~|*|S5d{@Ci3iS-s_DI zS8(0&5FnQcf6I=De`rC^Y*zUxT(7Bl&IPJUUFGa9Z9Ly70d7$ezw&Yvmg6g=G{?u) zZW&tQHmk`JMCVW|)@P>ofkgz`oeeQ$Ova9ELaHa&$kLs)s?X4d?dyNP@)u+Wl|lrA z`9GH>emz&DZO;Ghx_3~%kSoFYPRdDI;;tgoYZz(r3%tCQ01g_)FUGM+#hNiF`1vYi z-tD}mCLV}b;e^Q3bciU!$iF zFUS#L?X?%Sswkg@i#>moapm<{Ltx!?ZQo5s_veyDRveF13mBI0ZiDUlQxHL+;>cL6 z@d|z>jkEz#XeLET`z51l3<;jKwrd?&`MPIAD0#J(gN zcJvVT_^9n}pA)(=!s=8kroBbi9v--%4xt?qUe~u&ASk`qY{xVoX z9?-#2e662QuE5s0V&g1LD3-`e*@ZFe%eJ1e|TiIuZ-ZyCw~< zXu#@^SVO0}$%lWcv-y1a&u3xhj)~!(K~&X3Ff%V_;1s<(5}Gf-d86X_)H1p0uaZtdk@XI@vzV6WM@`>)vUi{K8g5D2tRrIu32}4K z3Ep}#RQ^hH^h?5n93ePo;ZM9p!CNjk(F^B2f08d%(@cW3Z9#U<_Nenu2QlA6NLoDd^%$d#68`#VJi)AvVT8 zJ)DZfeLAj>p8%!i6me9?E+BWUBWEr(9?^OZacEs}V`s;UZE&?vJvsQj zr89qn$!wVV4N*Jq5>K=+il{-y1?v)1hN}XrK8To_FvMtfchf98cKP-p?-OEri=~V4 zFdKEV)d&e!fY!>x=ZE6Zh~+a&;6(PNM{tl>80$;uX^GUKxE*I0)S2PS_)oH13#D3(l z6CrTs)P&7T)0CD=s{#A{+Vx!0raVp(eh1zI#jsA@`MiPGx>f00)qgmxupO8bvhWAcpRm#3H%){Q)*QITW zX*>t)jS6yfMH5P^tJOeBe#P(xlDx}iu2=2(>4EOca+mMx9%!EUxzi#X+q6$%V$GH5VU~n6w9P#$u zqzPmLi$gI@jfZ(xz)~8g=}?`5-vMkO`#T5~g-0TF zWPRf;l9|u^$&@zvl}v+#q~oDf&OipVq#0B%_d&3*Fejg9s~~^9A2|CCei<2^w+||f zOm|-Gjt_FJIP7>+YbRx{Ywk~{fnCHXCRsQ(bCbR2%dhp~P=bHTn;w3pd}>LlrVZ>w zUp*v~8UkiQ)Y@riwpQx;~dz&f(CB7zk{ek zfr$JGZDMDs!O1RNS7g^;zY)x)rvL)!rvcCrw+2{X!DJyfF$_l-2`5U=Ei z&8gwh@$>meRzjYU<55KnsO3SVdp@W> zbh*Ivw&;KPnjFIM^6@X`g1hs9AWntoA696mCzlAqTnR&I8Z{Hx9V8SftNGd+ld2+kqUKijZY_>rulx zm*@y3pGYqWS41S&bTFW=*8B$N5Y9-xt5a*l(Zhcu@Od~;$^J5Xd`OGm=7aaa45)Ax zi=kh+59j^p=x03^lQpUtnC_*FUPpYe_0(T6(%UoRuoe&Q^|o9mbVnN(^YN&WWXMwh zE+wc|m{Qg_>8F$%@iEV>Rx%> zbkBc2sso@?2%@m^aKaTT4KLUr4N-TJeNe_;`b$p-F?S3p+p>4$0%G=j+ax?bw>B;M zqQFhLl7F~_u0nrOCW>fN-6EswY^ShezG$-9cczwZP{T z+Ndx2lhRMh%HPdcerVDOYINL=A2Ug0x~00gZgUM2V10Lec?WIF>J`;GV~UZw2hi^y;I!*K&^v!I2OuVu zB(=xjikPo&mdw3p5|ifcMjnMEcMWXNDbk(I?Jo3*=Y*J~Qk2KtLy<7|B(OaVlb@i# zVd&@AJ}eAUf`F5GgeLpSp_xx_sfKp+Dbe8RB=~rYZx{uqvxj9HWm}4fuM)>=2IVWy zy;=8pg}l1>Bw7MtouM2Tj7WbvV(aelW@b6K)OsjcK=?b<31l%1nl4h`Sg)w8v@&rb z))=@SETGF9%kq6kI;DShiRcJDlhc<`)^>>n8nlOCdPi2O*>bbsbY*-S?hyJ@gyshM zdk6lW$CV!KLtC8Y&|>fOPgpu4w=edMK(orGg1an9QX>8XaZ-X72bq6^ukSdt$XPtz z3eIW?ZHD=k0}MKsE4wJ{H{a$|T&C(dF?E3Vlv`U6ZZ?Xmg@emb9*ST90kW&VJ&Bt_Ju2Pl83iFkt=pL@t&ou5}w^)^WDT{l{jm1ftn3b%rG10G89oRw1S zkUG91GT156e`p)wXS*rR75p;GiOL1NUHB%s)SNUIZk92Mby4^6Yg+{~HSayFJn9L= z#-jrCH$4fMZKQR?G?|cI#H@gCJXAx~7a_=K7yAPbc=Cwr8X|w~)x94{r!sbJdyrMX zvD?lgp*TisK6e07z9N27fE6a)tNMw_q+|U&Z*7Z6ZLRGtPTYTuNk&dv+t4=*wa!NI zjv^xW{1iPBye41s%lbOO(ym63%oGZ%r&PeU4pP6asT4#@=6(fJi^eDHb6aBO+0nH# z*la1WteE<{T|a+A4wO`X=C6%omnL&NF`Yky4H|kvIkPYK+F#$L@vh8E*xLFEnI54K zF~2JN6HgTdo3|dan3zBxxvLyr`0B#Pa2drqU6ZCY73jS< zAkbJi8}eYNI4Jj%p9p8Xm%n=@^{_+G|D;h($4-XvF+G3g?N*SyIo%ljRzKA3sJEmR zpHVW-E}dKxX9#f*x2NqTxQ?HUg}XmG`bLiDHe*=v6gAQLQm*5^>sdvcYyNRf={&F( z;|q%sHClEVZB(VKsEUrw=RJTy9B&xR;!Lq*OyB|ZZhtH{^Tw9O2?I{JsNAA2yBUco zVZu$aD`bDCBtXYA11;&6nD z{Dra;N$C)&$}ip~cDp%+cPxtS(|VNB<(uj}ez1-aL&zk8SK*Eo$(9PZW}Ricw(Hqu zYYkDf+U6J~XOLzQO)Yu*yAlVvy&ph7+vn6(tOI{{NI9`QcsdGogD*{IK*8|aA}E5- z>wAxMYQ}^+k$1qhZze~#=Wf}-X9_K@?Fkgn)9p{DJ!43r5)+JAlo|^4NaQGLS5JSO z0mD?~8R2nSqk@~7I9ajOMw5aOFE4S5o$QC+Hq*d%0VSYbWn^=}mm-Xu^a6H~lq#su zRYs!Vk#W;%=Smu03aiB!n)AMIKE$w_he=8~WWI8UR$mpK1q@aMs2QD{Q)3LgehIkx z8^^Gu5_Z)IsIzB?3Jk~~gsE9D9c_P4@hZDUdl@wKSl(#xym}=r4SDm;s)S>HSoNu! z%*+3XFg~;{w+Kb&r&J5c#t(-z=9Z01j~67nE;+FTWM%S9$0nHEc1&rhmN<(G5j;FhhK`nsG8`t>(g&r82 zx*#P-6dwlE$xCn|s>Cx04@Q612u^qUV>G|_2IrTM{5X9&(Yoc=S6?Q5;K3Z|I8H*6 zZ|+|0c8fUGZO1L#EP(Mp5=S|I0RrP~>8h*7v99}JI?W1=SRU-cZaxrNT-sRPdF((1 z=&0c|&c17w?SWHre~DA|Jz9QS&7Mp*@&@_J|s4&H2iap^P zsl32;wrJ&i7KM@CP>+hx^KtPGRS$Rc)OCG+fa|6|C{5H$`9A?Q0m}YdHT3I8 z`Z~x<0P5=x8{3A!8sYI-n`|fK$&Xe?Mv13ntbf1PJwtUH9O>qWGDyVo(VgiOD|=d_F*SUXt`?dk@<#QBGM!Gh9$z7Etyg8t z;{WLzy!%3P#GzGJR-j>jcXlxW{I|6TRY%cW4=d@o+>x(rQt?iJwk`6e^j4_eih5tA z#$Z^06qG6ah(z^m&LVmJi<;*>qU2cvT+rqP^RVE|K7d90KTGs% zCU}_3ne|bWRi=pB1W2kAfUYiQ%UF2C0C+%$zhAIRSmsrE@{h?{-=+)>HgXw1Wo#h| zWo~41baG{3Z3<-xZe(v_Y6>?tGM7>M2Nt)CFA1C-12;A_mr?o$6t_Q32~HaWH#Rnx zQThiIx1?wZKN|x#HaM42`UeydGdMU3FHB`_XLM*XAT>EKHkVLM2^Io1IhPUa11Nuu zcn4Hd+qyLZ0wTRiH6XnvAiZ~x-a({0PM;(xod%?}&ha z2>PPObMHO(|HpW5WQ?7?zHfeQtvPoxHYb~bv7izZWe-zFAu)m=A)qusMc3F81Oxzq zqC!BR2r(z82^@of{cR`aG=sUL;V6Hk^nV~!++ko0)}{)^V1c?QB;cM00ss;PfW)Oi z64F2*Km-Vs`WuLHmjgJ*t`OiJ3JF6KbE=@+eB9xVP8e*Rzi$CN5MBUCN=icD zcQ`=F73L0yfRO-QFvbbyimeC%BLK!I2pop-`9}&KStks}O5eyA;_k=+KzeWb=fn8yLMiV0D1eiF%(SJFNQ4Sa{usaNZH6Y*+7!r*Q z@jyah?f`6cfU%YiK;I39{0pq}7eD~;XLA4`A<)0l{fYjS2#)+642D2Zu5MtY4;<+T zaDXFV0DW~GA&fUh002fpe*u5N2s8>C5B3DZ5ny{P;CJg_fVz?)0F0gR&wOZzJKPO} z7DB@jzh)Hvl?J=bYDlOG%GDKy#Gr|P^`{DVhe5FW?j!u?a9xlnFQnh!D+f3d>hNm} zsE3=dDH8tB1E!_=2ZJ>c|J&vW!vMs9QsUxbAOP$k0Ok#G68=@d#K(UP_S^Z}jGZCC z&kf}UaKMfM3xGSouwTS}Xs{;?fN}SL1^E4^zmHg0McdTSLXZE?$p3OGE2F#teu5%W z06{5nAOHjci321g!~lN*|IA|mhX3i~KVr3z4k&=sU+rS|>F<_3|8Rikk3e_<|IDR_ z!ZH^I;Q2Sv9{|OG5bOo?f2RBIkpCace`Wb!iT>Xmse2$0zx_Oa0sbF9*cFcO`GdeR z*8_u9fi4QG1>}E2&0&8@R~H6_d$|6$R|^BiszC|qh-I`O=&pYd@a|uBI9eU<4TBoM zF%YM}So4?N^q1Zca3stCg@*rn!eFC-!2j}LMFw%fK1I-2YW{Y?up0VjT{R>G1^p#9 z5ixN9*xeoMLySFZ>`Dya2g0fn3iJL=W`M8|5{1Er0I)L#031;6#J>&}1Q2$^9tH}E z6*6{o?3QBP#Q%RN0r7Bm$2xy=0bBp?^>1;*VBRnY@!T{DA`|BPDy;3iT8YU^aDA9y zX}Kdh(Xuy648t|^%+E=OFiT_Np|TzH#jT+y2D4=mPkmJcZ6sobqHJWRAqN3U=n!{g7C+hQ5M`K(&S zXAXY+_%^4n_lf9Hfp&M5_=^Z?1{&XcY`H_p{S358IrkdTI9$qNrz&M>iR=tTt`S^| zcyOXqd;8BYbo0iM%T0CO5UCTE4CQN#(DLL9np!x}i zGG`SO=X5aRbCoC}uN$9^6+lqivG7xO5>--qW~qZa-w;dqC{)yDW9-(sU0hZYNW?l# z4k6k9>AAAF0fN2Rf>NeVJ*mxD5vm-Ch4@1DtpG|8n&1_2)}j2&K6 zHQawopr~2BY1sAcx=o_J`=PK1k&Sj<$9*Te5xuImCUB|GIfn|2fY>sW$39B09BmB5>9A`8$^O`+f1u8L!9oy4$(aHoZ0L zFHZ3fsqF_fr2E;!iVzhI31TfOANy#dO|5_Wz&dRwO+@Zp?;ILVzL9{e&SdVc1-sd7 z)$4tB`(FM~SrtG<&OIA0CA#`TH;7Bh51%B;EKrs7t0WaEc&t7|WEv@ic(6hm2iq2YFT+Rgltx}}C$_+Gy+eh)>S zJLMR@cO}%5H%*`A`1!AX{c0Eh7F|~_Mc}1%97}rRS27V6PGFKX&6P^KUaWt})UwT+ zEsp&lRcRTd2{l92K3VW=$=yp<=&DV~O(m^!0~_4(+DxyxEb3yY)_|t3 z3lHQ>G}QZhB5TY*@SgFYseyl1t%INTNZBx&F`VRJ*s`iEe_wdd^G(92RetmLS=J^U zPW^nv-Av8Uh!iT>MFZ~md(Vd!0zH$a1o}AY*quyEbQMl&cNLaX-1w%$Bs|Gx=;am+ zK%?w8)SM2)nku{|xA}wL95#7uc7WPIsab?AZH;c_= zy(b7vdW^eVc`!FtyG$3{Zp@x9ldzR`>0yIo98Ihwc9bcuEJwLe;E7zwG+8Bjk(p34 z+QuF}epW_pEle;h7#QFeHQqO%m^Ycu|Cl)oPIm383<4f;Lk@o|tkz9_n(`|r)hi3Z z`Rv4r{#L$ym|r_qzhSFuQZ!q8|e@21eax5pKK{i36S9KAK7E_5h)o?6yixg z^`ktyY)c4?w%^rCB^m>4@C^3Z4BgIdyh&s_YjzhuYkMP8RC_;ttFv#-bCy-gg~W$4 zo~~Xq*~d+WQFVV{W)+a)%&-s28IwLV3bU^Km~HQ?p`yUf8R8H7SmedFRg*p}9ye5E zWU$4t(bz2gv;95qKzqE}o!PA-D0gM|dMA@$a|6j+5gcKrUI6Ep_<+c{PNvw|5A-Ci zL*Z5TFBy9ZbjTz}U!TmyL57a&M^Q9$xRo*4I+{iATeN>aoE$gg@@9M`e3;8WvjDO5 zN$VYMEy~NjMoew1`!4p%PEfIy5J+W4aK*%Zk~3N}c3OrpE*&T&hbnxmv7p7L<_Wb7 z4evQ~(KD~Ne7&zeE4!r}P@BH(QoGPixMZ}?Ge0$IBQncL0mN0hyxMaOYsHBjA>T&$ zyN04B-|l~v-DNObZ=rn=mhbbF+CL>>-t#uqXZ}Ztg?X$!scQ9*`1h~o@RD`=c4vfN zoU+jEVrIkfE;V6U5pxDAb=#hL2b=5~;sGSZd!z@1dAZj0pnN6z#MiN@{QjHOJZsHn zQ(u;1G)gNG5+T|SOy)Gci|%>0`*Pnk*!{T6?+$-OSm*+Zc{*=(r`1G!X7CmMP}?x) zJE7BkKg_>5TJ;)h-(5lSZ_7ZneYb`f#{0$>O~*?8%aN5tQMGTE8D^W5)>50_v||t* zH6xl=`Bpe4!QSW*-2#?ygJ{0GXE0b?Gb4?Jh1W{=MoIP%=`*R2f zkY#`NM{r$5x0wRGlZq zNuOt5w`5(D^1a@=ZW5)k7%1F>dDN2Bn-Oun1jWL)|3oUb&wX@C>v2BHxyp23`;^+- zIlo2dh5*6vIln!qdivEDl5VYRgsRSG!~1`DqQPek27*Y`N!w97x9pf@`w{_6a1SJW zP8ORd5gf~>CU)luH1bO!Dn8xMTBTf`;|HGl4PQsLu_(Uhs1}^b>&`Ln1@*oM?++IZ zzd#*HCiH5189IK9bIaia*-(>(2u;UCJss)Vts&7di+*(S_UGlfD=7^EJVI*CNwa?^ z#9nNd*ji@$(y$TE9LTul_WHCpo^_Cr5S@%G<|AYgnt;bVG5nJE%La3SN&Y@g@hKFf zomb#eQSmOu`D0*30TB}t=u2O&@sefv_QUALgk5l%x#2daZ`kCmd#{>jEeaV6x=ywP{zkql0^qs5-^!nw9k6 zpSFiPN)v&iB`b!s+nvg+W+#t04yDuzTIdp;<%l0!D*!6Cp89!Rud*c<6Fnpbr?LLvT)z^HFqd(~k^gh$z_?gEPRWX@z5F)QAnp~8$tRs2pbnPN-OMke69ZY8M8^s!w&UvNqpNM zQ%aP{oyc8k!}o+g3tYA1{kVK(WT!fF34&dZR;}O???!*=>UMLUf8pk%yLxC5WO5C8 zgpp;Migh^>CQ>L9ab@7m9U65wX|CE&TV5$R_qoHfT4l#M?^n)5QNq?{$eJSz`KoH3nj zXNq|Whm(I#H$aoQn=x~{C|5=W`g}R#`hdGzJ3}>y#`!8wosyBFGYS%pX2GQGYjB^f zs60q-apT2*!Fx*ZBggi37G(i%%@j}bMRrmBOSP>U2&LXo@mOjJCdMM{UVc50v74gZ+)x@L(v+@y`E1!zlf@pC zQoquw+3(ln720oPV>zD1Kqf%^j6=QjeY&RV+>sfhg4+?H&-jyO8iPenb-C7%uKqsd zNP>UUxzl_N?O6k-+DCeIR!#U?gSI2 zP2b~gyITIpVHaR@gMd7Qa6a%sX%OX#o`{P>0r(i@aZf_@X&}W;-I3?=PUlIW_JX-z zj_%~5sKQJ_T$%2X^*3pA1{oSk&)bRN6d8ZlEmmZ%hbiAqz7btwL7gCEu&gDckjG|H z7zGfz*@&jw!#g##UV9muEH`cu^W)}KvRK{gm>~BcU5vbsTvY_$YOuWuUdJOF(soQQ zlahN)K7za9ytg>*^7z_=le5{p0cM?G+}+T8gA^_a42B%{)I-BpM%(f3{PB1kt4x2R z2QT|}t<1#-%3KXL}EHGOcDy!wMq%3NzeDV5{ zMIxrw6djWFa=NU~XLEh9Cxy@?FIIm=>{7$samAFNheGTfM?&P2ST6tSg!;}HqDN^T zHOkj!C3Bzbkmgqr=z6X%@i^0CNA4L&IZvB?vTx%HqZ;|S+dtM0k_Ey!yo7dK7~y4NY8)55^5Aq z!y{Iy?`Pg+)sAwE#4qdJD(ouTB&Sx4%`vG=1x#^+PuV&9wEgW2wF5pCyj?x=P>sk* zi6#p;<_fy%k(3g;nl5SGv}g2GhVTcfb@D5a42Ts+aUrBmESe4E$B>REH?-6@r4_m` zsk4a$#-{oskUlbJ5fxXfqECMn`V#WNF6I|5MFnN;xI-3I+@Y@ySc>U#$Z5M|O-o+Q z=q*U9g^%y*rP*;CI-5Tw2HXxT; zIeNjK{l+mxIp%5XNbj7%LHvn~vW;Paep!m>w@yCgo1(R$9>`Waj>aXeKC+#kdA9ll zupf5siEn3cyk4>k2@sT`H5BeCTGy?G1?4KtzUd#>nW(P6#!9j0w1#ILQ*3|OdE;uk zb*I91M?`k~d^o+2AD@3>55wG5UTrm+QxnB-FaJlGzG5M#Zjx5W;x;L!xPI1?AT(m? z0b{eU{pxu4k*=9K?I2L>=Lo-x*1Cwfl)OWFKv^SzV(h%9;Y1V@*--JbCEvL`qPW|i zy`*mAb#wias-ja&8mUk8Zh(W zgcTnKjo)*;H2lswuA*~NaGcac=bhno<5Nd5OSpQ|V)k_|(nsuJYicR-!x8=sGIsdn zy(FIBInQs;SD9GtL#7F=KZn3@W7svGm@+!YEmaVN`NV%5Vm!U4q&=X9Npv3feQar7 znLyc7%AecwZ@$rwu%1O!d4J*-_c~e?w|J#O(W$DY_{1s7RpL1><`C zOf)S6(LETAF+SVq;baNk=iL%^GwMHIoO@l)eoLg@gE&JCmMnJ8-!oT@F1dfi%}7L3 z(_DY@HRIE!eU#xV#*@QpV+2*6bMGl2k%7^d(31g&Vlimt5vLHfw(kQyn*5ls&CAs@ zI<*}_fmN89$ONUFEgewA2E|r4oM*Hz>%Nuud?3V&8II0tgMP+feQP19vt+O_3LJnZv@nR+?Hj(>a3Tmgj-l?keN$=tw|i}SWZ>8XEK zx19@R&G`w^$ICmL;UKiPO{}vgu6R=L{WyXeuiEG7kBkeFb7QLO4EMinC(rjI5RJKC zZ^1UIsg9nWPKwHvue*rHFA8@#Ti~f)?GJzJqjvZj6@{8nd z4-Z(fQ;19*5bw4wGuJUxoMsvb9n!8YZH`!QE3PF%Tlywr>GRXw^B$~w0t|XxsguBt zoI2TvOfsCp;!JkI;%`*umsc=J4Z|;l!qgG=K6new({kIrxfC1!1KWp8;tH3ZvIZr$ z&YuY+8v{2tGM7>M2NMx8GBPv@FHB`_XLM*XATc;Mm)@%f76mdeFfucjaX1Die~k47 zP+Z&AfQ#bp!L@M>?(Xgm!J&c1rEzx%!67)oHMl#$El6++4grF@U-mwdeg5~V?k%dC z^$j0$jWHKBxvIu{F$+gCptPd{_&pmlD?dO&NmGuE6~M~M!OY6ajzmqZ2?E;z|074D z)&{z`f*c+A|6w8F0yG7`$s|p|e{Y6Ljt&5MH#-0u2Y`*6pN)r~l@-9w%F6d2Lq`{W zfTXEA$O53m43Kwp0J0tCDX>_Fx~2iG?jHwO!#3*ap~ zKtoOupyC8{_?NNbzYLfF|2`Z58#CK~!~NU)uRtJ&znx9Z%^mHXOdY&H4psn5kR1@9 zBCW^__5?ElOdTx#GBmYwe|3EGH+44!*_oQX8T?(kDL`6G4Pg4#;lKNHHFp6yfnAwh zL3V%j$nsa1w_}!au#j-Hw+A|aU6KB(PZHz;G=DpHFP4A5Tw4c64+o$BkR`~$!t$>+ zEZm$}v>ZUrZa_K7f7`r?kp9T5fM5U@D=RB6J1+p}3;=qXTeJL?e?ZgA3HY~??Jx0L z4}LyQj!po}w>E%&AWPueKO`SlQ+FT$?BWLW^ZBRZe+h|=4PXH>2LsH2Rv-tYKhfXB zK+Ave+xvF`c>+GLzKtInfc3Ax|2`SMO_+tFgPqqO^S@s)i@b`Kii!;5zgzyVQe528 z6X5fngA4GUor@K~f5yhn3E+8q@cZ95s-~cS*YOWuIR{He0N=mLeLJWBlM8*(KyLQ`e^)C9HhtR$F$b%+iGI(< z$;`_6FFD9n8srJIPz8a_t^YMO|B`F{wQqJH2cW8>E9kEq=FN+h_5bMJmdxDt?G|x; z8_mB}z_%Ux-@H-|=8hJBEgL%*H^9`z#ncPw?OneS7r=+@Z7VH+o_`NBfQ8w?5&Y%? zcmsgi2Wt{7jXkv#QzWvfJNdD@d8*R{}3O5Me4uA!wO)L{zGg4 z7MVZ94q%b}LmU7Wxj)4DHobr7trmqp^j3@FA9|}r=?}ftqWp*6YEk(w@xIlf`iI`K ztN)?5>>7XQExYC)ddsf$hu*Sl|Dm_+I{ziUH!}S%e{sCkVruX7wi$n2aR0S=lbZdZ zw+hYwOI&X{b4R zn!CBYH3a_qK7HHa|M1@*NFdM?XpXcr?`SR%e`-?~+HqGSM&$8+cY>W0b@1Zj6@t!= z6oLD`TCjxR{Ut|1^rBeL>|XT!2G$XK>1(sCE-tIB(YkMXG(YZALkFy-**25!6qeA# zk*lH``<_B~x4sF7gM}6DX^Z->$Cc;|Jsy+u`R-iSfs7zcOVw?q^(j1U({E-JsEL?p ze|UUeUJewf&`z{IAJ)#WOE+qElo;(?s;REI2q>ewbbkMT24r3*OfBH4WC$fHwyVw78UF4&@Mt5C#f4%!PRdW3msGfrA zyiVm4Qho;67wsjr1d$_%wOfHPb>N+5{NGrDsKJ#67 z@AlPFepeDIg|}oj!@8ImuGl6FXh)SHok3|nD*L??ls8MEKvqD6wfyj@#vDwl4c zT+h(;>r5?o<(L5)TcBTtlMoIA8;gKdpN?wcU{2SakU{5@Ys`@<%*&EP6{5B7iwdoFBfBpLGx3ZTYAdpXKb9-w4mxLw8M>S6hbHPtZ%j^Ei)|DTr zSx-%tBP1(#Hy9&golnk}9Z&=kSk7;B8*@vulpxw5P(1W4YCs&Pz1q>~dlZaRbX;?2 zzRbz$w z=EnXY+qnE;!4e2nVVGle*&h_lK6W!yENJi;MZKQnFsVZqg3aY7hr3Sm6l$&Yzp6c|MbNK&ip**10cXO8?8 zKP7WNlirKZ35Cx==`v;21}xUo4laiCp7nf`kl7$S8^4690jm2iJN?M0L)|wPWmDRl zhE+{v1$w2zXAC8u5;-R_p~kFD_^pW4tTGDsP2s`H8=r^ye<)fhiy$Q#&Z&mVJF>90 zyu1wZ0Hu5?5QS10K3){@*=k}QS1jk##h_e5ruo#ii~@)zZf-9ASP_jt3>QlF=4%MZ zFhtB)#!JKFu7Oc^+^^wvpVp5S`@g+t3mKwlq2clRk5BAu#$W?hF77xZC}0D*_60~T zal2auY!5+Yt04DRFxsqyxFr zNr4}zz5A1$vn*|4qVPv3_J?b<&)o=h%VH{%c1fY^6@cnO%A^N4ln?DhV;QbJWRCeL zh!Aa$lHiNS1JfM^lE8_8RDQoDj7~llr6ohl<90aRB|FQDg7**76TXXpK0N@k7~2JY zxY}R}e|ilI`wK5X1Lq)Px4rC;#VB>gLR%aK;e+W2W}&=yeT!@|Yn)s54o+b0lt%N4 z<0lq96L?xHdM@nI@&kT@nqX`6FM7!?GM(fptzH1W+Bug#Ye@6#Uk#>>R^H`R>y8AI zX;h+-hP=r6U%#mDH%VQoUNxNSSU2=TEPn6we>RlHz>!-Ju$|Fx$$;Y%SQRpa3& zzLz!7S#(gvTE#)*bm<<(smAX%q>$#(dBx};G3)Q%eKQYsgvhB6=1(=PXF~M~CNqF1 z(8K3F=~M)`?$&)fo1Co1^~bXq>vm>#a=09`Us+MXJ@oO{B1B z&so%ORzV`H^)kMQ@2^7PRNw0m(1ggHf1d`g63bmXWs6;xyt8I#Ed%s1>A=?X{lX-= zxG_dAtijZ>E!AP4m!a8>0oxWcT5AhZl^JVVl_(x9Hn}~+T$y|GiJTfW%y>Q3(g}Rc zW3OF3KD-)#zj)`*Q;9cVh;beC{m`Ydghnl8f%QXI=CqiSVUFdT?w&}Kry*%Hf2j2? zeecXcOZee5D1ViDwB?y@x<^wafgmLPJ2N(_DVrh;bAw0IILI_S?5AvH#mm7hr9{ko z6DuJ?JoqZhHF_k?q&g{?@PNk%JYc4rn-q44qlh$-74i&M`Zrg;@`ShsH1fc!DH%PN zzF?~Pygk6_mrKQ+o`E~;PTnMIfBqvI5)fjR!|hD%>>ZMwLBGV)V&DdULCWT*ELz|9Gp@m2VL0<#kihLR z$&zF;Am*b~)a3U%j^$LeqT+nwl)HY`e6AD+vbM57_e$Y72677kDjejKe^V19JJh)U z#_43Y%Lu)Iyg+NlbmWFSd*k5`FGFUwdxUKWT44SsTGsu22!ql?ETv8`KXF`;RnPk1 z;4f)L``|;DWR7{9^)u@th8uAe3lb(7bbRG)gB|e-#Q9t zxAW5dq0)?RwB_f$vU-_Ie`q@bHLC*Q#_16Ac1P@QGv0p9{B#MQk4&mKZpVDcFw_)C zU4xlySl{~>+c2+;f)%tdq;l zr(`E?yP@4@XM!(~g>_Ro!!PiO24-nCDZqyAU0NI$BhGoPz41WXe+~UaGg9{ZfU{P6 zS(_*}Kj-K63S(7vT;DZQ-rq$d{*hzjwsoqTH=K_5#l}bb5g5OTw}gatF>zuLFVl5;pMbZYPdX^4z_b2$T(>NPB!FJ6IHppUPp zYT5GI+_9U5>B5fSCoUTH^Yf#8+>T~_ok^;fZY4E-)ya0u zw3dA$&h;Mce^KCtEQNgixkvErH4W@ER4$VsCB!rRmR=f@ciR1ccl18a+Os6+%EvQc(3D!6I$Ro#fSI6V5h z0vPMmXRaIFO1%D&Vs5?U_EggHj(lTX$TvE%Ol<)Pe|nq$={$+98}Y-(KL1h~%ota9 zQk#JE(IndnRb(sTZI(0qdLg)Z^vj!gV?!+n(%hzRDc*-`s+-#os6lNp(mZ-PgLp>`(Vj%y* z8#Kgrf0q!m{zWy2PREt!#V5>rB*x3@c!GVkVJlSkf;MrQ=%W`sr6TP{>lze;0iaz<5l*;fN8PZ6sxK(_G$suqx-C zh_yp>!gCsEiK)<0&*<8oC+VHb?s`4|Tl&&;$jSKaAfT%&YL{PF%pt((L6c>Esf9-O zHFkESXf1C=GnPqIGbBvFqXPNmwrrN*%SlbS_i)Ed$p-G^p*L{u76E$AHql(jd+6-i ze{|jhKtUYycV_pYKU| zYXmx`fo8VaZ@HN=n*m`JX`bgHTxdUn#rL66qDYrLD{yv92e*gZLoD?-k?rsZt)`No zsK#_~XiQb7ttzi2K}D;uNX4(q%CK?3f37A3d!PbKYB&+6pJi%2fiOOI(SKt80xP5B1IHi^%`Zh7_BpnlT)12Yw3#p-e>mzp_Ts}q$oO%1Q}sH?1uEOvA!9_Fpl3<)}1O%3z94uREDLDcWhl@ZCdho1hbICS#5ziD{W>cHd`&`ZoYkNe%V1ag#rz*9N1F&pn7L_yb-o$>(*e<2py4eUYJ^t1>_ z=ZP42SJ$vTUi52W3r)Cva5?qq#j@pmsZj=bi1hq~yY# z>{o>&4YrHUv>kaEqgf=bClwY@sF!I~VG69`AepG$bkN>p!!YbS+~aJrZ8d8=iiIf-?op&rJTR}}3j zs(J(^?52hZ&H1y~NN>_Al;lxV>j9^-c)1(GMbEYu7=6myf6(b`>u^ajMN#Z$W|7|U z#Odl;mQ}me4|Z!bS^L<8SW)W!aXKF2pBIrD2CF= zu0l3GDHGs-TZP4HJ>$H{44+(MScB`o7V^>0{;kCw=g6o?{e4a7x#kX#L(^XGJ`#fI z=2zkDS`*w;f6An`RU=(3qTVHAZ<+<$eYXnQAl%In@j~%>Ix-9KoU+uyQ;d((!Bo2~ z6h#l$T7A!51zb5;YHeeb!>u1o0iTiUSQ2;)o;2HwDUWvqDsEnFw?aSb&*@DouBex$1t;ZE}_y9K@&fk*|#{)Pm zv`w*k=dkDCOnQ_G8(D#5EzW&fZ)DSYuuSavX-3#gB)CrQYobaRrz6VGt1T7z(zDyb zoBEgi7?;%98A<3U{_r7QQrp%Sc995t&2*R(rU$iRy?sQv*2ja;dU2@(Sotm*mkxb1 zMBLFDe-({EfmwfNX+HB|%DdtUbMsV!w)ylf3&o*AGuFY+W8Ik189%KD^cT+R)I+f4 zONIc#46`!j$kK631lv9W60|@&Idffm047t`Vgz^od^*imzLb6Hr2_#>470k-_Um`C za9~||D8$e};Q@QP4v8<4f1+}k^CMfKXwQO6K=tk7 z&__Qsv)647)UJFM_LaCuS9*G*ccrT^H=vuc&p)HH*l^Sb7Ow$xllTLl2djpN;}Tka zky&w)P1kYSv3wiQRflGr@|CkR3-!sfom}V!1&EyrvM#(&N=*pzHAT!1X{6S@ca4xI#ptxL(y#8n5IgV+?VX+4T4` zodo&fI;5o8w`&CA@k}_WGEc%K8|vehe`&qrf!nR& z4+SDSa7~&%+YW6?!LN&inyl8x;uYggPFB!YvL^a$FLf*9q?;+j6O1!n&wVAzr($7VQfJnX-Ti(WwR9;%`0f4hjmL_F1Q z9%sx@%UAc2ZEd3Uy993D#gO!Kpxk>mX;DRA9HQ!u^oca#LbH|L?8j0x-qi5i)5Zh2 zprb^?7q(n1=rqj39w(hv%$O&kiPQG4j@}>bXS60LZ+-4j7sFkS3AcOA^5K;B0)Ggk zzYlPe^!m=7NQaOpGIUzNEh?hl2G}g$*;RP%gHnC)MS}>bel{R`oLQ3Fu}ZT zV$HfZu7qzj`jbM^DU5P)1-^X;LJI;Q^9sn@ukx3#$JLLykuNVDeD*HzkX$7w zD>FClt|%UetDu~)sfuTF6p&IL5>0;l5Qpe=5By@8%dQ<|+=htyiQ~etkwf_tmLm3> zXmwpDo2d>5Iis>-fBp|>EzMv({l^!H?5YE;WnPbk<maT=$>>)7^e zxVw*B;1c@u=s;CRMeRkiKC*q~X$!^#AB(Nevs(lOdR`jlfAVC>nAOUs_l*aXn(T}S zTCOFB)9K0$nOH%qM$_&Z>$jz3!-By!$%BXYh6hnH3oOgYD5XCdZqFnyNOkwdJ1lg0 zT0OQ0qLVQ#zn#*8Qm;gwEh0v)vE#k$EG~MMvv-`=^uBJH+=++ge7(83?}0u812%i(#XsFOWv(*j zXWnB*b1?b{h(|J?pGS+Pg$WbRA8DkHhlNwM zGg6{fovf;Rm;HIy1ubWa@k7Ch%}O+M#A2s5e>ML{bM$59iHswEK_zD?`53GLE*FC< zt4h`8uiuwH-tAf)etS6*&~?9m@DX176+qzAcqoFD{Wa&;bT1A1rEoa3HuMypYKOJp zVuchDN~T`l_u(bf>It8{61|c3=@z_}sz}tAT(B@kN#cZlLK#NWS8Jrm3~vD+M2_EQ zfAn=W8y%AmYv0;W-ZgIp)k&TvJo87bCgh_8hN${-F)M}Sfp96WDzOmdcvEPTdivg@jaM0r;e-pdX79@KZuz!vLgBM)h)J21X3E52Qr5vPp zvy8H$e}$CccCqP|5SN*l&O zlG`8@sK{sFg~3oUWAmgwg0>pwx6GAW8Y7j5xJM{X6qQM6j4xknnxO z5RP$drgpne{P4qCYI&IGwsO##f1<+1V{G2Hvv@aI`U5Z{Ell_& zepRG-FCT0yC!==PA!zrzKR#egHycnitVKx%NUga}bAmfqz$KT)NwObue_3Yo=(NY= zeY+`5e) zb4ZTm$ll<<&r0YbciT_me;8iB*ZdxTRzrT^`V#ZFsqCj6J@Z4~G14V@K>MeJ!s%C$V!=Bd2ON?In%7*0{2ne6 z=XvkE&#$3Q(L_S6;x@2Es)I^&@GB#T7NJ0ATook035yxXv)|<`t5F59r!HMvI_w?Y zDhadmjljDRo_^)>f6+YxIKr||#;=x(?;;qD)eGT0|6yUNxNSd~Mjt9BxOt;M=`lP! zfPuHkj8oRD+*bXhQcJ&&O~{F8OLA7OAN6SJb(Z$7EJ>NLhDXH`gf?D^)q)FQM2C*? zn{Tc6n%;fk6i}UHwH1?QyZ~tPw-Tulru}lZe&b`r6@1bHf3jqsTqPea^zel#R-hC1 z`uZSH%p~okOOS-3Xz)lN0JMryO8r5hQW>PM7SW}wCuLR`G!=M7)!Gz?QMZ3ol;$}Z z&o96tR31`c5$@AAj;8UPM0*Jc!sO?5@r1U-XN-4{P}9}&m)qQVCJN4E|9q1&_k;$} zKTBI<=KAtYf5smEgDH27GE^Az=(inq+Un~2C9|gSu_}6GM~PG9mnM~+c{atUP5+>) z`KF*w>f>GW!vzuIVaoROfz(n>!7e#2I0YD6O?1iHyH}&2{)!8t#k9RK6!TzE0mEG+ zd?7L3Ezvlx9qUjdZ3;C6eX#SS8k{nNl;)$=fqxpve~@)(K|adF^ngZhfYjEQq0Ra3 znbhktXZf;dJ@MlNz5&5%30__Bngn>CHATo`fSprFx^J$e1#Vk|&r-pitGQ}cr~q?7 zjKAn?V^^M!y+TIUKwdR*&7}`5o?7!MeSd0uRxSC1x?5>UI}ZOPIi$>08l3@bSN|=6 zhEIi~CkA6sO@CKHs&rN6?RS7ru~JlRf30cG-H4m)V@5N?!;jmj{`p_jtT_6~m3<(i zgVDMZiOT0zC)jqz00LyKY$+s|ieFeaSBZBscx8Q>&yJT(I3vrV)@hr{Uirl0Uf5q~G>*jyrPRz7RYRSgF^96@MjPnTQ38`2frU^$bzIN^%sX)7mL8M( z?1(|EQ-4y%^L~s4hJ#Xg?$QN*B6*EVII=nXK8lZMJz=q;+KCCBN|%^V4xP?-*C4Rm z_$*->f4W*aWF?uiNPx51^KR9O-Vt6YU=ONAMJ+9oQNN6#s@ z<5W zl6f#fE*dtB>KH`O4vTq2x`o?w9hNb`!G9`>h}vW|v zq{~SHDHd%S1IW~#(*2p~QqhxUG3Z^(&efno#-6?r{IPNF6oskj)&no2IHGu?l7Kr- zqt4IdWahH^uAONeS$_V>L$*U6YL~h=F%WfpO(PiEv-3laY1jpit&Sm9qyD~5gMV>@ zTsn3>(PoNx#=GUFIS96fo_#1`25g*!#kuIJP4D2a2`V}ocSue=q7HaH)9)RxU1XMb zhIP|e{%}huC9uk_FHZulTLa%1wZ44Woy6@eZ1p5o5;W_p7L_Ur4@<&_IR!S0M$NVr zEU<~PP>C%)s$zRbFHCHB!m${)T7Q*5&Z8Z5W_?mPS9U_sqYlgm>91Tj(&3IvBd>AI zHohkaD!VB0!2gj+Kxlx2W>kz<*?6^KFxP_K*q~mV7Q5V{VRrE@`}F-$*Vop#!A&`3 zN}36IQhL--dT-V59_I}P5?3j%`rNHok%uz9q=yd9P-G9hB{|(Ek zfMypybmraQckN!I6vD*X$(3p+<-uA2)WRTgPb-1fZ@D9(NAM~U(wZDn@RD$@M5MaE z&k{=EEZcKFs7PzXNn%zCHOx#p>EtChi)~Kuzr3u&?43k z&r&f(ykg14J4E_H+z!pkSbv)5BOJ|C0+^fi2+qaz4j#?U-kH1W1ylF@sy6LIHw<{m z)k&yqr6b^+#-de8_I|mvJpW`^FbQ9vr}VydFf+QjlF-&#Et(cCS70oVrY*C~4*6#> zeH1P!1>H|FLH;-&rm(rv5mnEGvhu($h)wyz$xHpz&uyd{!3Y+5B7ZZdJ%I=mau$8& z$SIQ|5Td1ibCP=aaGjgm&zIXb{;|CaZIq@zuuIXnA9=sY_Y5AKLG!_3?GG|*d zZmBj9Rb$E(nJo8`8a0)t{t8lP6f`By4-7G?;5hRT9}B8Yo6Kfic9?_jd*s;=3EkaM zvG-vG8sQa9xmg>2`+rG-mlLO)H^tQmvn0nK&Xcdk$*efzj<+I`H~H&zLsUeaSyO$D zgD0wBK&=F$d&t~G22>Id2})A;WPC>{%#wGUUqV%OzTMqBaCNWtQlPik!(WV(j_A@z zP;ZVCe~8~+MnPwnhVp{~tRJ*tyjJ%~9sHEF&Q_4{7_k1(uz%68+7L{B47a;(d&da# z40TWvYxLQOjG)n>yzSABc|67Ti+kMfzAH5)OIOz7x@bjWrFX1I6Dv)h@BoWN02H&$JPrJ}3r<=*|~Yb95b zC`xvj!`yR&s((rj*{6ZT{klg%NY?(l&Bk_S9HIi6_yt*jHeNWkugO1;tlq0LRAuaO zXsubPC4RQUAd~+rgOkp_f#X^!jca1Ah>XIfjY5<6(|A}ejLO8#P!e7m4TDKFFvjMW zb2<6}Ho)#gNw)n*ATPx;(&3_N30or)H)?n>hp{qB#(yCr?cuo+*8=1YddJp))X`@+ zc;xj@5Vlo7xI&?Ndznbnh?3?^qSMb-B=M~JM9bAglTKX~Owu0xp4dFH@4b(IOGp&o zOpL>0hbq6EQMrl^{GNA|yUtm)&?9{z)uWWgfN;4wd!Gz#0U0RyQX4-|>t$kWBQNZ8 zF^iy-Yk#xvKObO1^SBd(E;N9S;tpqv^|M&{C$5R*E83zTu)^0i!@5q~Pog*2c50HF zSE!j^=x8o-y)f_71Zw{?-50uxe5*%^sET(W%oL2*AGcaC7IK`DsKbRdKOt1gm-^iM z@|2Pr1cM6129!t11KQ}Hn15>~)gi-Am_U+5cYh*7pF=YrUyi`82FV52TK;~}bp%rq z5u2**frAC{;K=C_-Ee6rOh58|%Xn1rB!bdL=5=tndy#QF3Xyz_82gR;4lUm^w5Jxu z{^{&Rs0v$w<4Y?el;%=utP7ty1r7#lDTDXy(-wE#&W|tcWwf~-3E{0&6?Ov_bXiCi zkADGLy?Zi7#wSP)R`sdhAxtB-Sh5&h(7kg^exB)~;yJzIsh6Tg{$M{xE|S;!iQ{KN zR_{__9=Q1E+U*A1J;(2Su)$QIrot{Y6v6eQV(Gg6d_?KDh_Su27M8q5rk?}_t+gFN z)SYK2_c;Q}@x!HphR^4aX%kE}jiM;3Vt-d$;BoX z?6;PU@Ynhmb3w3ENwV5#@orcO4~Tgi^&;5Qe!3&7gPyvw<p#MxYR35E*g5J$9W#B%K0!v6YCDnPP3=Pr8(-F(H&kAPj6 zlD7GAehw+znkS=su2eK}d-h5%Jb$l#f}_slAgTwD9nP{0?+ko_2z;-)P2Y@yGrkqu z*~?><@x)?46AJpJobdK*R>OU;(_m13!rJyYi;JBfA7QLlI>3i#|;oJgsozt%}1VQ(ZZSL@pHC#8l#Cy=%}TBp-hnpbI=8a(;DndNks+1)eCSfV0rkp7ZaI%jwAM_ zVEew%(dhKq^05fh1hL<`KYwbz2zgE!E^{YpcXx7@-U;M8>5lA3`C`+etgynYN``ld zTf|BLEt09~qh-Ft_3a2dko#T#teSDIEQ~L*>=*$qr1r$%}3c05%hTTB?!LMMwhROEaXHYok%jL0cI%j*!YmfjRqIL1wDEd~q(ktceu|naC`t)DB;#!cz z)$wHZu$TF+i8+87 zZ>wLuG5^Aors;2e$Z0!*y?Q5op1~lvJ5`9+*Kmy(aEOLEr+@Zamy(-Xc&TFCyS<1> zgkTc(K6(I8{mRwe5<>NON5i4`0m`sV@hEX3L?ok#ATt+T9A&tbd5`y%Vnyh&(5wU%1k`7A#Ke>Yd09+g@J5+H>_K@zn)Y|a9yRyY-y;^I&IJ++`GMOOlx%`N~so=m%>HRWc^gx1vH62T**_nD)@%kdqZ>+rmwI48Y#m@P zj8^-VXtG-CO5S32b&9CI3H!zK6CkiYsZ7r8-AwHnZdq>gYfm<_XR|O4|4;9DkjPP3{O>3Ve`?ydDVI1OyX!lyc0} zzDRcDt&(Sa(&e*ezvtX|MG0F!V~kx2D}ZD*1Jlz}JuQt!>>M0lX-P&Qs&`69QXH1+ zb3xM`V@&(~VFTHKS}_$Ht`ZcN>J?m16EM#OmoH!6v4bBIQgG)Thb`y#syj>kdrV!N zlYcuDW>lbMq|HBkHe&h%VJonF^t8hO`e%zT5-NtLyy#>w@^}ggh5O3Lw{cAndg-qx zjXdcm>E!1Lx2uMA3o0>FBR$;K2iUdrv;@Cd`@Czw7A%6SX#a6ZzxmNRLrbY@)0ca;&;)RW_r~hl$yRw&MM|Bp10W~v&O)?O+KdXSZ|ee+k5`r z_x?G8PTl8tX-FhMJzZqoB$aCGP=6p$F1y)zpPcoCrU-{C3(Akq4@qgGX}8_ny}?F|W4uU88j9Mc$j z#_2=vY?(mZQmMgls*E}u2G#f5^z;Wc7Bgpc!o#%IZU^2V`ZC8z0EBG?DTf(x!M*q? zFY3FG?GewNJ;Xo6c2(%Yt$zxC6+e++<9G~lnt0KkLiBEM>F(DLWCp_5Nzqhrm$jp2QH zzZA#iNjrJ6t2Ds-p*C&UVu!$XuaUC!iyEv;%@Gq9cB0(j4?rr8*MA}6qWLXWg!|Id zReX^#rJ5D=Xb=Rx=nm1Qn5;Jg{ppD6ieqdei73My?d zI}gm%V#(XEsyggh)ifRvxv{BwfMY))-0df-m%OAG-0+|ttvR!&!-?_l7vj$A$&F%B z73u>*BkVRuA!m|wwtu47yq5));Yh=S=`*W4ocM^mAtr56LM~r1zDr5`Ch_J+zu*Wu zrdGeNkZ{hV9kheQ)6_hrK`u` znf22zxKUq7hV?X}9GEKd!&)q!aL(E*Jdc2=J;8vgw8s8c;(zy85HZ6&8@%Uyl0;7T zd$HYR!&cFO&3>zSBX6d1Zb)udB&1(e{#VHv^_s?6crc0x?F{<6b`tRSOPj3?8Jv9) zjkWT7lEXFxU8i%S2WJ`-S%r}`2joBn+cK5mwdgIMPYMpvDzM2k`g5!5!eKxLhgCtZ z{7M>4DL(r_fPV#B4Xio7Y&dIG0_rEe+c75dS;*2<4OQ{HVLIKt4mkm%)w+-cgdM65 zZ{f7E8sgg1j~D9=8bahqmv=yhw8jiS&m15#iR(x+X?lK0#eVVAgG1ZPb3E}W5BhJ$ zBnFDxD?C}tm5|G`hi_=o%uGwO&$>=Gcf-m-x^0QfXn!d&NIC~RMAb*T{bp7DGb=rx z$%<-9L+L={$eO;S_u;kfnzjyRFg4-4{g%}RCcl&#v7rq7CQ8tU9G@Yv7oez#yit_C zNzPrULv6XUk_ex?H{Ze6ANX*4_y7$-qI5h#_^l_wE;14W<>wJ$XT%Mqd%yeIjU93K zg&1!t<$tqCe8y;xt-w@%2L^qab9FTg<-V|H0jqz*Ja)1Ba){_^X}tKxwGI({ZHK)V zuBf=HqC;xq_g-@}eJ9i-y;d9{<#XedqDg)?gZ);Fbe>L$0rda~>mwyZ5CsEY)Rx^> zOf{+>%}md-9p{;c{ycW&Stg6X4`AIQo|`AC2!Br9l6eLyu{+$z0hQwkRIZG=gfgZQ zBZjLK5lxSvce!V6`x%H38CO%pyKvoH@mY^5>(X1+$!2@dcGhUyuDRJxk9d5qqvR}e zLxG<_W6+nQZ4jGi!j?XPr+e$T40_5Cm*lvHlCHXEN>RX2+%dbamww>2a!Na)5r9_C z&VTLjCdF$Q|5N+JrMu=uah8W7)GICb|Z9_<10U8V*AC$TRa^Gzwn1 z!&O2OFQf_Z$RhlHEe5Lr|e2=b0);@d*E8c{#F`R(jhD{ zsg#f@yk-r6n&ccpDYb_`>&Q&-5}^hI@_%=jwSLw~ ze5_XL_{Vh|`pIPqiLO-C1Cf3-VE*8;^7kn>13n)=SA6Vrny_Z1cINzQXe=pMcsC~s zg3mwJQ3uVv+1Vn9N?sggYD@%z3NAbAi8Z3bHg>dR9kF{bZIdZj>S&)R0wl&%cC zVH}aish`f8gIB*!A7Dwo7F{^TF>`dLN_yL@hZj<7tV0QRhO*=gwWTp;U6ol)b-Hhu zdEC?7d~Syfyx&+Y%y1*r{AvN+m6o|EbTEKic0ExcDg99*p2><`532=L6!dtvaCbaE zxLNd}*=IDXs!6S$gY5pMj0+l_eI)$1qWPN2JcXX8P+#r)Pol6$>@`1#4`--RZoQ(C zl!W0Q#r6b640Js;i+_dfzSlX4#fV;0sp9o9E#b4INxSN>xU|b4d5{DKe#QK_e+ZxJ z98d|xli|LtM7jng*#{Ba`V>VFHtof%cd3a)b-D{3C93P%xEKXDZvR7~{Ae_P&f4TD zx+)9I8djO;Hql)9Eoq8STFTZ65r0Ywq1y`k33Z&`6{Q2^ z;QA6iNJ13JggIBXi~~U+mM1zqn8laC+=Zm_!8^CMp)+%!%74`3Zg01}KK_NsUDu`Q zyxtri<9TR?smk(Dv>etIVLk+m&wVvF@#5*M1M^X!G0~^VVNe!k!oy1#onT9tmUJb_ z`v>7oW6n=g`+tb5bjkBIcD2YBedDJz^CjxVj(T70qU6u%#UhA}4-pZ7O4iPwzZ)Dxxg)3& zY?qFL7k^;M0wKdqO%baLPsp=|Zu4W!6G^L1gGgjDDoKRci(&+=DkIjAfu%W5Q}IWz zQqp*jn?{=NJbxM0L_(VRuQS&S&|7JyazpMxjyS)4jMjr7yu!Z*e%mypx&(%w5Rs&j z2CpMInKYnmp)a~6;&K?sWeR)3qL)b9(&e@OX)bLS_@jnjJpFJUC> zB6^5BPB6`%N#zBIX0ip{BWTBF5Nz#2*S-7xbjDLALcy=T4H$d5U%R>h5gAllwZFF{JP56gIaiOfUnn0 zv1LtjnXrRory}N6>kmvDCHG%PJC_;&5r5S1-Z;0RVbfoD67b(wVyD+-jQcXUT#7sQ z8I8P_CX+bt;Fk|Hn_MPqj;*O$Q$NXbjm$Owy4E?w&l|icRh{BilGZwwE|4B)OsoMZ zet&(x?u6dyn=h4P->`_iIVhuX|PEbpRuP}tR1FTDp7O`-k+_0%o;&;KX041Y}X z_>DnV*erOYv;%$ct~M&eZ6a=Dp@GBF8Y7REYkpN1lh!0L^PdcS-BKOX*HH{ke;k8! zzfVKsB)rhWwo7|Zu@?~6ejEeu<2nJ61@u7Sl5;9l==q5F2P+iEzhJY|jc7>N-rk^G zfS~-)(Y7kNR*q9&{}S0~%@~i1&wscdokgm^a{Hc`S2b)-8%N;uaN20lT3{an#!B&n zippmK!@#B<%nW?RHQ4R7rzm6B0x2IC&n5}!8xlS&*b1$RpN?zQPdwGB9+9JOd$lR@ znK?BFUy_$4=&+~0a8|U?K#k87EHrxrF@qU;dm;yeXS_*e0lhEE8O@?!0s}@?j!a0^UArO!>4|xr#OXD)ZZ)=>r6-Oy;>F+ecIgh z1KoC&d$0Dx~aqhWHz;(FDVzwa= zZSrVVD0KfZuXsi)W%kK>n19)K2LYiv-H&7Jo;fXwm7a#Qk>VQ|Z{jYtT9#ufCc0~_Z4`IIYa+$j6`s|rB zcWI5!$F(1S(#rk$7+>brgI-5tz{u)axm!iOmNOX;#C^wd3!_}37&?wyx@jiB3ZsJb z_gIKQ^>J(A;K`g%Tyw&l4H#wD+^O7*7o;}_r2&Xyx7<~6yMGt=rB>;WoFD!QY%604_Psr;GAJ_FNq4ekX73%>Ed-8`J$ooHAKlMKkmzbh;n!&c)s zd-=*M#lEEmcM_(4fK1o0`WJVh6Z-T@mpK+#9aa{n`0~gERiCj>xaSepd!}h`d}Z*U zcRdj&nHo%Soqw&6d0p;8Tvxa&%ZJ*GlUW9=ABSFZ6yT3!y0=wSH4Jmx;_DJ>E*6LT zi`-C4OJYJK8=c=Eb6dd3O|@vC9ht6~=OeS*``&Tq12*&y_$aIeT z1G9&MP!pLVv!MGJ>x=y9qr+6ov3}(Y$;5m5t;vD2=zkY@&w!8WJtgZ>D!CL{Y@#23 zeP7XD^%~S-YsR`A<98mET?F3!=5PWGJ^f8fB1i!Hi&>AlJ zQ9B>gPwAE;)EOo#8^K@V0yl5^JicFvO@HRlYk8OGb);=D=8D=_giwZ0kzb@|k$iei znmr#Yrp$gK0H#O<#_;%*rvRS~2eaR&1|su-dh@Mhn>$?V0!lGd1rfbmRapj^^9nBw zSf|l-|IPkApX9GrHCbz{&$N9>jg~H_e23g)>tTFbfd!;oMRJz1%l2flw-PLRqqoSS|LmOiOuEy*zV()g^?z&|J8ofp^_Ln;)s z;j~=V#IqzH$eIjmSLQ325ZqWiSbyIcY7*ZnsNG3&!Hlio@`3|sujP4$=B*$O)OCp3 zXFH0>rPbTxec7Nqs{(H45=_ZENmB;R=B)6C*6kZ2_|KFL?CH4t|t^9r}zp#z&u&e?)m#`)@^Q+G=HHW_A7k^JL_O(MS z9S<-DotklYln!wPVw7Nqb?hF$aivmIPs%)#emaQYE$CuuzU5a>N{Nwug4O$&)?M;P zPOi9ps((N)P1bh;wRE4{VcA*)2sybr6RzX8@a{HYz0?s5G+~?Jj0P(uJbS=9hZPs) zZEo zh7;`0sWrbG!TIA<5MLmMN*@}}jo)4avJ45&R$ivSC7Cir0Wc>D4bVZ2N?RJ*ZvkBxlOVT7UE#0sjBC$OJEo zyzNxo$v3=@%d#Js@k<9JtLCQ+8^MsX&p-9?pR`uOD;Je>$XRfz>za)y{K>Qg-ROaU zqw2wi-9E*cpr|S%0_&#b9|!Z!BhiXHIod zgWVdp0)D@A+UIXXK!4@SP+aR82{hx=#t^N-%p{b@|3bNXSF?rkkh$4Hr$ikjM(=?i zxizWMxt*Mx3S!CMEEEk}LSl1Tf)|-mZ7FLbSNO@&f6O$FDGAvwG!Igou`{);rQyN+ z_gR93m1aT@2t825F^>f#;Ct*v(ejv;B-#lkn}x6f)#?LMynpah20CRWr3bR^<|Ycy zC!ON-itOuY?YFO?41A5)zja)LwyQ2>RI3G`kh23Il5 z&=pho()foKd?4VUbx5hfV-9AdUA!CaD!KEzoMcR9sJeqbuC6W9+g93{(J$i~{AVVy zxjX)=2{}|-_kTjX1oVqOF_nEa$VXY(PDqzp-E<%_kguqhg}bk-ALK}+Ah6lzjrnr( z!K%5nWr`^yV5Z?D@i9NAO+g9`fFyN?Yy|l?DGazIkVU)1I z@*LECVe+f)PG_?v-|PN=IxJKKK2;J4ho9I zRRBJL4o}WzeRI6~Ms9O4I+9R7QT2CTNL4A@C32-J(-qnk&bwN$Y&MOQ!r8;PS>*r% zf*SoW^8wSo>bCi$gT1ct(vFUfr5nKruTETz8$F|@amEe(4tS%BeiYq-5LFHlaqo*k z<=&VZEPwZD+l~M6)W5U?hKE=%gC~`wnaM+^qfIvr{#Tip2mTEcXKlS{@`smIEV0YMX+iI2=eXU7(@I4r#+!9vI;Va4KlEald-0#_&G zIqLCC<3^I!F-)P~bT0rTCnjxoB0Cb0parxzr+=>$ftX*xeBVu#uf{A~CJ5aJMGF~v zcogP$>NkEZ(Fl2-gm!)RhJZ9WOajKOxFU-QM0KC9yW>7M1td>rkiH;jY@Ajh7B4I6 z&$D`dKm#eWJp5w9wZ~3wIZ&Lg*mC?}E20#ShQe;bwNRzwL+@nCq8}pa`G=~FmoluH znSXM_wLbpo_kbt6$?*_YEjqY*AL6?)my1u!vgN;01{bV3Kr6VRjZHBJyTazS!ya?SG)xo@ z%+M153T19&b98cLVQmU!Ze(v_Y6>?uG?!8O2NM%AH83&?FHB`_XLM*XATc*KH8_{x zs|OVXIWaLemvJ}-Cx49h1yCH?(gh0R?oP15-Q6X)yE_ap1b0oa0KwgZ2X~j??he6% z1t-DX-kfvKJ<0w4zv{iB3TCyg?%lojKuxZy!7Og+XbzNibZ}#4XJHcnyj9ZFWM>1g zv2n7nv2h?#Q)`0U?12B3BT;JuU0gwq4g!BzymbMZxq)SpW`AyALnTKCfV{gMfSnV- z&LhCiE5ODE;9z6p|F5B=ivU2<%mZWzP+|edJ30Vek*MD~I(fT*tZm%DdH(wmKxaV@ zVCU!OW%}D4AZ`zI0a=(i0F=z!Y=HLQj232g01ZbA5YWy0e?rg+*|@nm39zzydU~>$ z*}Jkhx>$?QGk*a*L2fnxb)YNI#RF&w_^V`qvY9>bpVC;6r~#TbAlLsWG#stmJk4By z0Im?5z!m;epR0un$jQx>#T8`tSB@M^6Xe{~{}pgQeA9Wmvj9v1&PhoZW$PlKc#((mYCG0^Hi7<~RNATNLc8+iQK0c?N${QJoWJYklO4tCzZ&Hp}P zR&hyrX%#ufe@gyWDIwwL1@L9&0x)xMass$Hd4B;s{5$~v|Ba$*2KtA^AD(g!R*nGv z|BwZD>A!?M{+R%}f7XK@@V~K?9l>)81kn8+b3-<6HVg1C_W#er{*@Yu@c*MXvj^FE|6>83TX#3`4k$T-H^Jfmn(6@mv$#q?OOU($|JBO5 znSX(ILEORG?%zg&T%|!?KucASn}yBav+*CX)?a&O2XX+aI=X`Xx>*3s>}+iRM+aUm zi+AAL!xcP|f2)At-TB{)QVtf5mVYf72R9GE%*Dmb8;K1(NgUkV0AF_SHd+F`{vKig zD~p4p8`uQ^uFfA|<>-R+*OBsYgXiWi(SLss4}ewTH{u1bzWt5(0IZU~5kG)c>R-gm z24I!`jo1OKGQSZQfK~oC0%uVCjlda{ej{)O<=+ULLFHe>2hO1S8*uFJk8a6E$haOy? z<8LxxXUG3hgNt$ctp*2k0zY^T|8qKA?Egsr_iVVptewC&l;iJibAcP-WasYsM|`lv z`FFGbLheAotdl6A0fbef3pW)wXAM7F2Fy|6N)`fh%tA7zE@?_qf6rQ8r)k&jg_FZ=AaDPC9Po}TV3fgeF5lsqPrGbA$#^cZf$@MYfDsXfAo7YiC3T5ZAbbxjW;+|cr0E&K!QljtSPqV_w4Ma z9rzBq6|zg7I@j5q4^1^6`_8jp#;bmyY;`!u8RMh&dC5s|iW*P12PC7;8b{c)7wh_Ckey8Q}mGfSyzefw;uPMqev_PjJY z2Fs;>xuYS$7bd5pV_+9%H1Sf8#9!B9a~UF5^S7HrAR6}28-I)tu1IUGvN)gby>R`^ z8`n?cQ+V|8^dxTdwo95SlH)eVY9_O_i2tdNTLFcZ9>J8zn^U+rQl?NVa7QU7ziL?} zHq$PB0y&`+wtw4WQkM5)WjEqu}PZTXgTCwu7ic!8`P*Da4F~nHcs!Xn!}Y1Dzm8Z zB%Ve_FRG<`MXlA-6-w)kK25|4Y0RkxwRFFt$5wn1zAoD~7$@#aTsEqqe4jS1Ki8xG z&3n{qxH_sXtad0+hepyMe(n$=lLo7Cw{1a?f}gnkqkpRZy?YXSI?>qEYz&G~1~_R{@t z)=$exK7Ti)b`h;Cq>pgMNh`dU!wdbwD~6}ZE54@msG^D%!6F#59v3A$$_NV8t<4tn zuS>Yg#@VwS(9^w^T5>NZ`&z&~)b6wzCn5K>Xe%#51AzBbn#=2?))CFcW&tUiKEJ?q z=5BC;>Z#Y!my(ga`dPIwP*47KdG6@t_0Fve47t;tT zfa*t-kC~qY0x>9pa?eZYV@mzBW=;o7kE5*c;qJ?^B^wLf?l9&pQ+tbZg)1_%xF;^Z zeqbbOD*R>3+e3W}5_?U>+r}Qoq$h-&2_MsBg_w2@f1|ZFEqz#f8`sL7&P77z3 zSoyYL4X@9tna|M-JMkg?r)HMT^s~#rNGY!%J=(V6vfCr>aA}JV!Vb2b9NQfWaElf= z2XUV4DA7-M;|H7p<~edkVObG9-YsB<0!g%ZcP=(hNu z*v2ryJ*mI+i*n$QkC$I*ljABw(I8cPXQb0U^_gvD>I?@y-c1Q|uV)pi+KbN*Gk@l4 zFOB-Fj}|0kTHuj|5twosYoa0LsCppLys4?#={c&1H@a0<&n$cUbHHTN?=<(2r) zEk@BHiDRn$wITW=v4lE-Vmck#JtpWdh!=?gnKBeYW|xF<&Aia$1w1`a01SGF1k=0E z&?Wmc7#}9fr6@Ap&)b*vHXlk&Pk%))XN*lrAsXnCE9Yq8?9(4ekQHdr6}IP4X+?4{ zv!pgvm6-&T4tygtIAu80(ZqNRb2dep&lZgFbfaklRu8T}NaL|M+obiu-s%zMTs+zZ znU$4zjxQ@$X*o?4n^sm{ZD%O)*TVfU!I8I~HTNXnPDL62>e?STsyUkkQGY4f4Q(DY zBB0|;Zu9{(nCID^X!?3kSPBClMZ4Q;x@lt!*NdX587o)rAo~eu_oFEvvtuIRBIdR) z*A#30W^pGlE#PKp9Y9+`>J+|#MkfqYuE*ZGcT@f;SF3BLqJ;HC2hkHg~!0DI%kv=_u9!fj&FZFty*K4^5?mx8>ng$h*s5dQzspdmXDGU zswzB*KBUvYn*jZ9>pyO+d z@7hn(I-JB|OV1=CzoigV8TnTq*5<4OS=vs0s*q315b58PCVvb0q6;@;Vh8mx#=~PE z3lW5PH+-uVoxWrC-D=ZQV`FU{yc=EP3=TO-*K{m=dh6iNCDN?_>M~s8+wOwjqox1K zFIASaI4@JRPYC^Il&rKxb^-J&1*X)R^y{zmK8tQQiW)@?m3J5u!NVD zjJebki>Hp@{4WxU@7^gi`oi;DAy`{KF+tiL`jyjsB?_1&KvNnvVKD=l!ObSpm>1WtYfYG;UvnViGqQ@^u7WulwWaudGGCGgDq)6U6?~JDNPmvs#unaq=7!O%Bs35St;+`K)4{$$ z#Y2VL?Ee_3BecP8y9?Up<=KCx>g1{gLe@0Y1NKW5cbPmU^EBPn4jfwr)-Qz5{Ye_7 z8pcZKb~HxECfHaKyQ*W@ztbaIKH^0NQ(lhB!A%IZ2dEn1V|9!)p10zuLw3C+KOH2u z$$vpJB~c8EW`0f#yxB3A$eEfPep=%5TU(Wx%KpB7YADzyi-P1ZxCLjy5x!CFMW#Cm zhm*@rE-J8zyM@`BvWtP|MoHSYi6=_kI&+(B9TK6@1vK!<$N^43eDnv5ZrISLK1iOR zK-)4ud_)N8KT!$c@woPJFN_m24@Qg}Kz{^zHCPO;o4TxG(YSQ)6QiQZogBeo#S+G@ zw3>)uA_StcYaD4DJO%V=6!u1fQr z*IA~mwRyH|q9ptO18hDbE5hq4DZOV@n7%(sxrJ{C>w8O4u5+%oScK78wK$p!+BI3h{r zEz60qAR~5+%w2tGyM_gJr_6N38-L2uMYsqUk)EN7xZO8|A_MYMBGkzur9{WV^l2`l~*jB)RqD<`>I*)hF{U4bQ0)uWqs68~To#y05<) z;D!^CH5Ok{AGUJ#gvUO8&13kmMJIp7SCXe_{C3a@i$MC+&*e zS9h7H$ZfRp5@t@Z`o7&_ zKITH}`wUDI&Is-NhJRm#xo{0cTSUXIp4!sHKjf@~JU8Jv3dot$sTC zboU`g)3S<)BVv1Dlf0HHeFb}?A1e(h8Ei zd}PN1gzbc8jeejdyfo7v1la8pm&{e%@ZMA8O_~R2&Wfc_s1YT3I!)SU;wCQG=(69| z3({VfOv5%97dfK~P0^iRet78NmgzzKnYD_#_ zv{hE|@o{NSik*+qZU|>L+OO%uO|nu zU=Vaa5YE&&cPpNDs;)Xt2Cw=KBy_8z3AvkC7>BxfY#Q!1sS=)`I+$TM9($wd%V5Z+ zi9A6{JuI=Zo2#6QQYQ+_mXzv5`1eaP;RX3iGu}FGGTN}V$b8C7hgp4-6**`-LVK>s>zKnFY!@jP#Y^v)mT#BwPYaMw^RO! z|1H)W+jb%qncwezcbcdM)Ie57+)IYecc z{Pu0p5o~AMvh#F$q|un;i~wCgqQAL1#_Xm%ajc;kl4+Ox?e%kLyMU{+kA|>}Y?Q{8n-HA^rPT9O|rMbAv8S716%) z0YTgxb+O5kLLToa0h zE9QuVoMv;>!o0@?)-v#*F^>#Uz5Shs9hF}t==Oxo+3BnUNzDsl#>H_*_CYPYKHbCT zd_Qv__zW1ne9jeXS~4^aEH%GPYZEQkrmZ&fI>RMJ>1EiOz(X9$gKU2vskHm{?NxvN zr>dlDwpj|73?kReNsYHvdr-o?v)-d9j|TIP^TtBt%th=}HQb6x-<2>i<(NfY@D`Hf zenGH@SyERH-*=Vw0XqUi*WGtz~U6a;vunbl5pHkOzPd8VE zZVsyS$k?|KHL+HD$F1+!@3@ZhgSdYrr#^VrDmN7RJ^sc|nO>3kr*DnN6X8rb7`YW` zJJ#<;jG(0Wu$x7p({jF3*hpS5lFZ0%59|ueqZ_rre&xj8=$a3z6r7HOU2MLeE09U%Yy! zGl?kXSvl=PDf%9y#>1kc+@iK&Km4i@Xi?L8I5OJ6WJABv*e`GvdY!^{%q2!?-saULF&#@ZvA zZfN`!_Z-|m7kwc3!nK;CZ7{yc$dYgJ(fp`{%I&5+j{S(G$$w;VXs*X%=Be8EN4bzt zrpL+cTFVCBxXR3_*;;r{OvAJgDhpo~8jLa1^i{?n$m-+Ieq5TwQq^v+mIfL*EP{;m*W-&P&1c5*hv1A2gvA&B zl5op3zddFlM@CJcsS^(udG**waKR>Wr(+qwXm3 zCel~dTi*mU2ZkhsFGk?rd!s33@#h&Y@*c?ZPov>hNL{hQ5 zPIS0&_^E%bOIYj_IA!5nuwY{=d(%a+-?Y7d3i<`muVPYc25!%tm_M(`gkCol)1TqZ zd#CGZ9Ud-Vp$reZg!=A5e^o%UzrrfS3wGn@!(0vB-*WD0er2lx^`@ie9V^WA^Ls(Q(G9G`|zvx1}4U$s5>`)+wg zzX!Isib%*=+qM=w%w~$bnW%1I6K|Xjq9jW5{)LeHJhRc1q7+9HHsBqb%4ZxCSmjS% zd@ue~hzNwowwc?L(`-bl6qFhSa4w6F1{K|Fj#A$2xMDD0KC*HF0E_g-_4jj6g-5 zM}e!>$;Q%QPGParE^J7ymaq;1ZM4qaGGu?E!PCR5lAniJ@j@Lh4c++BysIX9(Rlw+ zVjz8fue0jKH$Ngi=Io^g&2eesi*|UP=39UMH~mu8wdN_?gdE7(TC>9O=YT5tQBQgt zep#(o4R$|-3=`tarWwB`WQtdm3VeT? zh&}nJh7{bSw>YCxt+E}^KgnP`aOW|f-6}nabfq8^C%-0IsV;{*M0KfnTXO^!wne9r z=op{YOCye?1(z9pj9BtwXd9jDw=BacNW4kOkxu`v_#RVNJf`N8&JT06B*RbE1#{ob2@1Zq(ERTrk_;^|#>q&YUT|AC zSbo6ymGIRoQRIlxZe~92Emh!p-96nAaTqYz_>Qm*#-JJ9NIX1Y(8$SL)U@ZD!aXYz zH8>CmjS)uC6A&+ZWeF*gc9?&IUy53_Dx}Y8Ng8be>GJEN=#fTyirbmPN8g~NS*cPBaqMF`>kUAc?hkyij-WM@smT` z5sS{(2B!iT1|HQ}Eoc;iZ`{iAdJVUJb~RmT{(}dTWI_hI$}F>RxI$KHE+2DBxjVHQ zWOxw-ZQ;cN-i|%!9OHj}$0@LX9?XU!F!By~Qe?_*jNrSZEBhe!@SYagY%gQ_`~(}3 zje~WjqrggW&OV5bOt1VI(cyBi)349?TYkssmtRUB$`yH=s)TP1+PX#!-v>qRSraz@ zAPS`uZ>sY4t1z{pS6>)n=kw85+aZU6tcUM{W^cDE4kayr=-7XsSdj~*bc>}Sl&~%( z-@eB{TNB|D97+Aopjh0!)8Jifw?NxFay!`rt>tjH@8QY7MD^&~3N-=Fx+zQao8bL2F`2XlFH+ z$=G&{I=%#?T_1>nZ;oMo{+y3HmZrQJLEGbKHSX!gr-=O9?3a@1q32UR^jV!|bbAE* z89K9-El@AL1=>= zEebD#5U$EeSQt<>&=sJ{|8)JBi#V~!N?PC3hl78!#rL|p_VLNHPQZ}WHScD_C+%KI zQd0NCN9Lo$dbNcS=;Z{^arNTA+1010u>{fCCPD!Awy)|ne@yFj5N3?GxP!|El7 z#e$?R%UrTtlwCkkns?>~Vv@FYk9K1kG&eBD*Qs%2yt8vV8KR%3z z>LPzK5n}VGb(GPiC@L_K>;&NUq(`MQ%*Kgn>iP$o6Ed=zrBi=EoZFy2TAP|`qTpLj zLz@y6|0;`ys7hEAvLWEDCHh3(^$YQ2RADQY`+CbW7XCniIq!O$a71n06K6v?w7a6^ zaYeMMgi6evjW#WFLib5mek@n8rLMvMt?YlBxJgZ}TXG3_e#nR~2^;JN!!W~W*CUe+ z`>YYi%N(P}XTTeb8f5KXQ%727Sx}#!sZeK{UTS6ER}Jb>e_7ktzY;2kfZ()1+4zQi zN!FamfEo1|IfJ|6Ac{#mB`#IkUQzECyJVExXZ+n8gIjyx#?5O%2gM%tRxpv{h-H6K zK#o9K(rHJ1g1zDPDp7xB{e1^Cv7IPAIK0h@somjV)H)-IGd1-cUE4qr9nN1+N$uFk^I z@83;a7YTk@$zA!yh|-eqUh0Y%W60B8RexXJQg8x!`08jw6~Wh#!ptA{6zWc%>sn0l zgLjSu3Fhqt!5py)rF|``q!kH`+q|K*tcf5A3ORlD_yhm)rE0w8#j*ZHBRxZn$o5adysXc}Z$_A2Lu9|Wgswsaf6Cms7%(++{ zCghIDk8kS4a`02s?Cy239?O64YARM_|AEfQQ7kVI`>Eo~w~vdj7@{VZWy$<2dFm$a zFID9Y;*@6(!?mjPyT3lgN(WP7etsGet8X#Ge9R*2_RHQLfD|H z`Jh`VD^FZ7+@NGkpr4TC`hprPa<*3~920*ynzG*Pu*Mua^$V_WB0GO^O7j{PPEjqz z;aex{f=~8nJ1j>n9zM(u z(Ke&mg66xfv5)UCt@#I(_`ZA}FRo^OtL z*3P7PEoNs3zg!;B<{}UiGy6d_H1iRW4|CS%V&rnU`l&dUfUo!d;VRay3PH2zlmu8= z4>5QK4o5}hd5wS30%x|B->dgUOqg(}>Yv1jjuyY-*y8!iHC)+1IqXe32Gjk zDC@@%wMI>;gypZa?R)_T7ILcT+7=)$CDRUD1^Y-2VF? zaG7IJ7ktE;KW)cwP+>%b*ve&$Pps*G^mZ~{khEzpc_$BcKD~y^Vjr{5r)QKveuX|0 zfW^i#CxCbs!!8Lz+AjVm1iGA>-gi7dhH3ZOZ;v*VQfzEY>oB!<9oby{`~{v@b{}{C zJtkh22y}naj$T0|NAH6rWaW)k6ZW;yA-SLX-V%3VZE7ALKH&VTx1!7&(63NQYHl?B zU_G)+e6ked+6fX!MdJR)Vi;z|g{tv z{3~m5F)*Tho3WbZnkD)RPsjE~&^IOd+WHUEaHm`W^_&EZmMfq^>s$ zNvnU~QPUN|x6JcM0e0I0EYPZN^h`O+v1QEDBY897?5D&oOs-~z7p>aLz8(@twsoPa z(8Yg$FWc>5vf=m1p7q<{=zT&jsf5ZsMEv<{z>$M!zGL{N_GAaz4pMDY)t<;V&7n3l zGK)ZG?-xwae3olRt*ASSV^@GO+~v<@ino8)pp`O90N<@b6{$%FS{A2W)*5oEIS69P zeQ(}DCF{)`8q8*qwxN5*h~ESR35VAeC^Y!kCDx`2VLYb8+`24pB1^>(?=~S@bl~GcV?CtUc9k?B1R3AeTPN%nV=-B zVh?*vd-^7H#_05JgTN(8BGitJotkoE=j}I2Z$D}xthsLOdxHVtY^?t9ObNeDd>oKCohy0{+GirwR!;XKTBz%{r z+SVp68e$vnJA=L%rszw~jnj(C)B97KDrJTZuIU%?dm|Rsess~qOIy{B&iB`fAdB+v za?hf7?(-;_?fv9%8vFDXF zqn8-$hOO`{(`k5D9r!aYQ|5nEqQS$;G=(dk5cF%=RxaL|d_h>ja0;|YTFCOW#fdYo z6}5~gMA4`K;moFRGm5+ip^nGdqX--h3s)EiX?+@YdG?Oa=044MfoGwQ;@X-}&YcE@ zoiy_*YxX}7WH3$&%B381PJpjHFL{fHn4Yj|?*y zxM?)rRMLfwFTr9&_Wgg<<$7;AESs(=Skty?)1$1D=h3XM?+xXFLSpv5+P<;b`PBZJ z1c}()VkOkf0du2}_sly4qmZY=mj{+|Ui!22PnXj_b3~#MIT1s3Jc+CiEhw|3>`L4d zpc_NFs>PS2mQqXkeRU@MDu}4$4GFh5R155_SErT`mi3!E&xn7AHevQIK^^#Um$UNt zLzEj$88Ww`QKPk|@`5lYY577^LDbk4A_yQ5R>mTGjs)ImOdM%0)$vEznaOS^F>Jp| zp)0>lI!u1%12(bRcdzaAECxFo&tuulSL#Dk^)Y@fYP_Zi9`01EYrGhY7BC}nqyv(| zn^T!Q+iOj{5RQMJOeHOB*0w>`mJj~HhQHYHz80y&Zefd-HJO``2!1I1R{Aq3N(2&> zO-9uZYAWGSC0n;ZfMO-%o9shR9%*+yZKMs5W!)LsGHNI%0umB}Ga4=>q-CH=9sm64 z`V@o3y+ihDgO%gBe4r*aii1l^GNYi@tKcIa?UZHhQgPPCRok(6W> zIrBgxA^R#fGkYniaH9rlxJ<8u{dDL}yRvV4UzydV^#^as3eHax{FEfFv?2v@?~8rK zmSciPv&&D_?t}ApuBOLGH1pc5lhbd=1<55`qZn<&bH1}VQLA%G z5MkTdpbIM@A~8EU)-D!7;|#6Dn$@jQQeQYFhV4t4S!rsU5)|IXqJyCVQO;T5Rhg@? zVr|eXeXP>!twAW&RB+EmaCp8fn%=-p&FD5_7Ko=YDm7`~4nessHS^8$}^j=2{lcG?nA!a6CBpnE8 zjfAgAGD@We6gk%au5Bwu{9zV)CZXTcetlcV+Meg@j{PE+UBeCvN}+9xakAz0YLe%J z@m+r?C%z;;#kaW8K+uX`w8*FyY9)6VuRA1~mPEm)qttVCA^r&YIN+pSVeIhhV!wwVsR8(z8U z4L0-Y0f$MD!v+B(Re$>978j3os2YDN3zdIgbuOEXfp>o2s%;i`0Jid4j+^Q5O#dJv z^T}|+5_Pr7R3^feYNatspmd9A%a7A5mk0T2t#^K{kwv$z%H+(E1Y>6;I3PyW;Jxdq z;B9X|cQH!TmDymO7YANFwTT)E*@}4+*QU)+&k1uX{i`68*NdMSkEF!4;=rL^h{1p4 zDF?Wabsj9m1Fnxa&keotmY|4GVzJrbuhV&tEZSkQ*s^ZuNg!yO!pmW|1G0=_QfARM?%vtOXL5Zh=go1wYOVnkz9M zF23Sqi@N_x*nVYjy?@WQjh?>IBchx-|0 zg1w2HND@GoHBlo~KlngwThaqO#tR|Do%d(Kw8uXRmoU8OG#T;jGaCFCZD8z@QjM>D zsy=NeB56qzVL`~AZmxes+1Oyvg!;OXAw4GS6+J@Y-V{ajXV;p!FM#`8D?y;^iIGhYJvkXde znbv+F`mDioH)A~;*P9#h6CrT91x9;3RlglCv)2u>F9`tV*?#-xp(Vz+EFmLeSUN0E zkGQ%(5ZCKy{MNn=L*W2>s5%j`x-pF6{;NM@(%o59%OHP!5c9oTP0mf?WJjq)7&U@e z%`YKWim^<3q#px*&Q#h;j>THb@R)wbC%~XYPdRHd?E$4n(ggN9u75ZdO|sDpgTU&0}z!8y5cBR{Ui3O*?xMf!*Y5Q71`A5@pgpWWEVTGqgJqwVy%AdtZB8@$P>q)Xpjl`V-U82YU*`&ETl7>%j8@J3kx>}!`KaV>j zzJ!0I4Ouz9FL};$!$TzK2R)+4yAb<4BO2m>32z!MsI04QkG<$FUyEo{L&oLW4D9;? zTGJDA;SwbEu!X3Q(abzGl%;9A#+QZuogE^J>!& zRs>w~6$|PwENdb4ZLjpVXBxIm(n~)v3RL-fkV2N0t{II@X(TTyx)uV3 z?ch#Q90$GR(QOaVMMsw|Pmg(xu%-+Z^|eTU zxolO;ZS5+e%Cn{BLgdDm(_}Fx7PUs=Yxit1dyj6ijjn*8l`sGFg(Hv~bGw%SZKes? z6khnFJ>y|X5)L_Wzh5-m7lw$aI`V&MT}$UpL*lWiJB^L%TJo(I21$Vdd&b^`x2M^i zHjQ%iTdF}Z?uPf;o;p>J6x0}e?!H@Wyf8TJT3G|+?pepGuQ?lg+t6x6Zp{z<#&}NZ zRuN=9r(Xs486WGQT$8&-i#cBEBWkiE&5)f6M(JCF@V>ssYl9G?dtm;wHphRguG%}6 zCz4+0^k|Pt-fj2^4?bIVq+#sq_TpQfpJ-iGjJFAlH{2UK3>Oak_;zLD?kwN7%R8UU zT}fNqq#9^b`2ts)ypRhefBsUyd`~bpl-+xh&U2cTBDNf)9cg|Y*&A|wwyA2gLzdIf zijR5bDCfly&1a4pqtTvktn0NR6SUP4Uk zrf;PTfsl48XBfRqiB$^He0c2ig&*x}blc*xbl0=#dmZ)Hd+dmBsC7xGiG7+WgUGi- zyZtpt9ugN85xf#!+{7IFD{ut;!A~b{>!|tq(eamEm8B_qeOq)Y@9cl5UXu$C$2d;{ z!BfNOZnRkP5rd3G-?DfRfpX>CmBDlJD=qvnI&AS)@E$w^Cd_C`r?{peDlX%8-L%@jk0(jR>7p{7dDv^#`zP_h9=?@`{b z@^5U~ArDHe&o7;zs7d!Yd##`gr(g)vsrl8IF6=@t61 zc6V&~ra`v!(cvg6(%G|4G0u4d)o?hLyubc%iUXOjPOVbB<8oEwt>%X{Q@fE!UhK91 zl4p4R<_UiZ`CAuiWK5XxH91U}X^HV8b8x8fZd_T8kB4Zzgrj{-82|*iOm4wTM;9V8 z4BIZFO%X5k;l26IFy%}Vw)2z3l3_B=5>E=8j8twU>J#(c}2We3csNI`WnY@ zUn-f&+(c3CW%0t@MTNHgZ7tR4B$uq~DCK4^xlcg&X}4a`T>8{&Q>`j=dRPJ3UKzzr z7h;Dj3o>@5rH}=wR*$I1lWQM&fx{GZ_QoS%>GVhdZHRvXmCARxxmN+IlZkJbmnc-z zgS~&%ry+vOKjZ`IwP=jpacF`HwW>cGk{48KHq2bKw00Q8Gp{oV@)0_w@adzGw~(u* zIS$JDQ^$ufq0cTguelG)Dv>hZ<4Rv@Mf-gcWxCwYeRS9%7RjG6mqtW$-$MRk;RGJ*m? zzrJ@g^){C)eTl<=MkN{kWEbGuN2nHm=I5?={)&gyUT$RELqwM2b%js?NWpue^@C38 z_AiJH9!gdac%SOc$T5H~EpbA-+M%ZLovG%dI;TUmst=v*%go}v^rhdB{EScH9AJN< z$TWZ8)dhxlVfXz8F~0%l{x9wi7r;6{uSePO;+QWDYayNIzB)8hSH|hAJ0@W)RZsa5 zl5S8E^Q3vy+8BwN@TB1^twMckp)8{sC*_BSX1wvO*#7*R`zZ&5rT6rGI$5Qu%)NA8 z%lMGQ;#fZoA3+*TWLTcm*hw^*x-fr&soFanr$TZjIc<~t3?|SY?5hRme^i~sm}!u~ zqh@H9XxD719kJ}KzcE8CyV?3As2qenE7qH4*clpPN^3}AD^VT%y>e;1a5hNX-YTUcRLb8sL`~qfW^k+J9TIa+Mq3L?p++_g3o2AB7qI%$pt*drFi&aSU* zFQ480r7Si|o_kTto)PT5W}X}88;u5;!&pO_EW9M;kmBAmw|z12ZijzacvlHaAGP8f zY+AsekBITGX~R z%q_Owv$=qWasxM)3n|1eaGKUj%H6UrLZlg}m&SASvb!@d`yR>g4c-z&@4a?kOXxdM zny@CsXWG4AKLO+j8}}8n>&CmQIY$lYMb_wpon1~s zrCn25TH>w85|LxN)G{_V-RkrP6jMkx}CjEc7aCyRP{?~0JYn^-8 zp~A^|Ilf0^@_oBSf!ENVGlRM!hnr5fbNxd%r5u99180l!91uy$}%^af0tH zFYcKR@n_p%EM;R|M10jO2rUUA&1MfZdit&&Q-2zNqUoH@*AAm#$b)e#)?f7;wi5FH zIM=p`*JDoo8l!(6^bayhX%?Elj|r{X)eX8Hfk+qtX8QGw!KNZYDJ@AZZq>d7h-gAV zjy?1Cgs-%t3dbo|C40118iPP=jNl2)*?x{RGp&vJ)5I!{9mCW|nstyqB*nLq1nfaL zNHmmG>%{g)NBjF|R^vXddV9K(R58d?EzA8Oa+zJ5KNNqRXY|y8%ZNTeOjox|gJuv* zykqoG4lYp*Z1e(I+6ofs!2)#mMgvh$j~mU%H$d~bibg8AvZdKD%`wn!p!79$eT9p; z>|hUpXm;T`z#d7cG@WvsU}4LX>RpmK5M=Lx%3gp#ETJpyjjcT%@EV6d*@Cn!3JzCh ztSlp)`E7r2soG(PoHwGGFJ>A-E0(n4dDK<)ZxmgVJJcps0<*2+OHxc0obJkaczj#3 z`K02F<2%QPKFoE{`uvNkeOg7KKzPxx;-NX0`88qQi~wMw?^v9$&H~39Ih51CZaOOd zzcx)V;e}d%{w>&?EwGCZlObcyp0Qr7Ql?x_<|BWvWou{5+fTyFP!1R zkVbOW=<^6v%Wm65u1olRlLkvn3w5Ac!gy<0^gyqO#^%YHy>6LI+DsJIK0~861638} zKfA*gFaj$NgW4hC&FD9v1vTjRUFbE1a(Oc87-4LRRv4-tP0FZTG7*K;- zssLr;$>M@Ffx=|qrYkhQyldmvY|KcS%JF}52|>3vf3K_e3|?#$mX<}~y_L})<&3s zYh99L=8EHQDiNrZ30(JpIJBG7e7x@_TVKmi#61JaXgeargFspM_DrAh51^`t9VSJw zs^C>^j|h}*grB0dIAV@>4O(HHpJIPNqN44_^LkkEI(uF#Eh7L>Zk$Wt z4*5h8)~0gvvUlJVTMTT0fo&@&A2AHh6vVf3RMB$ngY;D6aL|F^?T)FTHM0Py5k<*6>3E_r%i9nR@H&G47I)h7QL+L z+|mbyfAhQTi)C0O^v*(3iqgZe9+ib|&!DMPXA55Y-eSPh3(h;1C=VtO1cK_YLVs7i zT%;_xXue&hd@9J};GuXfeRqF_R1h@LSwvD=;;kQTtg(TF6>!`cgCA>3Z&k#zg9;S0 z5-K)Ym;QbtrIIn{*5;u(#B6Fl9Is8HXja(6iISjR)lAr0f7v!4H|UQzc>`;?LGw++_cSXuYU=wUB>w%ivSOdzWdSRd=*8=4KkI-7z83Py>czGci3I zeiD9ITHuirXNI#Hmi9?6|q6 z)KL+K{n+<-;UmmUdhFPZ)hkB8j)3GrSD>I=9Cjz{1V#ASf=9 z)C?+}Y?ivs^|!|j2{;Gq9B7SCEJ4k(a`R`IMtdu_JVUp7N-nEEdV9Rf56E=srhI)~ zJ+{EK%W8}bH|3c315bbr^*t7xgboriUX-?lglLAmjsIGj@X>`G9O+8@AB^g^L4zDo zZff{)jLWM$vq^u73P*FpL)$7oBA6J(DsSK3_iEfb0ELSM@=>dBV8L`MKN16Kx9T z2o*75RWmG(S7+9U9Av)MUEk!*x=um&cdn}RA1RSPhS>@~Q?U-b- zMIL@SxbLFME7)!3_&|TiY<_2lv1ict{mj??PugNL%@A@8@zwY5vK>TpKE~F*jzqz& zswQ;3+u=bs(&IUDMUDV>KddjWvMj;HEGI;uqbtTavG%-^V_v})d1<`e>kxW#(c<=s z{ESmj_=tb23guNBR>`(YF&~a53?pY2i=F~cv)HQ!+7CWTU(1MEW}VQvKi67lIhkE& zAl{DTVIbn0$YBET@;m>zeFPca75F>Gf%!cKBh5e7QDsB>AzcBZ&aamdKwkzw^dxCR za4}h^+OS7_@hZ3G=K7B_XVq^_%Z7d-K?WG#34m6%N;@poYU~62hqkBd@^v^W*o*Rb zq$0uL5CfRNEr=^OOPkz)_0>8|X(A+z@cRu4KwyAmjakZR`;kuESYJuL3T19&b98cL zVQmU!Ze(v_Y6>?uHkVQQ2NSomRSN7L0ya38yx{{U5IHe73NK7$ZfA68AT&8QGnWzU z11Nv(Tx(Ms$F}{>ub7WXo$^e-AE#WWt{)-s1LFX;iJg3qLB>h|5lO^Oe*Il*_edHE zL?lD*&8bS2Y4!B%>E6$^clSuvCReI5Xq~H!wN*-$NxssSqo&f4!@4pWB<_7dBBsEN1a87>p&7~F|18QCy^MBJFFZsZaHWwD%^hx zHh^bL7!&2f<2kV7DyJ=(_mwl?Cn_o(3xaWU666NY&Ku5y5g$?&wXPi17PUbqhpnh> zYE!Ka6{90L+4L#K@A#yhB4?Hdca)K(NvN1!HjT72ooKiu(VS2c)~{bM)Y{X=8QGqeA*es zF+AaDn+RFqYFH@-JmHmL4^MdUB=D_$q=o{X@R=M0`Xt>VMm!NL%uFtH@GgIC#fT^1 zk{rVmm>9fhJOK;f4x~blbHQ>tM!2k2q~j9?Ceuklnm{=cDB*~7d`qXGhM;bzJ!?VnG^2y$6;tJ&+1hnJ(H{#butN``j&$AjL( z;pgfdm+;xwCNk1>o$^mutvA9H01j2iADo{LG2=S{cCtiZm(LX2&m^CO^>Urt>nVLZ zo~PHQ{j@Dzxn7b;*4=6z4oAoRk+7>qBsQHP)9YHI=OgMTX#=Ud!SKAGtn#s}oI-&;i32HFuZ^o<3P%J#wQL}a)Y@V96?oPQikRSid ztu^angzK`7^hgG%#Gu*~2B4sfe)!EJ2PA2_ru{Yx8RSC$|}+MTFPh- zS73C-8nx`5D$~`pzRF+DEH(?Z$}AvL9qb&hv_TDLH)i(Jr%(0I<4JEa9)6ez&%bHc z3{+@t9ZEUW++P6#z-VA~3v2&H9I{NyVP2*sR^1Abe6@d?>yowATx3m)ifbL(kUUC- z^+++T#|~m0T5K(lZo+!BFm?9pvCGmvug8Mpw92eSYoP^a+AtXZUmxA%Nv1M zW?Xk85cPkRjX-v7;XW+GHFY_yiY?Q%oDSV(RczfoXMFteD`kHCS}AS) zy?i%SakhD(D$eF^>mgKJDB!HR1-F_JgCtE2^Q;VIa)_ukk;AOope{b^4NfP+fB*LB ztamxq|8v4?r0=s^u!gFiJ4%OC$Ws<$0&*(qip+l_kM%iYW;wnX)PwVje*J#bkio^A z!ehEMa>@ePpX8M5XhWg^f8A+9OmepL=h*Nrpaeqnyp!EfwSH?oPlehR(&#ENPcG1u zkrbrWp^jBD(ngL%&pxzKn<@ccELzUP1O^Skoju2STv3ZS1|Nkac|4xvLG+RbBnvRc zau0dqy^^Hh(u4qV3f!Zu9d2l1ckPlww)2D)WYb(lQM2H<;%dW!rX!m z%ae+dI%qF5zY0ITy%z(e_JJJw06N$xB;-oW5csS^aRpr|YRn*VQN^NAu$H_*o5-w@ zbOHI*qQ#~Lqf{b+@nItDM`5V}xQCv|@>N5iJ*aVK=Y1A!K4v?i-eXcL~@_t zlNsZ5I43f5`|ap_JRGlC+8P;fmJ=vZG{7DS(o*$FDUP^e0Kh1Z+i3u4wNkPQ9)N$o zH2L8nywhU4PP?c#o|gJRZ$kYc_EzEtL2QCOLFF}>0*2E*Vs{(F!we<&o^x6qp(A7? zvIDze9Ba|*9Hosie-wY><_%W0;(MY=e0KTk>bU6Mf-`foZ83-8mYbM^dgh_8FEe5sR&^4dwZVH`(nM;VM3e8$9(OyZ3WF~GNyjHTo< zQ8E$PvS4arT?rHxnOJ8E>P0Vf#ru;Wdm@ET1cHL~)QA_v2y!>H$k=V?T5*4p7rh}! zFiwpe?Id}dYH^F4GgT72ra4V7bK+i@Jqx4}pPHV~m8JnE0+!P!9;m_lHa# zF63G3c@mElLgBO*{4F=QL&$#up=JIcJWl&0x1M?_Z9CpB!pt2VYX`ofb#8he1(OrF z`X}W7KOm+As%w0?A?%eQ3@|<%ZLO#3OXxknMISN)Dn80t`m!U^KE`{dcI6) zp9Zj1S;9}|DP!_%JG;1-my)iXUjT-YYi7&yTMpcPbp&1RbyOgeIl+I4Zr@7Io>R06 zETd*VM-r75nmKi$;9)D@!ODDsGTpmHfYL9mtjv61NsE#N)tsPaU^%U|Oi?aM8k=WTv}J_*#g$#d~FltUWl?k0k^)T zvr$V5(DF2sm63E}rt_5N%FLzpK~lG#&R*!G)yjw3EDQyiysL(*{RJgqVux8$+Y{X9 zw@PMrTlI3x5!lOhnz1#fEccz_p2^eLS*p?5ogzT#OG}?Dj>CVJARfeblSDwGm#5<8 ziot8kUP_nnzP@cUeqf~#+I$z!io}H1t2Au3LVdQgTQzU5>V~l zyfc%`TApuLFcN?AVxe}0omEuS!N0)iSOug-T1s;11!-yN?v!1+JAVr(A=0r*Hwa62 zN(u-FNJw`{cgN*_?mhS6zTCI@e&;(g=ggUhnfb=&(>qqj-->AF;$5=?T-$Y}$>qw> zkAlQ4!{vs7BR`u*Tx#{p^^4EbV@$2Y^Q_72+xk?q@S)*Mkv);(A5~k`Ln;+G75hB? zXugO&Q$?Ic+Zra2XfA)>KW#0L-4xU~p5xqaooO=o;i%-q>ht>x@1dt#RN}W*#u~vE zz~d%k?{%Ks6%X%^#Oq%QL#HfT69sjzza_ub(_|Is8t)W%XdVb&0x6f3Bh#rXIT;+Q zS9Y(_)`<3zyGo$KB3SReENsL0PmSoeg3RX$sC5Yh35+}+IQq2ys8uIF9IE+F1no?q z^==isRmi9={cv3%F~wO{bo6T|ip|tI9?&UOt*bTuRa?-H{9CZ&2o zFok06(R7V?!;| zzuLDQZ%rIMm{yN1Qxt7kt~qd9KbXaE>5DxHKX>I!zzMSwC~EPw76!)pAV#sb zYk6bKmb_HS5EHPUb>nIGLEge&>_@-aQ+4()Sr@85li4u{Hug`&gbu_QwtFZmJsOW3Aqk*WUX22R|>@U(n3;(t{$E2dI%!oE0_UR)=J|hBSiErR=DcjbdTi^vg8*hvE>|N5~tP=#?@wfa?4@+EY zV>@Z&OOHo(Mzzh40m-rSW~8)yJPUFamnX_qa%I|k;uGG_r6f2bHYS)Y#U$=hU2lP4Qh9)dlquof)ga|3YXkqZ)+#VQlq?!2q%;G|Dj|lrl9pAbK z0F!GeY#sCHscU90@98+&L$aV{nFw~*JSaNvjI4Fy9w?WC+6cuD0v`9S2M=YpYrB3r zfnp68l80>ThKk5(>AitJ_OcST%u%b37{2~3{4sTjFar!^DyR%41jr`N{hg z8h!$!{yD7eV|!t>G;6d#a2&mv;8Fw1)qA?5%Ih58XDr*)CaO0spDyzj$H4akwvXLb zv$Tr%O;{{Z2Qg`DQosZ)!Ox@Y1m~NrzP97O!pB|cp>~`Hp|V;<=XjQNgrVjHiSBiv zfQ6jjC(F&=dSH%@jM@V5?q-6@J5qc(8-ydJx# z?|uVq#SZ2{X>m1I6b{bXi%df*e3?q_=?5zTQCv!)&qgEX-ezk;G5SXnAX5o|a)^Bp5~?z;FsRa`|GHQh9;`=)GyAE5`*`axJM=I4L3Gi)?A z+6i?20>#*vUlRqJDL$hnwQb9OZTef=>We`^Ze`|YQ;+%2pLe2Tu|h-kna0npealq{ zAO$~c>nSrcFXT>*Au4Wv4~Yeb?EkLU%w$@!$||}wi%J~C{*>?5cq|kp6-2vi(~E7$ z+8@uqWzR4rFWQRzlsi%yn2Fex?A`dH~Y0fAK#>fA-dHy>OJs!3p{<_cL&*z4!KVm8Dj%227)W!)ghOdEV zdy|*pXf+b^wYWo4f;gg>XYLm8CbN1e7dxhiB!zX7Vux6Ara2rM0xue|phkv-GB$t{ z9{ufcFnCjB@9W{a%E{i|ndLjTRywtL9Lmgt_SuCE#lJiI&F3{Aw-0RC$^VcadXbfW zUZmg{4YiH_UtTq~MuQ7b-8#+^#O+u1KW+1+dz zEc$D6gXfj>?0Y@rIx%n~f#Tt$b(R(wJX9hh-Ydyi#_5CuT1#s-%JXqKMY@w z9r?Hfh3H_@ilZ=$Tr<>q_j~Od-x26+p5A-tu+%I)X!GLG`cG3{ZWQqkg2@3g{{F1p zq^xpB0=YF@sHEwFUFPNNGnI-UKB3G2s=-Gc_cB>0!@z59!1vk*dOcxc8h(Ak?AU&H za(#Vr;_K_HSr)hG(0(@=y2*r!%FK`#Qtg^KR! z4n=&91c7-3xgk8l+&sL3d;;v;d@S7DEUX}~inEoBg$I=Gjm%qK?*C(g*og#1qC+4; zh_)4U%2$E{Ah4XRo4W^{fPm2d{si(0@(cd=Phg#)k=C37SnT7;m>Nix(?BO(3y?e1lIdqm2lU~owwJmto?{WQb9c``*sAvGUY@$};; z2R)-wF&(ZohlOTGo3c+POHu@5o~T8WGGjk(PNb$y(K0w$k$E|cn`iY)E(?gu=iS$L zSpWz9YgravdJ1$5%(uLuwjf<$)K+aq{!i){Sdz=H(FqG&=vQKS6Jqi)$d$XEqiFp` zC6S9ikue`#RvpEN>r4yb34h`9jxa1VGsNXN|B!S9hsr<%+e=IzOhSw7q>wk5UKC%@ z_!78NzGF-X4HAzMb?C97kyC^}i`#nFfoTr7WX9&N#&MwvYoZmS_~HT~S|DU^QB1L^g)Q6=>TqhygVxHmnBSP(O;fr<^obEZ2b}p z&f9HjysYAaj@+c?2zBl11UAn|Ud*x%D-k3b2?AgK#u z7p0`w4RxDhe5@}O%h`0m&eQf9Qb|~`V%h&n*(R~{6vUTU^>MC`&!K1{hN2n z3!|GVs+cuy6ny5hbE45WZ6lpPqp zGE~w3#kAQf;M_^ASmp<)pNEYR`0~InQJ7*BR zKXS{NyR{3$hc3}mkId3q%SOGA?L9*k)NDnz@80--2eLnj^@yca9v`&Yq@G+12!oHz zb|ooI5I{4Jm;L(HRge0!vd!K!>n}WlwB27$f@tO8?+z)mmiaZ|&wUGK5e8 z=5}VWj`JK2Ran87%!uu;)e#@ww!cZ!;Q>=3f`KVP`LeXhR=i5%k!(I`*C4$~cqMlx zwC6PY=kMND!^O7nH{|gNp_2_af-vKW}DlguydDw({l1y z9g`=DJ3Z;|*Gd1v&XS)sw1<6P<6RApCMezMZ1AS^Ni}@js6BFoMmYsU>Qh`@Nsn*C z$^f0PpwIt2bjZh9*7`YDY-e_k<+DZY0S4dSRkXfu*nUXY2j#IodB5D$7aomn!mmPS zD{OHhMrlmak2BjPm01`Lj&`g{$*f9$Xhb0Wac8i{9Aa`wWG+ojCEx6$xduc|E#JPF+}P*b!&V{Z zd6xh9=FyNuDU0@(bdi}wvfv~Usw?}k|3LF|EqR@!e}+SZ<3fMY+l+T+5(mX|EJSFw zcMfQxc}&St_RKwFO%=mrP9$CQIh@|3f6Zy&76~scaGL8HumMBlWex2AQuQ1Jp@#wY z3fNw*e)7G)V6E^nDJwAr_D-Sws_|!MiRFRG`U(fdhQrn-sIvO`d6NN7NXU$f7jftA=5i)hrY*c?@CrOSo?~?N& zXG6BVPc^GK`r3{ShBg7L7ayX({S*d5R0~k`SKU736F4w|m{vAjrgW78wa5wT<2smj z>kh310=MgPD_ObDTGJOLw6{qid8QwXw_m;C_glPa`+V|dl6$?;Hhm{mHyFqA0}AYR zg|Zh$i`&%6<0V)T)JVg`KM!yX{_75`I$|QB-ybaDZDAtXJT4G^B8~E^P7eaZ<1upF zRg&ql7_hY!>EHnnL86xqS+bFTenW-nu3PzFp%KD1$?%0`4~OdRO}qZ0uPGa{^ZdTV zD2A(l?!-xF_qf*QEK8?2!Ie;3@(Rz-BrM$iG2IZcg3B0H@r!#RVgbI?43hrKfj_8G zPjB4cc=XKxA?3*TxUQdQp#^ODYWExoiTVi2tx~^g?@7qp6)NdSv~M+*3_bOUXA=Li zstby(JkIa@wcZpEOb?J(QN9THcv^Gq@a^pz5mdaNH5}=v)a437q1Jn?MWJGnenFKd zNoDdoQX&O-78cHxG|c&$S0BPM;-`(;C_ZIzjk{o-BomSjnP0wmMW6(jh+{UtR-uKN?rF37WWJ&YBJKq|L3_ag+>C+W`MVlFv zHHk&1^h7eLkk=DY~W(7=0^A8Lk$uuM|}awl|3yYH;=mi4A`=0~LPi27G$ofZ1q80+rm+iP5(T3D?_WBpz z%zt(C_pLj=JJiTO&7HTg=Ukwnr&hc(8I0OG<|L2{PZ7ayT3gn&L3l(nx|0u&h~-Qn z5ZF~DM=bnajFjXZO0pfa3;MJ}(eu$pOD~QnFY75cHE3dxpl~Gt>7REe@#n@I_TTqg9&UtvB8UzU z;^8p?fwgS?pma}75Ll1Sh=-1k?mtRRXJ-#OA#Nc$ZaPg6SkVdQ{3Hqgk0e89B+C0v zfLlm_8^SHf`^1ytkrm>G2tj0p`T69er-hWEPSzeabiCY%Q3Mtn zz=SYxFqJ3mJG?&d{=uobhg4m4h*@KbqVD@`*Jni4cbqzHM&um%V;UC^jZs~lHAx{j zA%7M1MJ`b>obH|7cXck3;u{ogHVno>Hf^{eE{(bYBaRk5n~vxUlFiGljUxBoLm48M zLytp4(XZ`&k|J}Jt;S*idHp^<4ddBQ0F;h+d4$NmBU(z(1ANF-PYgj7wk+$}aXrnJ z+@q_aZj@DjEvW1uasX17b@)6jIG=2;6Va61Q@^Sk`pVKM(y`M7Z|cRIDiYtDQ(xXe zqH4Htw-d|qzDICnDkQpxIl{3=N80DL)&QXvIbQQ8eSs>x#?&}iiP~txzr18Tpg{53 z9z!LYq87ejPB8uKB3O46>mt<41;qod>hgjLExZ$#8%H&S-UakLble={O9&4)K63}7 z5c=;B)2LiK&H(al{H#<{G}874l;<)%WK}2GpMQ%j<2YeixyFWkXq>ciq}s!lHvgzG zTX|uzk7k70t@k?wH=9a1*j62D3+QO#R040XX|J^fAt^=C~J z_Q%+BN13@nlSH;-#1(J(i*=oqqPU$2pb*{&p!~b`MCtU+Umc5+H>Pp*UEhvRlJB6Q z_x-p|m>{q4|1}1(c-IWslvLQrf=CyjgA`XHd*W=Q=HwW+P_(@{gMuW!mCXX49B1=|WVJpTw&W+S zXmjjhY>`f{QAq^5Iamnur_FSjC($12V9!paVHHRpF%y#169jOh6!WSj`<4u8K{MG- zwZ9ep64lxK{-nBgjlv>N;(kA@#=Gs6T_F3beK0rBWak~8-@f0T=Z2sl11$AtcJj+} z4cn);vYyqq_$+x1s}>r}`2ZPvv1#1==xW&oa7`kDQj$TxTRmb)ODyx#iD5BC6m!-j zKaA6`nY{jS0{~+iUwd>+Zon_`Fh!?`TpT~$XgFv3olc|`EsCCReut>F`R)vBIh$6r zc0K2(N2hp!osp1JYuH1{$UWLiE4t%=4g`ep+Q| z9PYbbJ{$W>pPIBE<*SCt8vgc|pc{4QR^t8VkV@FpH$Jj@E~4YZ@uSEqYErT+Ih}1e zZ#&_=YFuM3X<=IvE@;m~y0aHn`KVu25L3Mqf9n_C!OD#mqc3_&eJ+BlIj5~Z}KaMD2HvwO5No)%d9j}W~bhIhH7~TyL zzj}WeK@`ymnqW;{;C(s*_MaB!7~p%DL@sHWH>m*xf8jN;t1$XiC2zS zM+1pbar35<87(GoOK+8w{X4xYWH~JsM|81!4Kz!usgZ0@T!QRXd!X>LJeVzv$^?kSQhi)i8Vva)xof683~LyM1J#d*My!XvKLghp;T3&Zt}O$~XSFUf5B~ z6oqb4{OcZGV1%mIj8^`@yUgQgg<9X2-$zn;lL~S+*bG=7 zy!lyZ7GGXWBY)n=ROv9lw>0kacD8HoAdr5!AQDb1LF4j+Yq^dbEhvrP>K^clwoakE z`i}g<1lT0$ps{mYAGXf3RT3ewWR)InOazc<0^9J@yxcrZYe=}~r_jaVTBk8Un zK>j_ER&{*BfHr`U&&QO(`cHX4y=0e8SY<%xsfWwpU)%gz^D8>jvJiMLaKYUN$%$Y5&iyAy8%Yoz+4yn%Ju=q-zZ zmTu$mN2D!#>lLX10A+6_A9!osD*zsN>)cyET+OBLy3SAyL;9R!SHsGDPHC$*kyoV^ zzfv3i6zV}a^=%WBojI38U!ZKGWJpR3zMUXI zzc-@Byvf@^Tac9aCzefq797MLy%de)_h!GLJs4J|d5We(&}D#LJ% - Image Processing Division\newline - National Institute for Space Research\newline - Av. dos Astronautas, 1758\newline - 12227-010 São José dos Campos, Brazil + Ecosystems Services and Management Program\newline + International Institute for Applied Systems Analysis\newline + Schlossplatz 1\newline + A-2361 Laxenburg, Austria email: vwmaus1@gmail.com - url: \url{www.dpi.inpe.br/maus} - Telephone: +55/12/3208-7330 - - name: "Gilberto Câmara" + url: \url{www.iiasa.ac.at/staff/maus} + Telephone: +43/2236-807-550 + - name: Gilberto Câmara affiliation: INPE affiliation2: University of Münster - name: Marius Appel @@ -19,27 +19,25 @@ author: - name: Edzer Pebesma affiliation: University of Münster title: - formatted: "\\pkg{dtwSat}: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in \\proglang{R}$^1$" + formatted: "\\pkg{dtwSat}: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in \\proglang{R}" # If you use tex in the formatted title, also supply version without plain: "dtwSat: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in R" # For running headers, if needed short: "\\pkg{dtwSat}: Time-Weighted Dynamic Time Warping" abstract: > - The opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. However, few of the proposed methods for remote sensing time series analysis are available as open source software. In this paper we present the \proglang{R} package \pkg{dtwSat}. This package provides an implementation of the Time-Weighted Dynamic Time Warping method for land use and land cover mapping using sequence of multi-band satellite images. Methods based on dynamic time warping are flexible to handle irregular sampling and out-of-phase time series, and they have achieved significant results in time series analysis. \pkg{dtwSat} is available from the Comprehensive R Archive Network and contributes to making methods for satellite time series analysis available to a larger audience. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and evaluating the results. + The opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. However, few of the proposed methods for remote sensing time series analysis are available as open source software. In this paper we present the \proglang{R} package \pkg{dtwSat}. This package provides an implementation of the Time-Weighted Dynamic Time Warping method for land cover mapping using sequence of multi-band satellite images. Methods based on dynamic time warping are flexible to handle irregular sampling and out-of-phase time series, and they have achieved significant results in time series analysis. \pkg{dtwSat} is available from the Comprehensive R Archive Network and contributes to making methods for satellite time series analysis available to a larger audience. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and assessing the results. keywords: # at least one keyword must be supplied - formatted: [dynamic programming, MODIS time series, land use changes, crop monitoring] - plain: [dynamic programming, MODIS time series, land use changes, crop monitoring] + formatted: [dynamic programming, MODIS time series, land cover changes, crop monitoring] + plain: [dynamic programming, MODIS time series, land cover changes, crop monitoring] preamble: > \usepackage{amsmath} \usepackage{array} \usepackage{caption} \usepackage{subcaption} \usepackage{float} - \usepackage{framed} - \usepackage{listings} - \usepackage{siunitx} - \usepackage{latexsym} + \usepackage{microtype} + \setlength{\tabcolsep}{4pt} documentclass: nojss classoption: shortnames output: @@ -51,7 +49,7 @@ vignette: > %\VignetteIndexEntry{dtwSat: Time-Weighted Dynamic Time Warping for Satellite Image Time Series Analysis in R} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8} - %\VignetteDepends{rticles, dtwSat, ggplot2, scales, Hmisc, reshape2, xtable, tikzDevice} + %\VignetteDepends{rticles, dtwSat, ggplot2, scales, Hmisc, reshape2, tikzDevice} --- @@ -62,6 +60,7 @@ opts_chunk$set( message = FALSE, error = FALSE, results = "hide", + cache.path = "./cache/", cache = FALSE, comment = "" ) @@ -83,11 +82,9 @@ if (other_user) { #options(tikzDocumentDeclaration = "\\documentclass{jss}\\usepackage{siunitx}\\usepackage{latexsym}") ``` - - ```{r , echo=FALSE, eval = TRUE, cache = FALSE} # Install dtwSat package -#install.packages("dtwSat_0.2.0.9000.tar.gz") +#install.packages("dtwSat_0.2.2.9000.tar.gz", repos = NULL) ``` @@ -95,9 +92,10 @@ if (other_user) { library(dtwSat) library(ggplot2) library(scales) -library(reshape2) +library(Hmisc) new_theme = theme_get() +new_theme$text$family = "Helvetica" new_theme$text$size = 8 old_theme = theme_set(new_theme) @@ -106,30 +104,28 @@ page_width = 5.590551#in 14.2#cm page_height = 9.173228#in 23.3#cm ``` -\footnotetext[1]{This vignette is based on the paper: MAUS, V.; CAMARA, G.; APPEL, M.; PEBESMA, E. dtwSat: Time-Weighted Dynamic Time Warping for satellite image time series analysis in R. Submitted to the Journal of Statistical Software.} - \sloppy # Introduction - -Remote sensing images are the most widely used data source for measuring land use and land cover change (LUCC). In many areas, remote sensing images are the only data available for this purpose [@Lambin:2006; @Fritz:2013]. Recently, the opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. These methods include multi-image compositing [@Griffiths:2013], detecting forest disturbance and recovery [@Kennedy:2010; @Zhu:2012; @DeVries:2015], crop classification [@Xiao:2005; @Wardlow:2007; @Petitjean:2012; @Maus:2016], planted forest mapping [@Maire:2014], crop expansion and intensification [@Galford:2008; @Sakamoto:2009], detecting trend and seasonal changes [@Lunetta:2006; @Verbesselt:2010; @Verbesselt:2010a; @Verbesselt:2012], and extracting seasonality metrics from satellite time series [@Jonsson:2002; @Jonsson:2004]. Given the open availability of large image data sets, the research community on Earth Observation would get much benefit from methods that are openly available, reproducible and comparable. However, few of the proposed methods for remote sensing time series analysis are available as open source software, the main exception being the BFAST and BFAST-monitor algorithms for change detection [@Verbesselt:2010; @Verbesselt:2010a]. This paper is a contribution to making methods for satellite time series analysis available to a larger audience. +Remote sensing images are the most widely used data source for measuring land use and land cover change (LUCC). In many areas, remote sensing images are the only data available for this purpose [@Lambin:2006; @Fritz:2013]. Recently, the opening of large archives of satellite data such as LANDSAT, MODIS and the SENTINELs has given researchers unprecedented access to data, allowing them to better quantify and understand local and global land change. The need to analyse such large data sets has lead to the development of automated and semi-automated methods for satellite image time series analysis. These methods include multi-image compositing [@Griffiths:2013], detecting forest disturbance and recovery [@Kennedy:2010; @Zhu:2012; @DeVries:2015], crop classification [@Xiao:2005; @Wardlow:2007; @Petitjean:2012; @Maus:2016], planted forest mapping [@Maire:2014], crop expansion and intensification [@Galford:2008; @Sakamoto:2009], detecting trend and seasonal changes [@Lunetta:2006; @Verbesselt:2010; @Verbesselt:2010a; @Verbesselt:2012], and extracting seasonality metrics from satellite time series [@Jonsson:2002; @Jonsson:2004]. Given the open availability of large image data sets, the Earth Observation community would get much benefit from methods that are openly available, reproducible and comparable. However, few of the proposed methods for remote sensing time series analysis are available as open source software, the main exception being the BFAST and BFAST-monitor algorithms for change detection [@Verbesselt:2010; @Verbesselt:2010a]. This paper is a contribution to making methods for satellite time series analysis available to a larger audience. In this paper we describe the \pkg{dtwSat} package, written in \proglang{R} [@R:2015] and \proglang{Fortran} programming languages, and available from the Comprehensive R Archive Network at [http://CRAN.R-project.org/package=dtwSat](http://CRAN.R-project.org/package=dtwSat). The package provides an implementation of Time-Weighted Dynamic Time Warping (TWDTW) [@Maus:2016] for satellite image time series analysis. -The TWDTW method is an adaptation of the well-known dynamic time warping (DTW) method for time series analysis [@Velichko:1970; @Sakoe:1971; @Sakoe:1978; @Rabiner:1993; @Berndt:1994; @Keogh:2005; @Muller:2007] for land use and land cover classification. The standard DTW compares a temporal signature of a known event (*e.g.*, a person's speech) with an unknown time series. It finds all possible alignments between two time series and provides a dissimilarity measure [@Rabiner:1993]. In contrast to standard DTW, the TWDTW method is sensitive to seasonal changes of natural and cultivated vegetation types. It also considers inter-annual climatic and seasonal variability. In a tropical forest area, the method has achieved a high accuracy for mapping classes of single cropping, double cropping, forest, and pasture [@Maus:2016]. +The TWDTW method is an adaptation of the well-known dynamic time warping (DTW) method for time series analysis [@Velichko:1970; @Sakoe:1971; @Sakoe:1978; @Rabiner:1993; @Berndt:1994; @Keogh:2005; @Muller:2007] for land cover classification. The standard DTW compares a temporal signature of a known event (*e.g.*, a person's speech) with an unknown time series. It finds all possible alignments between two time series and provides a dissimilarity measure [@Rabiner:1993]. In contrast to standard DTW, the TWDTW method is sensitive to seasonal changes of natural and cultivated vegetation types. It also considers inter-annual climatic and seasonal variability. In a tropical forest area, the method has achieved a high accuracy for mapping classes of single cropping, double cropping, forest, and pasture [@Maus:2016]. + +We chose \proglang{R} because it is an open source software that offers a large number of reliable packages. The \pkg{dtwSat} package builds upon on a number of graphical and statistical tools in \proglang{R}: \pkg{dtw} [@Giorgino:2009], \pkg{proxy} [@Meyer:2015], \pkg{zoo} [@Zeileis:2005], \pkg{mgcv} [@Wood:2000; @Wood:2003; @Wood:2004; @Wood:2006; @Wood:2011], \pkg{sp} [@Pebesma:2005; @Bivand:2013], \pkg{raster} [@Hijmans:2015], \pkg{caret} [@Kuhn:2016], and \pkg{ggplot2} [@Wickham:2009]. Other \proglang{R} packages that are related and useful for remote sensing and land cover analysis include \pkg{landsat} [@Goslee:2011], \pkg{rgdal} [@Bivand:2015], \pkg{spacetime} [@Pebesma:2012; @Bivand:2013], \pkg{bfast} [@Verbesselt:2010; @Verbesselt:2010a], \pkg{bfastmonitor} [@Verbesselt:2011], \pkg{bfastSpatial} [@Dutrieux:2014], \pkg{MODISTools} [@Tuck:2014], \pkg{maptools} [@Bivand:2015], and \pkg{lucc} [@Moulds:2015]. Using existing packages as building blocks, software developers in \proglang{R} save a lot of time and can concentrate on their intended goals. -We chose \proglang{R} because it is an open source software that offers a large number of reliable packages. The \pkg{dtwSat} package builds upon on a number of graphical and statistical tools in \proglang{R}: \pkg{dtw} [@Giorgino:2009], \pkg{proxy} [@Meyer:2015], \pkg{zoo} [@Zeileis:2005], \pkg{mgcv} [@Wood:2000; @Wood:2003; @Wood:2004; @Wood:2006; @Wood:2011], \pkg{sp} [@Pebesma:2005; @Bivand:2013], \pkg{raster} [@Hijmans:2015], \pkg{caret} [@Kuhn:2016], and \pkg{ggplot2} [@Wickham:2009]. Other \proglang{R} packages that are related and useful for remote sensing and land use analysis include \pkg{landsat} [@Goslee:2011], \pkg{rgdal} [@Bivand:2015], \pkg{spacetime} [@Pebesma:2012; @Bivand:2013], \pkg{bfast} [@Verbesselt:2010; @Verbesselt:2010a], \pkg{bfastmonitor} [@Verbesselt:2011], \pkg{bfastSpatial} [@Dutrieux:2014], \pkg{MODISTools} [@Tuck:2014], \pkg{maptools} [@Bivand:2015], and \pkg{lucc} [@Moulds:2015]. Using existing packages as building blocks, software developers in \proglang{R} save a lot of time and can concentrate on their intended goals. +There is already an \proglang{R} package that implements the standard DTW method for time series analysis: the \pkg{dtw} package [@Giorgino:2009]. In the \pkg{dtwSat} package, we focus on the specific case of satellite image time series analysis. The analysis method implemented in \pkg{dtwSat} package extends that of the \pkg{dtw} package; it adjusts the standard DTW method to account for the seasonality of different types of land cover. Our aim is to support the full cycle of land cover classification, from selecting sample patterns to visualising and assessing the final result. -There is already an \proglang{R} package that implements the standard DTW method for time series analysis: the \pkg{dtw} package [@Giorgino:2009]. In the \pkg{dtwSat} package, we focus on the specific case of satellite image time series analysis. The analysis method implemented in \pkg{dtwSat} package extends that of the \pkg{dtw} package; it adjusts the standard DTW method to account for the seasonality of different types of land cover. Our aim is to support the full cycle of land use and land cover classification, from selecting sample patterns to visualising and evaluating the final result. +This paper focuses on the motivation and guidance for using the TWDTW method for remote sensing applications. The full description of the method is available in a paper published by the lead author [@Maus:2016]. In what follows, the \autoref{the-time-weighted-dynamic-time-warping-method} describes the application of TWDTW [@Maus:2016] for satellite time series analysis. The \autoref{dtwsat-package-overview} gives an overview of the \pkg{dtwSat} package. Then, \autoref{classifying-a-time-series} focuses on the analysis of a single time series and shows some visualisation methods. We then present an example of a complete land cover change analysis for a study area in Mato Grosso, Brazil in \autoref{producing-a-land-cover-map}. -This paper focuses on the motivation and guidance for using the TWDTW method for remote sensing applications. The full description of the method is available in a paper published by the lead author [@Maus:2016]. In what follows, Section \ref{dtwsat-package-overview} gives an overview of the \pkg{dtwSat} package. The Section \ref{the-time-weighted-dynamic-time-warping-method} describes the application of TWDTW [@Maus:2016] for satellite time series analysis. Then, Section \ref{classifying-a-time-series} focuses on the analysis of a single time series and shows some visualisation methods. We then present an example of a complete land use and land cover change analysis for a study area in the Mato Grosso, Brazil in Section \ref{producing-a-land-cover-map}. # The Time-Weighted Dynamic Time Warping method -In this section, we describe the Time-Weighted Dynamic Time Warping (TWDTW) algorithm in general terms. For a detailed technical explanation, refer to @Maus:2016. TWDTW is time-constrained version of the Dynamic Time Warping (DTW) algorithm. Although the standard DTW method is good for shape matching [@Keogh:2005], it is not suited *per se* for satellite image time series analysis, since it disregards the temporal range when finding the best matches between two time series [@Maus:2016]. When using image time series for land cover classification, one needs to balance between shape matching and temporal alignment, since each land cover class has a distinct phenological cycle associated with the vegetation [@Reed:1994,@Zhang:2003]. For example, soybeans and maize cycles range from 90 to 120 days, whereas sugar-cane has a 360 to 720 days cycle. A time series with cycle larger than 180 days is unlikely to come from soybeans or maize. For this reason, @Maus:2016 include a time constraint in DTW to account for seasonality. The resulting method is capable of distinguishing different land use and land cover classes. +In this section, we describe the Time-Weighted Dynamic Time Warping (TWDTW) algorithm in general terms. For a detailed technical explanation, refer to @Maus:2016. TWDTW is time-constrained version of the Dynamic Time Warping (DTW) algorithm. Although the standard DTW method is good for shape matching [@Keogh:2005], it is not suited *per se* for satellite image time series analysis, since it disregards the temporal range when finding the best matches between two time series [@Maus:2016]. When using image time series for land cover classification, one needs to balance between shape matching and temporal alignment, since each land cover class has a distinct phenological cycle associated with the vegetation [@Reed:1994,@Zhang:2003]. For example, soybeans and maize cycles range from 90 to 120 days, whereas sugar-cane has a 360 to 720 days cycle. A time series with cycle larger than 180 days is unlikely to come from soybeans or maize. For this reason, @Maus:2016 include a time constraint in DTW to account for seasonality. The resulting method is capable of distinguishing different land cover classes. -The inputs to TWDTW are: (a) a set of time series of known temporal patterns (*e.g.*, phenological cycles of land cover classes); (b) an unclassified long-term satellite image time series. For each temporal pattern, the algorithm finds all matching subintervals in the long-term time series, providing a dissimilarity measure (cf. Figure \ref{fig:twdtw-example}). The result of the algorithm is a set of subintervals, each associated with a pattern and with a dissimilarity measure. We then break the unclassified time series in periods according to our needs (*e.g.*, yearly, seasonality, monthly). For each period, we consider all matching subintervals that intersect with it, and classify them based on the land cover class of the best matching subinterval. In this way, the long-term satellite time series is divided in periods, and each period is assigned a land cover class. +The inputs to TWDTW are: (a) a set of time series of known temporal patterns (*e.g.*, phenological cycles of land cover classes); (b) an unclassified long-term satellite image time series. For each temporal pattern, the algorithm finds all matching subintervals in the long-term time series, providing a dissimilarity measure (cf. \autoref{fig:twdtw-example}). The result of the algorithm is a set of subintervals, each associated with a pattern and with a dissimilarity measure. We then break the unclassified time series in periods according to our needs (*e.g.*, yearly, seasonality, monthly). For each period, we consider all matching subintervals that intersect with it, and classify them based on the land cover class of the best matching subinterval. In this way, the long-term satellite time series is divided in periods, and each period is assigned a land cover class. ```{r twdtw-example, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_height/3.5, fig.align='center', fig.cap='Matches of the known temporal pattern to subintervals of the long-term time series. The solid black line is the long-term time series, the colored lines are the different matches of the same pattern ordered by TWDTW dissimilarity measure, and the gray dashed lines are the matching points.', fig.pos='h'} n=4 @@ -143,12 +139,12 @@ df_dist$y = 1.8 plotMatches(mat, attr="evi", k=n) + ylab("Time series Pattern") + geom_text(data=df_dist, mapping = aes_string(x='to', y='y', label='label'), - size = 2) + + size = 2, family="Helvetica") + theme(legend.position="none") ``` -To use TWDTW for land use and land cover classification, we need the following data sets: +To use TWDTW for land cover classification, we need the following data sets: - A set of remote sensing time series for the study area. For example, a tile of a MODIS MOD13Q1 image consists of 4800 x 4800 pixels, covering an area of 10 degrees x 10 degrees at the Equator [@Friedl:2010]. A 15-year (2000-2015) MODIS MOD13Q1 set time series has 23 images per year, with a total of 23 million time series, each with 346 samples. @@ -156,7 +152,7 @@ To use TWDTW for land use and land cover classification, we need the following d - A set of ground truth points, with spatial and temporal information and land cover classification. These *ground truth* points are used for validation and accuracy assessment. -Based on the information provided by the user about the images to be analysed, our method maps them to a three-dimensional (3-D) array in space-time (Figure \ref{fig:3-D-array}). This array can have multiple attributes, such as the satellite bands (*e.g.*, "red", "nir", and "blue"), and derived indices (*e.g.*, "NDVI", "EVI", and "EVI2"). This way, each pixel location is associated to a sequence of measurements, building a satellite image time series. Figure \ref{fig:3-D-array} shows an example of "evi" time series for a location in the Brazilian Amazon from 2000 to 2008. In the first two years, the area was covered by forest that was cut in 2002. The area was then used for cattle raising (pasture) for three years, and then for crop production from 2006 to 2008. Satellite image time series are thus useful to describe the dynamics of the landscape and the land use trajectories. +Based on the information provided by the user about the images to be analysed, our method maps them to a three-dimensional (3-D) array in space-time (\autoref{fig:3-D-array}). This array can have multiple attributes, such as the satellite bands (*e.g.*, "red", "nir", and "blue"), and derived indices (*e.g.*, "NDVI", "EVI", and "EVI2"). This way, each pixel location is associated to a sequence of measurements, building a satellite image time series. \autoref{fig:3-D-array} shows an example of "EVI" time series for a location in the Brazilian Amazon from 2000 to 2008. In the first two years, the area was covered by forest that was cut in 2002. The area was then used for cattle raising (pasture) for three years, and then for crop production from 2006 to 2008. Satellite image time series are thus useful to describe the dynamics of the landscape and the land use trajectories. \begin{figure}[!h] \begin{center} @@ -168,6 +164,7 @@ Based on the information provided by the user about the images to be analysed, o \end{figure} + # dtwSat package overview \pkg{dtwSat} provides a set of functions for land cover change analysis using satellite image time series. This includes functions to build temporal patterns for land cover types, apply the TWDTW analysis using different weighting functions, visualise the results in a graphical interface, produce land cover maps, and create spatiotemporal plots for land changes. Therefore, \pkg{dtwSat} gives an end-to-end solution for satellite time series analysis, which users can make a complete land change analysis. @@ -185,6 +182,7 @@ The class \code{twdtwRaster} is used for satellite image time series. This class + # Classifying a time series This section describes how to classify one time series, using examples that come with the \pkg{dtwSat} package. We will show how to match three temporal patterns ("soybean", "cotton", and "maize") to subintervals of a long-term satellite image time series. These time series have been extracted from a set of MODIS MOD13Q1 [@Friedl:2010] images and include the vegetation indices "ndvi", "evi", and the original bands "nir", "red", "blue", and "mir". In this example, the classification of crop types for the long-term time series is known. @@ -193,17 +191,17 @@ This section describes how to classify one time series, using examples that come The inputs for the next examples are time series in \pkg{zoo} format. The first is an object of class \code{zoo} with a long-term time series, referred to as \code{MOD13Q1.ts}, and the second is a \code{list} of time series of class \code{zoo} with the temporal patterns of "soybean", "cotton", and "maize", referred to as \code{MOD13Q1.patterns.list}. -From \code{zoo} objects we construct time series of class \code{twdtwTimeSeries}, for which we have a set of visualization and analysis methods implemented in the \pkg{dtwSat} package. The code below builds two objects of class \code{twdtwTimeSeries}. The first has the long-term time series and second has the temporal patterns. We use the plot method types \code{timeseries} and \code{patterns} to shown the objects \code{ts} in Figure \ref{fig:example-timeseries} and \code{patterns_ts} in Figure \ref{fig:temporal-patterns-soy-cot-mai}, respectively. This plot method uses \code{ggplot} syntax. -```{r, echo = TRUE, eval = TRUE, results = 'markup'} -ts = twdtwTimeSeries(MOD13Q1.ts, labels="Time series") -patterns_ts = twdtwTimeSeries(MOD13Q1.patterns.list) -MOD13Q1.ts.labels -``` +From \code{zoo} objects we construct time series of class \code{twdtwTimeSeries}, for which we have a set of visualization and analysis methods implemented in the \pkg{dtwSat} package. The code below builds two objects of class \code{twdtwTimeSeries}. The first has the long-term time series and second has the temporal patterns. We use the plot method types \code{timeseries} and \code{patterns} to shown the objects \code{ts} in \autoref{fig:example-timeseries} and \code{MOD13Q1.ts} in \autoref{fig:temporal-patterns-soy-cot-mai}, respectively. This plot method uses \code{ggplot} syntax. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} library(dtwSat) names(MOD13Q1.patterns.list) head(MOD13Q1.ts, n = 2) ``` +```{r, echo = TRUE, eval = TRUE, results = 'markup'} +ts = twdtwTimeSeries(MOD13Q1.ts, labels="Time series") +patterns_ts = twdtwTimeSeries(MOD13Q1.patterns.list) +patterns_ts +``` @@ -216,7 +214,7 @@ plot(ts, type = "timeseries") + plot(patterns_ts, type = "patterns") ``` -TWDTW uses both amplitude and phase information to classify the phenological cycles in the long-term time series. The EVI peak of the "soybean" time series has a similar amplitude as that of "cotton". However, the "soybean" series peaks in late December while the "cotton" series peaks in early April. The EVI peak of the "maize" time series is at the same period as the peak of "cotton". However, the "maize" time series has smaller amplitude than the "cotton" one. Therefore, we can improve the time series classification by combining shape and time information. +TWDTW uses both amplitude and phase information to classify the phenological cycles in the long-term time series. The differences in the amplitude and phase of the cycles are more clear when we observe the EVI signal in Figures 3 and 4. The EVI peak of the "soybean" time series has a similar amplitude as that of "cotton". However, the "soybean" series peaks in late December while the "cotton" series peaks in early April. The EVI peak of the "maize" time series is at the same period as the peak of "cotton". However, the "maize" time series has smaller amplitude than the "cotton" one. Therefore, combining shape and time information we can improve the time series classification. ## Detection of time series patterns with TWDTW @@ -229,7 +227,7 @@ matches = slotNames(matches) show(matches) ``` -To retrieve the complete information of the matches we set \code{keep=TRUE}. We need this information for the plot methods of the class \code{twdtwMatches}. The argument \code{weight.fun} defines the time-weight to the dynamic time warping analysis [@Maus:2016]. By default the time-weight is zero, meaning that the function will run a standard dynamic time warping analysis. The arguments \code{x} and \code{y} are objects of class \code{twdtwTimeSeries} with the unclassified long-term time series and the temporal patterns, respectively. For details and other arguments see \code{?twdtwApply}. +To retrieve the complete information of the matches we set \code{keep=TRUE}. We need this information for the plot methods of the class \code{twdtwMatches}. The argument \code{weight.fun} defines the time-weight to the dynamic time warping analysis [@Maus:2016]. By default the time-weight is zero, meaning that the function will run a standard dynamic time warping analysis. The arguments \code{x} and \code{y} are objects of class \code{twdtwTimeSeries} with the unclassified long-term time series and the temporal patterns, respectively. To perform the alignment between the time series the default TWDTW recursion has a symmetric step (for more details and other recursion options see \code{?stepPattern}). @Giorgino:2009 provides a detaild discussion on the recursion steps and other step patterns. For further details and other arguments of the TWDTW analysis see \code{?twdtwApply}. In our example we use a logistic weight function for the temporal constraint of the TWDTW algorithm. This function is defined by \code{logisticWeight}. The \pkg{dtwSat} package provides two in-built functions: \code{linearWeight} and \code{logisticWeight}. The \code{linearWeight} function with slope \code{a} and intercept \code{b} is given by $$ @@ -241,7 +239,7 @@ $$ \omega = \frac{1}{1 + e^{-\alpha(g(t_1,t_2)-\beta)} }. \label{eq:nonlineartw} $$ -The function $g$ is the absolute difference in days between two dates, $t_1$ and $t_2$. The linear function creates a strong time constraint even for small time differences. The logistic function has a low weight for small time warps and significant costs for bigger time warps, cf. Figure \ref{fig:logist-time-weight}. In our previous studies [@Maus:2016] the logistic-weight had better results than the linear-weight for land cover classification. Users can define different weight functions as temporal constraints in the argument \code{weight.fun} of the \code{twdtwApply} method. +The function $g$ is the absolute difference in days between two dates, $t_1$ and $t_2$. The aim of these functions is to control the time warp, e.g. a "large time warp" is needed to match a point of the temporal pattern whose original date is January 1 to a point of the long-term time series whose date is July 1, on the other hand to match January 1 to December 15 has a "small time warp". If there is a large seasonal difference between the pattern and its matching point in time series, an extra cost is added to the DTW distance measure. This constraint controls the time warping and makes the time series alignment dependent on the seasons. This is especially useful for detecting temporary crops and for distinguishing pasture from agriculture. The linear function creates a strong time constraint even for small time differences, including small time warps. The logistic function has a low weight for small time warps and significant costs for bigger time warps, cf. \autoref{fig:logist-time-weight}. In our previous studies [@Maus:2016] the logistic-weight had better results than the linear-weight for land cover classification. Users can define different weight functions as temporal constraints in the argument \code{weight.fun} of the function \code{twdtwApply}. ```{r logist-time-weight, echo = FALSE, eval = TRUE, out.width=paste0(page_width/2,'in'), fig.align='center', fig.cap='Logistic time-weight function \\code{logisticWeight} with steepness \\code{alpha=-0.1} and midpoint \\code{beta=100}. The $x$ axis shows the absolute difference between two dates in days and the $y$ axis shows the time-weight \\citep{Maus:2016}.', fig.pos='!h'} # Maximum time difference in days max_diff = 366/2 @@ -262,9 +260,9 @@ df_weight = melt(df_weight, id.vars = "Difference") names(df_weight)[-1] = c("Functions","Weight") ggplot(df_weight, aes_string(x="Difference", y="Weight", group="Functions", linetype="Functions")) + geom_line() + xlab("Time difference (days)") + - theme(text = element_text(size = 10), - plot.title = element_text(size = 10, face="bold"), - axis.title = element_text(size = 10), + theme(text = element_text(size = 10, family="Helvetica"), + plot.title = element_text(size = 10, family="Helvetica", face="bold"), + axis.title = element_text(size = 10, family="Helvetica"), legend.position = c(.3,.85), legend.background = element_rect(fill="transparent")) + scale_linetype(guide_legend(title = "")) ``` @@ -273,43 +271,44 @@ ggplot(df_weight, aes_string(x="Difference", y="Weight", group="Functions", line \pkg{dtwSat} provides five ways to visualise objects of class \code{twdtwMatches} through the plot types: \code{matches}, \code{alignments}, \code{classification}, \code{path}, and \code{cost}. The plot type \code{matches} shows the matching points of the patterns in the long-term time series; the plot type \code{alignments} shows the alignments and dissimilarity measures; the plot type \code{path} shows the low cost paths in the TWDTW cost matrix; and the plot type \code{cost} allows the visualisation of the cost matrices (local cost, accumulated cost, and time cost); and the plot type \code{classification} shows the classification of the long-term time series based on the TWDTW analysis. The plot methods for class \code{twdtwMatches} return a \code{ggplot} object, so that users can further manipulate the result using the \pkg{ggplot2} package. For more details on visualisation functions, please refer to the \pkg{dtwSat} documentation in the CRAN [@Maus:2015a]. -We now describe the plot types \code{matches} and \code{alignments}. The code bellow shows how to visualise the matching points of the four best matches of "soybean" pattern in the long-term time series, cf. Figure \ref{fig:twdtw-matches}. +We now describe the plot types \code{matches} and \code{alignments}. The code bellow shows how to visualise the matching points of the four best matches of "soybean" pattern in the long-term time series, cf. \autoref{fig:twdtw-matches}. ```{r twdtw-matches, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/3.5, fig.align='center', fig.cap=c('The four best matches of the "soybean" pattern in the time series using a logistic time-weight. The solid black line is the long-term time series; the coloured lines are the temporal patterns; and the grey dashed lines are the respective matching points.'), fig.pos='!h'} plot(matches, type="matches", patterns.labels="Soybean", k=4) ``` -The next example (Figure \ref{fig:alignments-all-patterns}) uses the plot type \code{alignments} to show the alignments of the temporal patterns. We set the threshold for the dissimilarity measure to be lower than $3.0$. This is useful to display the different subintervals of the long-term time series that have at least one alignment whose dissimilarity is less than the specified threshold. +The next example uses the plot type \code{alignments} to show the alignments of the temporal patterns (see \autoref{fig:alignments-all-patterns}). We set the threshold for the dissimilarity measure to be lower than $3.0$. This plot displays the different subintervals of the long-term time series that have alignments whose dissimilarity is less than the specified threshold. ```{r alignments-all-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.5, fig.align='center', fig.cap=c('Alignments and dissimilarity measures of the patterns "soybean", "cotton", and "maize" to the subintervals of the long-term time series using a logistic time-weight. The solid black line is the EVI time series, and the coloured lines are the alignments of the patterns that have dissimilarity measure lower than three.'), fig.pos='!h'} plot(matches, type="alignments", attr = "evi", threshold = 3.0) ``` +\autoref{fig:alignments-all-patterns} shows the alignments of each pattern over the long-term time series, note that we can rank the alignments by their TWDTW dissimilarity, i.e. alignments overlapping the same period usually have distinct dissimilarity, which can be used to rank them. In the figure we can see that maize (blue lines) and cotton (green lines) overlap approximately the same time periods, however, they have distinct dissimilarity measures, and therefore, can be ranked. Observing the time period from January 2010 to July 2010, both soybean, maize, and cotton have at least one overlapping alignment, however in this case the cotton pattern matches better to the interval because its dissimilarity is lower than the others. + ## Classifying the long-term time series -Using the matches and their associated dissimilarity measures, we can classify the subintervals of the long-term time series using \code{twdtwClassify}. To do this, we need to define a period for classification and the minimum overlap between the period and the alignments that intersect with it. We use the plot type \code{classification} to show the classification of the subintervals of the long-term time series based on the TWDTW analysis. For this example, we set classification periods of 6 months from September 2009 to September 2013, and a minimum overlap of 50% between the alignment and the classification period. This means that at least 50% of the alignment has to be contained inside of the classification period. -```{r time-series-classification, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.5, fig.align='center', fig.cap=c('Classification of each 6 months periods of the time series using results of the TWDTW analysis with logistic time-weight. The solid lines are the attributes of the time series, the background colours indicate the classification of the periods.'), fig.pos='!h'} +Using the matches and their associated dissimilarity measures, we can classify the subintervals of the long-term time series using \code{twdtwClassify}. To do this, we need to define a period for classification and the minimum overlap between the period and the alignments that intersect with it. For each interval, \code{twdtwClassify} will select the alignment that has the lowest TWDTW dissimilarity taking into account the minimum overlap condition. For example, in Figure 7 the interval from 1 September 1012 to 28 February 2013 has three overlapping alignments, maize in blue, cotton in green, and soybean in red. Without a minimum overlap the function \code{twdtwClassify} would classify this interval as maize, which has the lowest dissimilarity in the period. However, if we set a minimum overlap of 50\%, the function \code{twdtwClassify} classifies the interval as soybean, which is the only class whose alignment overlaps the interval during more than 50\% of the time. The interval of classification are usually defined according to the phenological cycles or the agricultural calendar of the region. The classification interval can also be irregular, for details see the argument \code{breaks} in \code{?twdtwClassify} + +In the example bellow we classify each period of 6 months from September 2009 to September 2013; we set a minimum overlap of 50% between the alignment and the classification period. This means that at least 50% of the alignment has to be contained inside of the classification period. We also use the plot type \code{classification} to show the classified subintervals of the long-term time series. + +```{r time-series-classification, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.8, fig.align='center', fig.cap=c('Classification of each 6 months periods of the time series using results of the TWDTW analysis with logistic time-weight. The solid lines are the attributes of the time series, the background colours indicate the classification of the periods.'), fig.pos='!ht'} ts_classification = twdtwClassify(x = matches, from = as.Date("2009-09-01"), to = as.Date("2013-09-01"), by = "6 month", overlap = 0.5) plot(ts_classification, type="classification") ``` -Comparing the results of the classified time series in Figure \ref{fig:time-series-classification} with the crop cycles in Figure \ref{fig:example-timeseries} we see that the algorithm has classified correctly all the eight subintervals from 2009 to 2013, which are, respectively: "Soybean", "Cotton", "Soybean", "Cotton", "Soybean", "Maize", "Soybean", "Maize". +By comparing the results of the classified time series in \autoref{fig:time-series-classification} with the crop cycles in \autoref{fig:example-timeseries} we see that the algorithm has classified correctly all the eight subintervals from 2009 to 2013, which are, respectively: "Soybean", "Cotton", "Soybean", "Cotton", "Soybean", "Maize", "Soybean", "Maize". -# Producing a land cover map - -In this section we present an application of TWDTW for land use and land cover change analysis using satellite image time series. Our input is a set of images, each covering the same geographical area at different times. Each pixel location is then associated to an unclassified satellite image time series. We assume to have done field work in the area; for some pixel locations and time periods, we know what is the land cover. We then will show how to obtain a set of template patterns, based on the field samples and how to apply these patterns to land cover classification of the set of images. In the end of this section we show how to perform land cover change analysis and how to do accuracy assessment. The satellite images and the field samples used in the examples come with \pkg{dtwSat} package. -Our method is not restricted to cases where the temporal patterns are obtained from the set of images. The patterns for the TWDTW analysis can be any time series with same bands or indices as the unclassified images, such as in the examples of Section \ref{classifying-a-time-series} above. -## Input data +# Producing a land cover map -The inputs are: *a)* the satellite images for a given geographical area, organised as a set of georeferenced raster files in GeoTIFF format, each containing all time steps of a spectral band or index; and *b)* a set of ground truth samples. The satellite images are extracted from the MODIS product MOD13Q1 collection 5 [@Friedl:2010] and include vegetation indexes "ndvi", "evi", and original bands "nir", "red", "blue", and "mir". This product has 250 x 250 m spatial and 16 day temporal resolution. +In this section we present an application of TWDTW for land cover change analysis using satellite image time series. Our input is a set of images, each covering the same geographical area at different times. Each pixel location is then associated to an unclassified satellite image time series. We assume to have done field work in the area; for some pixel locations and time periods, we know what is the land cover. We then will show how to obtain a set of template patterns, based on the field samples and how to apply these patterns to land cover classification of the set of images. In the end of this section we show how to perform land cover change analysis and accuracy assessment. -The region is a tropical forest area in Mato Grosso, Brazil of approximately 5300 km$^2$ with images from 2007 to 2013 (Figure \ref{fig:study-area}). This is a sequence of 160 images with 999 pixels each for 6 years. We also have a set of 603 ground truth samples of the following classes: "forest", "cotton-fallow", "soybean-cotton", "soybean-maize", and "soybean-millet". +As an example we classify approximately 5300 km$^2$ in a tropical forest region in Mato Grosso, Brazil (\autoref{fig:study-area}). This is a sequence of 160 images with 999 pixels each for 6 years, from 2007 to 2013. We also have a set of 603 ground truth samples of the following classes: "Forest", "Cotton-fallow", "Soybean-cotton", "Soybean-maize", and "Soybean-millet". The satellite images and the field samples used in the examples come with \pkg{dtwSat} package. -\begin{figure}[!h] +\begin{figure}[!ht] \begin{center} \includegraphics[width=\textwidth]{./study_area.pdf} \end{center} @@ -317,13 +316,17 @@ The region is a tropical forest area in Mato Grosso, Brazil of approximately 530 \label{fig:study-area} \end{figure} +## Input data + +The inputs are: *a)* the satellite images for a given geographical area, organised as a set of georeferenced raster files in GeoTIFF format, each containing all time steps of a spectral band or index; and *b)* a set of ground truth samples. The satellite images are extracted from the MODIS product MOD13Q1 collection 5 [@Friedl:2010] and include vegetation indices "ndvi", "evi", and original bands "nir", "red", "blue", and "mir". This product has 250 x 250 m spatial resolution and a 16 day maximum-value composite (MVC) for each pixel location [@Friedl:2010], meaning that one image can have measurements from different dates. For this reason, MOD13Q1 also includes the "day of the year" (doy) of each pixel as a layer, which we use to keep the time series consistent with the measurements. + The data files for the examples that follow are in the \pkg{dtwSat} installation folder *lucc_MT/data/*. The *tif* files include the time series of "blue", "red", "nir", "mir", "evi", "ndvi", and "doy" (day of the year); the text file *timeline* has the dates of the satellite images; the CSV file *samples.csv* has the \code{longitude}, \code{latitude}, \code{from}, \code{to}, and \code{label} for each field sample; and the text file *samples_projection* contains information about the cartographic projection of the samples, in the format of coordinate reference system used by \code{sp::CRS}. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} data_folder = system.file("lucc_MT/data", package = "dtwSat") dir(data_folder) ``` -In this example, we have stored all the time series for each band in one single file. In this way, we can use the function \code{raster::brick} to read the satellite images. The algorithm also works when the time steps for each band are split in many files. In this case, the user should call the function \code{raster::stack} with the appropriate parameters. Because of processing performance, we suggest that interested users group their images in bricks and follow the procedures given below. +We have stored all the time series for each band in one single file. In this way, we can use the function \code{raster::brick} to read the satellite images. The algorithm also works when the time steps for each band are split in many files. In this case, the user should call the function \code{raster::stack} with the appropriate parameters. Because of processing performance, we suggest that interested users group their images in bricks and follow the procedures given below. ```{r, echo = TRUE, eval = TRUE} blue = brick(paste(data_folder,"blue.tif", sep = "/")) red = brick(paste(data_folder,"red.tif", sep = "/")) @@ -335,10 +338,16 @@ day_of_year = brick(paste(data_folder,"doy.tif", sep = "/")) dates = scan(paste(data_folder,"timeline", sep = "/"), what = "dates") ``` -The set of ground truth samples in the CSV file *samples.csv* has a total of 603 samples divided in five classes: 68 "cotton-fallow", 138 "forest", 79 "soybean-cotton", 134 "soybean-maize", and 184 "soybean-millet". Reading this CSV file, we get a \code{data.frame} object, with the spatial location (\code{latitude} and \code{longitude}), starting and ending dates (\code{from} and \code{to}), and the \code{label} for each sample. +We use these data-sets to create a multiple raster time series, which is used in the next sections for the TWDTW analysis. \pkg{dtwSat} provides the constructor \code{twdtwRaster} that builds a multi-band satellite image time series. The inputs of this function are \code{RasterBrick} objects with the same temporal and spatial extents, and a vector (\code{timeline}) with the acquisition dates of the images in the format \code{"YYYY-MM-DD"}. The argument \code{doy} is combined with \code{timeline} to get the real date of each pixel, independently from each other. If \code{doy} is not provided then the dates of the pixels are given by \code{timeline}, i.e. all pixels in one image will have the same date. Products from other sensors, such as the Sentinels and Landsat, usually have all pixels with same date, therefore the argument \code{doy} is not needed. This function produces an object of class \code{twdtwRaster} with the time series of multiple satellite bands. +```{r, echo = TRUE, eval = TRUE} +raster_timeseries = twdtwRaster(blue, red, nir, mir, evi, ndvi, + timeline = dates, doy = day_of_year) +``` + +Our second input is a set of ground truth samples in the CSV file *samples.csv*, which has a total of 603 samples divided in five classes: 68 "cotton-fallow", 138 "forest", 79 "soybean-cotton", 134 "soybean-maize", and 184 "soybean-millet". Reading this CSV file, we get a \code{data.frame} object, with the spatial location (\code{latitude} and \code{longitude}), starting and ending dates (\code{from} and \code{to}), and the \code{label} for each sample. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} field_samples = read.csv(paste(data_folder,"samples.csv", sep = "/")) -head(field_samples, 2) +head(field_samples, 5) table(field_samples[["label"]]) proj_str = scan(paste(data_folder,"samples_projection", sep = "/"), what = "character") @@ -346,186 +355,151 @@ proj_str ``` -## Creating the time series and the temporal patterns +## Assessing the separability of temporal patterns -After reading our data, we need to create the time series for analysis. For this purpose, \pkg{dtwSat} provides the constructor \code{twdtwRaster} that builds a multi-band satellite image time series. The inputs of this function are \code{RasterBrick} objects with the same temporal and spatial extents, and a vector (\code{timeline}) with the acquisition dates of the images in the format \code{"YYYY-MM-DD"}. The argument \code{doy} is optional. If \code{doy} is not declared, the function builds a \code{RasterBrick} object using the dates given by \code{timeline}. This function produces an object of class \code{twdtwRaster} with the time series of multiple satellite bands. -```{r, echo = TRUE, eval = TRUE} -raster_timeseries = twdtwRaster(blue, red, nir, mir, evi, ndvi, - timeline = dates, doy = day_of_year) -``` +The classification is highly dependent on the quality of the temporal patterns. Therefore, it is useful to perform an analysis to assess the separability of the temporal pattern. Ideally, one would need patterns that, when applied to the set of unknown time series, produce consistent results (see the guidelines for single time series analysis in \autoref{classifying-a-time-series}). For this reason, before performing the land cover mapping, we perform a cross validation step. In this way, the users would assess the separability of their patterns before classifying a large data set. -We now need to identify the temporal patterns. Usually, this can be done using the collected field samples. In the next example we use the function \code{getTimeSeries} to get the time series of each field sample from our raster time series. The arguments of the function are a set of raster time series, a \code{data.frame} with spatial and temporal information about the fields samples (as in the object \code{field_samples} given above), and a \code{proj4string} with the projection information. The projection should follow the \code{sp::CRS} format. The result is an object of class \code{twdtwTimeSeries} with one time series for each field sample. +In the next example we use the function \code{getTimeSeries} to extract the time series of each field sample from our raster time series. The arguments of the function are a set of raster time series, a \code{data.frame} with spatial and temporal information about the fields samples (as in the object \code{field_samples} given above), and a \code{proj4string} with the projection information. The projection should follow the \code{sp::CRS} format. The result is an object of class \code{twdtwTimeSeries} with one time series for each field sample. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} field_samples_ts = getTimeSeries(raster_timeseries, y = field_samples, proj4string = proj_str) field_samples_ts ``` -After obtaining the time series associated to the field samples, we need to create the template patterns for each class. For this purpose, \pkg{dtwSat} provides the function \code{createPatterns}. This function fits a Generalized Additive Model (GAM) [Hastie:1986,Wood:2011] to the field samples and retrieves a smoothed temporal pattern for each band (*e.g.*, "blue", "red", "nir", "mir", "evi", and "ndvi"). We use the GAM because of its flexibility for non-parametric fits, with less rigorous assumptions on the relationship between response and predictor. This potentially provides better fit to satellite data than purely parametric models, due to the data's inter- and intra-annual variability. +To perform the cross-validation we use the function \code{twdtwCrossValidate}. This function splits the sample time series into training and validation sets using stratified sampling with a simple random sampling within each stratum, for details see \code{?caret::createDataPartition}. The function uses the training samples to create the temporal patterns and then classifies the remaining validation samples using \code{twdtwApply}. The results of the classification are used in the accuracy calculation. + +A Generalized Additive Model (GAM) [Hastie:1986,Wood:2011] generates the smoothed temporal patterns based on the training samples. We use the GAM because of its flexibility for non-parametric fits, with less rigorous assumptions on the relationship between response and predictor. This potentially provides better fit to satellite data than purely parametric models, due to the data's inter- and intra-annual variability. + +In the next example we set the arguments \code{times=100} and \code{p=0.1}, which creates 100 different data partitions, each with 10% of the samples for training and 90% for validation. The other arguments of this function are: the logistic weight function with steepness `-0.1` and midpoint `50` to \code{weight.fun}; the frequency of the temporal patterns to 8 days \code{freq=8}, and GAM smoothing formula to \code{formula = y ~ s(x)}, where function \code{s} sets up a spline model, with \code{x} the time and \code{y} a satellite band (for details see \code{?mgcv::gam} and \code{?mgcv::s}). The output is an object of class \code{twdtwCrossValidation} which includes the accuracy for each data partition. The object has two slots, the first called \code{partitions} has the index of the samples used for training, the second called \code{accuracy} has overall accuracy, user's accuracy, producer's accuracy, error matrix, and the data used in the calculation, i.e. reference classes, predicted classes, and TWDTW distance measure. + +```{r, echo = TRUE, eval = FALSE, results = 'markup'} +set.seed(1) +log_fun = logisticWeight(alpha=-0.1, beta=50) +cross_validation = twdtwCrossValidate(field_samples_ts, times=100, p=0.1, + freq = 8, formula = y ~ s(x, bs="cc"), weight.fun = log_fun) +``` +```{r, echo = FALSE, eval = TRUE} +load(system.file("lucc_MT/cross_validation.RData", package = "dtwSat")) +``` + +\autoref{fig:plot-accuracy} and \autoref{tab:cross-validation} show the 95% confidence interval of the mean for user\'s and producer\'s accuracy derived from the hundred-fold cross-validation analysis. The user\'s accuracy gives the confidence and the producer\'s accuracy gives the sensitivity of the method for each class. In our analysis all classes had high user\'s and producer\'s accuracy, meaning that TWDTW has high confidence and sensitivity for the classes included in the analysis. The cross-validation results show that if we randomly select 10% of our sampels to create temporal patterns we can get an overall accuracy of at least 97% in the classification of the remaining samples with 95% confidence level. +```{r plot-accuracy, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_width/2, fig.align='center', fig.cap='User\'s and producer\'s accuracy of the TWDTW cross-validation using 100 resampling-with-replacement. The plot shows the 95\\% confidence interval of the mean.', fig.pos='!ht'} +plot(cross_validation, conf.int=.95) +``` + +```{r, echo = FALSE, eval = TRUE, results = 'asis'} +twdtwXtable(cross_validation, conf.int=.95, digits = 2, caption="\\label{tab:cross-validation} User\'s and producer\'s accuracy of the TWDTW cross-validation using 100 resampling-with-replacement. The table shows the standard deviation ($\\sigma$) and the 95\\% confidence interval (ci) of the mean ($\\mu$).'", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` + + + +## Creating temporal patterns + +In the last section we observed that the land cover classes based on our samples are separable using the TWDTW algorithm with high confidence level. Now we randomly select 10% of our samples for training and keep the remaining 90% for validation. The first set of samples are used to create temporal patterns and classify the raster time series, and the second set of samples to assess the final maps. Ideally we would need a second independent set of samples to assess the map, but it would be very difficult to identify different crops without field work. Therefore, we use the same samples used in the cross-validation (\autoref{assessing-the-separability-of-temporal-patterns}). +```{r, echo = TRUE, eval = TRUE} +library(caret) +set.seed(1) +I = unlist(createDataPartition(field_samples[,"label"], p = 0.1)) +training_ts = subset(field_samples_ts, I) +validation_samples = field_samples[-I,] +``` + +We use the function \code{createPatterns} to produce the temporal patterns based on the training samples. For that, we need to set the desired temporal frequency of the patterns and the smoothing function for the GAM model. In the example below, we set \code{freq=8} to get temporal patterns with a frequency of 8 days, and the GAM smoothing formula \code{formula = y ~ s(x)}, such as in \autoref{assessing-the-separability-of-temporal-patterns}). -To produce the set of template patterns using the function \code{createPatterns}, we need to set the temporal frequency of the resulting patterns and the smoothing function for the GAM model. In the example below, we set \code{freq=8} to get temporal patterns with a frequency of 8 days. We also set the GAM smoothing formula to be \code{formula = y ~ s(x)}, where function \code{s} sets up a spline model, with \code{x} the time and \code{y} a satellite band (for details see \code{?mgcv::gam} and \code{?mgcv::s}). ```{r, echo = TRUE, eval = TRUE} temporal_patterns = - createPatterns(field_samples_ts, freq = 8, formula = y ~ s(x)) + createPatterns(training_ts, freq = 8, formula = y ~ s(x)) ``` -We use the plot method \code{type="patterns"} to show the results of the \code{createPatterns} in \autoref{fig:temporal-patterns}. -```{r temporal-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_width/1.5, fig.align='center', fig.pos='!h', fig.cap='Temporal patterns of forest, cotton-fallow, soybean-cotton, soybean-maize, and soybean-millet based on the ground truth samples.'} +The result of the function \code{createPatterns} is an object of the class \code{twdtwTimeSeries}. We use the plot method \code{type="patterns"} to show the results of the \code{createPatterns} in \autoref{fig:temporal-patterns}. +```{r temporal-patterns, echo = TRUE, eval = TRUE, fig.width=page_width, fig.height=page_width/1.5, fig.align='center', fig.pos='!h', fig.cap='Temporal patterns of Forest, Cotton-fallow, Soybean-cotton, Soybean-maize, and Soybean-millet based on the ground truth samples.'} plot(temporal_patterns, type = "patterns") + theme(legend.position = c(.8,.25)) ``` -After obtaining the template patterns for each land cover class, it is useful to perform a pre-classification analysis to assess their quality and their informational content. Ideally, one would need template patterns that, when applied to the set of unknown time series, produce consistent results. For this reason, it is advisable that the user performs a pre-classification step, along the lines of the individual analysis described in Section \ref{classifying-a-time-series}. In this way, the users would assess how good their patterns are before classifying a large data set. +Our method is not restricted to cases where the temporal patterns are obtained from the set of images, such as in the example above. Once can also use patterns from a different set of images or defined in other studies, as long as these temporal patterns stand for the study area and their bands match the bands in the unclassified images. ## Classifying the image time series -After obtaining a consistent set of temporal patterns, we use the function \code{twdtwApply} to run the TWDTW analysis for each pixel location in the raster time series. The input raster time series in the object \code{twdtwRaster} should be longer or have approximatly the same length as the temporal patterns. This function retrieves an object of class \code{twdtwRaster} with the TWDTW dissimilarity measure of the temporal patterns for each time period. The arguments \code{overwrite} and \code{format} are passed to \code{raster::writeRaster}. The arguments \code{weight.fun} and \code{overlap} are described in Section \ref{classifying-a-time-series}. Here we set the parameters of the time weight (logistic function) base on our the experience about the phenological cycle of the vegetation in the study area. In the next example, we classify the raster time series using the temporal patterns in \code{temporal_patterns} obtained as described above. The result is a \code{twdtwRaster} with five layers; each layer contains the TWDTW dissimilarity measure for one temporal pattern over time. We use the plot type \code{distance} to illustrate the TWDTW dissimilarity for each temporal pattern in 2013, cf. Figure \ref{fig:plot-dissmilarity2013}. +After obtaining a consistent set of temporal patterns, we use the function \code{twdtwApply} to run the TWDTW analysis for each pixel location in the raster time series. The input raster time series in the object \code{twdtwRaster} should be longer or have approximatly the same length as the temporal patterns. This function retrieves an object of class \code{twdtwRaster} with the TWDTW dissimilarity measure of the temporal patterns for each time period. The arguments \code{overwrite} and \code{format} are passed to \code{raster::writeRaster}. The arguments \code{weight.fun} and \code{overlap} are described in \autoref{classifying-a-time-series}. Here we set the parameters of the time weight (logistic function) base on our the experience about the phenological cycle of the vegetation in the study area. In the next example, we classify the raster time series using the temporal patterns in \code{temporal_patterns} obtained as described above. The result is a \code{twdtwRaster} with five layers; each layer contains the TWDTW dissimilarity measure for one temporal pattern over time. We use the plot type \code{distance} to illustrate the TWDTW dissimilarity for each temporal pattern in 2008, cf. \autoref{fig:plot-dissmilarity2008}. ```{r, echo = TRUE, eval = TRUE, results = 'markup'} log_fun = logisticWeight(alpha=-0.1, beta=50) twdtw_dist = twdtwApply(x = raster_timeseries, y = temporal_patterns, overlap = 0.5, weight.fun = log_fun, overwrite=TRUE, format="GTiff") ``` - -```{r plot-dissmilarity2013, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Illustration of the TWDTW dissimilarity from each temporal pattern in 2013. The blue areas are more similar to the pattern and the red areas are less similar to the pattern.', fig.pos='!h'} -plot(x = twdtw_dist, type="distance", time.levels = 6) +```{r plot-dissmilarity2008, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Illustration of the TWDTW dissimilarity from each temporal pattern in 2008. The blue areas are more similar to the pattern and the red areas are less similar to the pattern.', fig.pos='!ht'} +plot(x = twdtw_dist, type="distance") ``` -The results of the example above can be used to create categorical land cover maps. The function \code{twdtwClassify} selects the most similar pattern for each time period and retrieves a \code{twdtwRaster} object with the time series of land use maps. The resulting object includes two layers, the first has the classified categorical maps and the second has the TWDTW dissimilarity measure. -```{r, echo = TRUE, eval = TRUE} -land_use_maps = twdtwClassify(twdtw_dist, format="GTiff", overwrite=TRUE) +The results of the example above can be used to create categorical land cover maps. The function \code{twdtwClassify} selects the most similar pattern for each time period and retrieves a \code{twdtwRaster} object with the time series of land cover maps. The resulting object includes two layers, the first has the classified categorical maps and the second has the TWDTW dissimilarity measure. +```{r, echo = TRUE, eval = TRUE, results = 'markup'} +land_cover_maps = twdtwClassify(twdtw_dist, format="GTiff", overwrite=TRUE) ``` ## Looking at the classification results -The classification results can be visualised using the \code{plot} methods of the class \code{twdtwRaster}, which supports four plot types: "maps", "area", "changes", and "distance". The \code{type="maps"} shows the land cover classification maps for each period, cf. Figure \ref{fig:plot-map}. -```{r plot-map, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Land use maps for each year from 2008 to 2013.', fig.pos='!h'} -plot(x = land_use_maps, type = "maps") +The classification results can be visualised using the \code{plot} methods of the class \code{twdtwRaster}, which supports four plot types: "maps", "area", "changes", and "distance". The \code{type="maps"} shows the land cover classification maps for each period, cf. \autoref{fig:plot-map}. +```{r plot-map, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Land cover maps for each year from 2008 to 2013.', fig.pos='!ht'} +plot(x = land_cover_maps, type = "maps") ``` -The next example shows the accumulated area for each class over time, using \code{type="area"}, cf. Figure \ref{fig:plot-area}. -```{r plot-area, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Percentage of area for each land use class from 2008 to 2013.', fig.pos='!h'} -plot(x = land_use_maps, type = "area") +The next example shows the accumulated area for each class over time, using \code{type="area"}, cf. \autoref{fig:plot-area}. +```{r plot-area, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Percentage of area for each land cover class from 2008 to 2013.', fig.pos='!ht'} +plot(x = land_cover_maps, type = "area") ``` -Users can also view the land cover transition for each time period, by setting \code{type="changes"}. For each land cover class and each period, the plot shows gains and losses in area from the other classes. This is the visual equivalent of a land transition matrix, cf. Figure \ref{fig:plot-change}. +Users can also view the land cover transition for each time period, by setting \code{type="changes"}. For each land cover class and each period, the plot shows gains and losses in area from the other classes. This is the visual equivalent of a land transition matrix, cf. \autoref{fig:plot-change}. ```{r plot-change, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Gains and losses in area from the other classes. The $y$ axis shows the actual class; the positive direction of $x$ axis shows the gains and the negative direction of $x$ axis shows the losses of the classes indicated in $y$. The colors indicate from/to which classes the gains/losses belong.', fig.pos='!h'} -plot(x = land_use_maps, type = "changes") +plot(x = land_cover_maps, type = "changes") ``` -We can also look at the dissimilarity of each classified pixel setting \code{type="distance"}. This plot can give a measure of the uncertainty of the classification of each pixel for each time period, cf. Figure \ref{fig:plot-dissmilarity}. -```{r plot-dissmilarity, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='TWDTW dissimilarity measure for each pixel over each classified period. The blue areas have high confidence and the red areas have low confidence in the classification.', fig.pos='!h'} -plot(x = land_use_maps, type="distance") +We can also look at the dissimilarity of each classified pixel setting \code{type="distance"}. This plot can give a measure of the uncertainty of the classification of each pixel for each time period, cf. \autoref{fig:plot-dissmilarity}. +```{r plot-dissmilarity, echo = TRUE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='TWDTW dissimilarity measure for each pixel over each classified period. The blue areas have high confidence and the red areas have low confidence in the classification.', fig.pos='!ht'} +plot(x = land_cover_maps, type="distance") ``` -## Assessing the classification accuracy -In this section we show how to assess the accuracy of the TWDTW method for land cover classification. To do this, we split the ground truth samples into training and validation sets, using the function \code{splitDataset} from the package \pkg{dtwSat}. This function splits set of time series in the object \code{twdtwTimeSeries} for training and validation. The argument \code{p} defines the percentage used for training and the argument \code{times} gives the number of different partitions to create. This is a a stratified sampling with a simple random sampling within each stratum, see \code{?createDataPartition} for details. In the next example we create 100 different partitions of the data. Each partition uses 10% of the data for training and 90% for validation. The output is a list with 100 different data partitions; each partition has the temporal patterns based on the training samples and a set of time series for validation. -```{r, echo = TRUE, eval = FALSE} -set.seed(1) -partitions = splitDataset(field_samples_ts, p=0.1, times=100, - freq = 8, formula = y ~ s(x, bs="cc")) -``` -For each data partition we run the TWDTW analysis to classify the set of validation time series using the trained temporal patterns. The result is a list of \code{twdtwMatches} objects with the classified set of time series for each data partition. To compute the *User's Accuracy* (UA) and *Producer's Accuracy* (PA) of the classified time series we use the function \code{dtwSat::twdtwCrossValidation} that retrieves a \code{data.frame} with the accuracy assessment for all data partitions. -```{r, echo = TRUE, eval = FALSE, results = 'markup'} -log_fun = logisticWeight(alpha=-0.1, beta=50) -twdtw_res = lapply(partitions, function(x){ - res = twdtwApply(x = x$ts, y = x$patterns, weight.fun = log_fun, n=1) - twdtwClassify(x = res) -}) -assessment = twdtwCrossValidation(twdtw_res) -head(assessment, 5) -``` -```{r, echo = FALSE, eval = TRUE} -load(system.file("lucc_MT/cross_validation.RData", package = "dtwSat")) -``` - - -Figure \ref{fig:plot-accuracy} shows the average $\mu$ and standard deviation $\sigma$ of *user\'s* and *producer\'s accuracy* based on a bootstrap simulation of 100 different data partitions using resampling-with-replacement. The *user\'s accuracy* gives the confidence and the *producer\'s accuracy* gives the sensitivity of the method for each class. In our analysis all classes had high *user\'s* and *producer\'s accuracy*, meaning that TWDTW has high confidence and sensitivity for the classes included in the analysis. The average, standard deviation, and the 99\% confidence interval is also shown in Table \ref{tab:assessment}. -```{r plot-accuracy, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_width/2, fig.align='center', fig.cap='User\'s Accuracy (UA) and Producer\'s Accuracy (PA) of the TWDTW method for land cover classification. The plot shows the averages and their confidence interval for 99\\%.', fig.pos='!h'} -df = melt(assessment[,-1], id="label") -df$variable = factor(df$variable, levels = c("UA", "PA"), labels = c("User's Accuracy", "Producer's Accuracy")) -ggplot(df, aes(x=label, y=value)) + - stat_summary(fun.data="mean_cl_boot", fun.args=list(conf.int = .99), - width=0.5, geom="crossbar", size=0.1, fill = "gray") + - geom_point(size=0.2) + facet_grid(. ~ variable) + - scale_y_continuous(limits = c(0,1), labels = percent, breaks = seq(0,1,.2)) + - xlab("") + ylab("Accuracy") + coord_flip() +In this section we show how to assess the classification. \pkg{dtwSat} provides a function called \code{twdtwAssess}, which computes a set of accuracy metrics, and adjusted area such as proposed by @Olofsson:2013 and @Olofsson:2014. The inputs of this function are the classified map (an object of class \code{twdtwRaster}), and a set of samples for validation (an object of class \code{data.frame} or \code{sp::SpatialPointsDataFrame}). Besides coordinates, the samples should also have starting dates, ending dates, and lables compatible with the labels in the map (for details see \autoref{input-data}). The output of \code{twdtwAssess} is an object of class \code{twdtwAssessment}, which includes four slots: 1) \code{accuracyByPeriod} is a list of metrics for each time period, including overall accuracy, user's accuracy, produce's accuracy, error matrix (confusion matrix), and adjusted area; 2) \code{accuracySummary} has the accuracy and adjusted area accumulated over all time periods; 3) \code{data} is a \code{SpatialPointsDataFrame} with sample ID, period ID, starting date, ending date, reference label, predicted label, and TWDTW distance; and 4) \code{map} is a twdtwRaster with the raster maps. The next example uses \code{twdtwAssess} to compute the accuracy of the maps (\code{land_cover_maps}) using the validation samples (\code{validation_samples}) with a 95% confidence level. +```{r, echo = TRUE, eval = TRUE} +maps_assessment = twdtwAssess(land_cover_maps, y = validation_samples, + proj4string = proj_str, conf.int=.95) ``` + + +The results of the assessment in \autoref{tab:map-error-matrix}, \ref{tab:map-accuracy}, and \ref{tab:map-adjusted-area} are accumulated over the whole time period, i.e. the total mapped area is equal to the surface area times the number of maps. It is possible to assess and visualise each period independently from each other. However, our samples are irregularly distributed over time and some classes do not have samples in all time period, which limits the analysis of each time period independently from each other. + ```{r, echo = FALSE, eval = TRUE, results = 'asis'} -assess_mean = aggregate(assessment[, c("UA","PA")], list(assessment$label), mean) -assess_sd = aggregate(assessment[, c("UA","PA")], list(assessment$label), sd) -l_names = levels(assessment$label) -names(l_names) = l_names -ic_ua = t(sapply(l_names, function(i) 100*mean_cl_boot(x = assessment$UA[assessment$label==i], conf.int = .99)))[,-1] -ic_pa = t(sapply(l_names, function(i) 100*mean_cl_boot(x = assessment$PA[assessment$label==i], conf.int = .99)))[,-1] +twdtwXtable(maps_assessment, table.type="errormatrix", digits = 0, rotate.col = TRUE, caption="\\label{tab:map-error-matrix}Error matrix of the map classification based on TWDTW analysis. The area is in the map unit, in this case $m^2$. $w$ is the proportion of area mapped for each class.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` -assess_table = data.frame( - Class = assess_mean$Group.1, - - MUA = sprintf("%.2f", round(100*assess_mean$UA,2)), - SDUA = sprintf("(%.2f)", round(100*assess_sd$UA,2)), - CIUA = sprintf("[%.2f-%.2f]", round(as.numeric(ic_ua[,1]),2), round(as.numeric(ic_ua[,2]),2)), +As we can see in \autoref{tab:map-error-matrix} only nine samples were misclassified, all of them from the reference class "Soybean-cotton". From these samples six were classified as "Soybean-maize", and three as "Cotton-fallow". As we see in \autoref{tab:map-accuracy} the only class with producer\'s accuracy lower than $100\%$ was "Soybean-cotton", reaching $72\%$ with high uncertainty ($\pm13\%$). The user\'s accuracy for all classes was higer than $95\%$, with maximun uncertainty of $\pm5\%$. To visualise the misclassified samples on top of the maps we can use the plot \code{type="map"} for objects of class \code{twdtwAssessment}, such that `plot(x = maps_assessment, type="map", samples="incorrect")`. The user can also set the argument \code{samples} to see correctly classified samples \code{samples="correct"}, or to see all samples \code{samples="all"}. - MPA = sprintf("%.2f", round(100*assess_mean$PA,2)), - SDPA = sprintf("(%.2f)", round(100*assess_sd$PA,2)), - CIPA = sprintf("[%.2f-%.2f]", round(as.numeric(ic_pa[,1]),2), round(as.numeric(ic_pa[,2]),2)) - ) +The \autoref{fig:plot-map-incorrect-samples} shows that the misclassified samples are all in the map from 2012. The six samples of "Soybean-cotton" classified as "Soybean-maize" are within a big area of "Soybean-maize" and the three samples of "Soybean-cotton" classified as "Cotton-fallow" are near the border between this two classes. This errors might be related to the mixture of different classes in the same pixel. -x_assess = xtable::xtable(assess_table, - format = tab_format, digits = 2, label = "tab:assessment", alig=c("l","c","c","c","c","c","c","c"), - caption="User\'s and Producer\'s Accuracy of the land use classification based on TWDTW analysis. $\\mu$ is the average accuracy, $\\sigma$ the standard deviation, and CI is the confidence interval of 99\\% using 100 resampling-with-replacement.") +```{r plot-map-incorrect-samples, echo = FALSE, eval = TRUE, fig.width=page_width, fig.align='center', fig.cap='Incorrect classified samples.', fig.pos='!ht'} +plot(x = maps_assessment, type="map", samples="incorrect") +``` + +```{r, echo = FALSE, eval = TRUE, results = 'asis'} +twdtwXtable(maps_assessment, table.type="accuracy", show.prop = TRUE, digits = 2, rotate.col = TRUE, caption="\\label{tab:map-accuracy}Accuracy and error matrix in proportion of area of the classified map.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` + +In \autoref{tab:map-adjusted-area} we can see the mapped and the adjusted area. This is the accumulated area over the whole period, i.e. the sum of all maps from 2008 to 2013. As the "Forest" and "Soybean-millet" did not have omission ($100\%$ producer's accuracy) or comission ($100\%$ user's accuracy) erros, we immediately see that their mapped area is equal to their adjusted area (\autoref{tab:map-adjusted-area}). To help the analysis of the other classes we use the plot \code{type="area"} for class \code{twdtwAssessment}, such that `plot(x = maps_assessment, type="area", perc=FALSE)`. \autoref{fig:plot-area-and-uncertainty} shows the accumulated area mapped and adjusted for all classes. In this figure we see that our method overestimated the area of "Soybean-maize", i.e. the mapped area ($110173564\;m^2$) is bigger than the adjusted area ($104927204\;m^2$) plus the confidence interval $4113071\;m^2$. Meanwhile we underestimated the area of "Soybean-cotton", i.e. its mapped area ($18782634\;m^2$) is smaller than the adjusted area ($26260270\;m^2$) plus the confidence interval ($4805205\;m^2$). The mapped area of "Cotton-fallow" ($47600561\;m^2$) is within the confidence interval of the adjusted area ($45369285\pm2484480\;m^2$). To improve the accuracy assessment and area estimations the field samples should be better distributed over time, which would also allow for better land cover changes assessment. -addtorow = list() -addtorow$pos = list(0) -addtorow$command = paste("Class & \\multicolumn{3}{c}{User's Accuracy (UA) \\%} & \\multicolumn{3}{c}{Producer's Accuracy (PA)\\%}\\\\", paste(c("","$\\mu$","$\\sigma$","CI","$\\mu$","$\\sigma$","CI"), collapse="&"), "\\\\", collapse = "") +```{r, echo = FALSE, eval = TRUE, results = 'asis'} +twdtwXtable(maps_assessment, table.type="area", digits = 0, rotate.col = TRUE, caption="\\label{tab:map-adjusted-area}Mapped and adjusted, accumulated over the whole period, i.e. the sum from the sum of the maps from 2008 to 2013. The area is in the map unit, in this case $m^2$.", comment = FALSE, caption.placement = "bottom", table.placement="!ht") +``` -xtable::print.xtable(x_assess, add.to.row=addtorow, include.colnames = FALSE, include.rownames = FALSE, - comment = FALSE, caption.placement = "bottom") +```{r plot-area-and-uncertainty, echo = FALSE, eval = TRUE, fig.width=page_width, fig.height=page_height/2.7, fig.align='center', fig.cap='Mapped and adjusted, accumulated over the whole period, i.e. the sum from the sum of the maps from 2008 to 2013. The area is in the map unit, in this case $m^2$.', fig.pos='!ht'} +plot(x = maps_assessment, type="area", perc=FALSE) ``` @@ -533,20 +507,24 @@ xtable::print.xtable(x_assess, add.to.row=addtorow, include.colnames = FALSE, in # Conclusions and Discussion -Nowadays, there are large open archives of Earth Observation data, but few open source methods for analysing them. With this motivation, this paper provides guidance on how to use the Time-Weighed Dynamic Time Warping (TWDTW) method for remote sensing applications. As we have discussed in a companion paper [@Maus:2016], the TWDTW method is well suited for land cover change analysis of satellite image time series. +The overall accuracy of the classification with a 95% confidence level is within 97% and 99%. With same confidence level, user's and producer's accuracy are between 90% and 100% for all classes, except for "Soybean-cotton", which has wide confidence interval for user's accuracy, between 59% and 85%. A small sample size will likely have large confidence intervals [@Foody:2009], therefore, this analysis could be improved by increasing the number of "Soybean-cotton" samples. In addition, our access to field information is limited to a set of samples irregularly distributed over time, which are not enough to assess each mapped period independently from each other. Nevertheless, the results of the accuracy assessment show that the TWDTW has great potential to classify different crop types. -The main goal of \pkg{dtwSat} package is to make TWDTW accessible for researchers. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and evaluating the results. The current version of the \pkg{dtwSat} package provides a pixel-based time series classification method. We envisage that future versions of the package could include local neighborhoods to reduce border effects and improve classification homogeneity. +DTW based approaches have achieved good results for land cover classification [@Petitjean:2012; @Maus:2016], however, a reduced number of points in the time series will negatively impact the accuracy. Remotely sensed images often present noise and poor coverage due to clouds, aerosol load, surface directional effects, and sensor problems. This leads to large amount of gaps in satellite image time series. Therefore, methods that deal with irregular temporal sampling, i.e. irregular sampling intervals, have great potential to fully exploit the available satellite images archive. DTW is known to be one of the most robust methods for irregular time series [@Keogh:2005; @Tormene:2009]. It was successfully applied for satellite time series clustering using FORMOSAT-2 [@Petitjean:2012] and using MODIS [@Maus:2016]. @Petitjean:2012, for example, showed that clustering based on DTW is consistent even when there are several images missing per year because of cloud cover. However, the effect of a reduced number of samples in the time needs to be better evaluated in order to point out the limiting gap size for satellite image time series analysis using DTW based methods. -The \pkg{dtwSat} package provides two in-built functions for linear and logistic time weight. In the current version of the package the parameters of the weight functions are set manually to the same value for all land use/cover classes. Future versions of the package could include methods to search for the best parameters to be set class-by-class using field data. +The DTW approaches will search for the matches of a temporal pattern, therefore the number of points in the time series should represent the phenological cycles of different land cover types. The number of available observations might be a limitation for sensors with lower temporal resolution, such as Landsat. We believe that this limitation could be addressed, for example, by combining TWDTW analysis with pixel-based compositing techniques [@Griffiths:2013; @White:2014]. These approaches have become more popular with the opening of the USGS Landsat archive and could be used to increase the availability of gap-free time series [@Gomez:2016]. -To aim for maximum usage by the scientific community, the \pkg{dtwSat} package described in this paper works with well-known R data classes such as provided by packages \pkg{zoo} and \pkg{raster}. We are planning improvements, so that \pkg{dtwSat} can be combined with array databases, such as SciDB [@Stonebraker:2013]. We believe that combining array databases with image time series analysis software such as presented here is one way forward to scaling the process of information extracting to very large Earth Observation data. +\pkg{dtwSat} provides a dissimilarity measure in the n-dimensional space, allowing multispectral satellite image time series analysis. Our experience using MODIS data sets has shown that n-dimensional analysis, i.e. using RED, NIR, NDVI, EVI, and NDVI, increases the separability among classes when compared to single band analysis, for example using only EVI or NDVI. Further studies on multispectral data using TWDTW analysis will help to optimize the selection of bands. +The main goal of \pkg{dtwSat} package is to make TWDTW accessible for researchers. The package supports the full cycle of land cover classification using image time series, ranging from selecting temporal patterns to visualising and assessing the results. The current version of the \pkg{dtwSat} package provides a pixel-based time series classification method. We envisage that future versions of the package could include local neighborhoods to reduce border effects and improve classification homogeneity. +The \pkg{dtwSat} package provides two in-built functions for linear and logistic time weight. In the current version of the package the parameters of the weight functions are set manually to the same value for all land cover classes. Future versions could include methods to search for the best parameters for each land cover class using field data. +Nowadays, there are large open archives of Earth Observation data, but few open source methods for analysing them. With this motivation, this paper provides guidance on how to use the Time-Weighed Dynamic Time Warping (TWDTW) method for remote sensing applications. As we have discussed in a companion paper [@Maus:2016], the TWDTW method is well suited for land cover change analysis of satellite image time series. -\section*{Acknowledgments} -Victor Maus has been supported by the Institute for Geoinformatics, University of Münster (Germany), and by the Earth System Science Center, National Institute for Space Research (Brazil). Part of the research was developed in the Young Scientists Summer Program at the International Institute for Applied Systems Analysis, Laxenburg (Austria). Gilberto Câmara's term as Brazil Chair at IFGI has been supported by CAPES (grant 23038.007569\/2012-16). Gilberto's work is also supported by FAPESP e-science program (grant 2014-08398-6) and CNPq (grant 312151\/2014-4). +The TWDTW algorithm is computationally intensive and for large areas one should consider parallel processing. The algorithm is pixel time series based, i.e. each time series is processed independently from each other, and therefore, it can be easily parallelized. To aim for maximum usage by the scientific community, the \pkg{dtwSat} package described in this paper works with well-known \proglang{R} data classes such as provided by \pkg{raster}, which offers the option to work with raster data sets stored on disk that are too large to be loaded into memory (RAM) at once [@Hijmans:2015]. We are planning improvements, so that \pkg{dtwSat} can also be combined with array databases, such as SciDB [@Stonebraker:2013]. We believe that combining array databases with image time series analysis software such as presented here is one way forward to scaling the process of information extracting from very large Earth Observation data. - + +\section*{Acknowledgments} +Victor Maus has been supported by the Institute for Geoinformatics, University of Münster (Germany), and by the Earth System Science Center, National Institute for Space Research (Brazil). Gilberto Camara's term as Brazil Chair at IFGI has been supported by CAPES (grant 23038.007569\/2012-16). Gilberto's work is also supported by FAPESP e-science program (grant 2014-08398-6) and CNPq (grant 312151\/2014-4). diff --git a/vignettes/references.bib b/vignettes/references.bib index af97bfb..dd88c03 100644 --- a/vignettes/references.bib +++ b/vignettes/references.bib @@ -601,3 +601,76 @@ @article{Hastie:1986 volume = {1}, year = {1986} } + +@ARTICLE{Tormene:2009, + title = {Matching incomplete time series with dynamic time warping: an algorithm and an application to post-stroke rehabilitation}, + journal = {Artificial Intelligence in Medicine}, + volume = {45}, + number = {1}, + pages = {11 - 34}, + year = {2009}, + doi = {10.1016/j.artmed.2008.11.007}, + author = {Paolo Tormene and Toni Giorgino and Silvana Quaglini and Mario Stefanelli} +} + +@article{Olofsson:2014, + title = {Good practices for estimating area and assessing accuracy of land change}, + journal = {Remote Sensing of Environment}, + volume = {148}, + number = {}, + pages = {42 - 57}, + year = {2014}, + issn = {0034-4257}, + doi = {10.1016/j.rse.2014.02.015}, + author = {Pontus Olofsson and Giles M. Foody and Martin Herold and Stephen V. Stehman and Curtis E. Woodcock and Michael A. Wulder} +} + +@ARTICLE{Olofsson:2013, + author = {Pontus Olofsson and Giles M. Foody and Stephen V. Stehman and Curtis E. Woodcock}, + title = {Making better use of accuracy data in land change studies: Estimating accuracy and area and quantifying uncertainty using stratified estimation}, + journal = {Remote Sensing of Environment}, + volume = {129}, + number = {}, + pages = {122 - 131}, + year = {2013}, + note = {}, + doi = {10.1016/j.rse.2012.10.031} +} + +@article{White:2014, + author = {J. C. White and M. A. Wulder and G. W. Hobart and J. E. Luther and T. Hermosilla and P. Griffiths and N. C. Coops and R. J. Hall and P. Hostert and A. Dyk and L. Guindon}, + title = {Pixel-Based Image Compositing for Large-Area Dense Time Series Applications and Science}, + journal = {Canadian Journal of Remote Sensing}, + volume = {40}, + number = {3}, + pages = {192-212}, + year = {2014}, + doi = {10.1080/07038992.2014.945827}, +} + +@article{Gomez:2016, + title = {Optical remotely sensed time series data for land cover classification: A review}, + journal = {{ISPRS} Journal of Photogrammetry and Remote Sensing}, + volume = {116}, + number = {}, + pages = {55 - 72}, + year = {2016}, + note = {}, + issn = {0924-2716}, + doi = {10.1016/j.isprsjprs.2016.03.008}, + author = {Cristina G\'{o}mez and Joanne C. White and Michael A. Wulder}, +} + +@article{Foody:2009, + title = {Classification accuracy comparison: Hypothesis tests and the use of confidence intervals in evaluations of difference, equivalence and non-inferiority}, + journal = {Remote Sensing of Environment}, + volume = {113}, + number = {8}, + pages = {1658 - 1663}, + year = {2009}, + issn = {0034-4257}, + doi = {10.1016/j.rse.2009.03.014}, + author = {Giles M. Foody} +} + +