Merge pull request #779 from spencermountain/dev

Dev

README.md

@@ -328,6 +328,7 @@ _(these methods are on the `nlp` object)_
 - **[.verbose()](https://observablehq.com/@spencermountain/compromise-constructor-methods)** - log our decision-making for debugging
 - **[.version()](https://observablehq.com/@spencermountain/compromise-constructor-methods)** - current semver version of the library
 - **[.world()](https://observablehq.com/@spencermountain/compromise-constructor-methods)** - grab all current linguistic data
+- **[.parseMatch()](https://observablehq.com/@spencermountain/compromise-constructor-methods)** - pre-parse any match statements for faster lookups
 
 ##### Utils
 

builds/compromise.js

@@ -1,4 +1,4 @@
-/* compromise 13.4.0 MIT */
+/* compromise 13.5.0 MIT */
 (function (global, factory) {
   typeof exports === 'object' && typeof module !== 'undefined' ? module.exports = factory() :
   typeof define === 'function' && define.amd ? define(factory) :
@@ -3440,7 +3440,7 @@
 
   var fromJSON_1 = fromJSON;
 
-  var _version = '13.4.0';
+  var _version = '13.5.0';
 
   var _data = {
     "Comparative": "true¦better",
@@ -13906,6 +13906,12 @@
     nlp.world = function () {
       return world;
     };
+    /** pre-parse any match statements */
+
+
+    nlp.parseMatch = function (str) {
+      return syntax_1(str);
+    };
     /** current version of the library */
 
 

builds/compromise.mjs

@@ -1,4 +1,4 @@
-/* compromise 13.4.0 MIT */
+/* compromise 13.5.0 MIT */
 function _typeof(obj) {
   "@babel/helpers - typeof";
 
@@ -3434,7 +3434,7 @@ var fromJSON = function fromJSON(json, world) {
 
 var fromJSON_1 = fromJSON;
 
-var _version = '13.4.0';
+var _version = '13.5.0';
 
 var _data = {
   "Comparative": "true¦better",
@@ -13900,6 +13900,12 @@ function instance(worldInstance) {
   nlp.world = function () {
     return world;
   };
+  /** pre-parse any match statements */
+
+
+  nlp.parseMatch = function (str) {
+    return syntax_1(str);
+  };
   /** current version of the library */
 
 

changelog.md

@@ -10,6 +10,10 @@ While all _Major_ releases should be reviewed, our only two _large_ releases are
 
 #### [Unreleased]
 
+#### 13.5.0
+
+- **[new]** add `.parseMatch()` method for pre-parsing match statements
+
 #### 13.4.0
 
 - **[change]** stop including adverbs and some auxiliaries in `.conjugate()` results

package.json

@@ -2,7 +2,7 @@
   "author": "Spencer Kelly <[email protected]> (http://spencermounta.in)",
   "name": "compromise",
   "description": "modest natural language processing",
-  "version": "13.4.0",
+  "version": "13.5.0",
   "main": "./builds/compromise.js",
   "unpkg": "./builds/compromise.min.js",
   "module": "./builds/compromise.mjs",

src/_version.js

@@ -1 +1 @@
-module.exports = '13.4.0'
+module.exports = '13.5.0'

src/index.js

@@ -7,6 +7,7 @@ const Phrase = require('./Phrase/Phrase')
 const Term = require('./Term/Term')
 const Pool = require('./01-tokenizer/Pool')
 const tinyTagger = require('./02-tagger/tiny')
+const parseMatch = require('./Doc/match/syntax')
 
 function instance(worldInstance) {
   //blast-out our word-lists, just once
@@ -65,6 +66,10 @@ function instance(worldInstance) {
   nlp.world = function () {
     return world
   }
+  /** pre-parse any match statements */
+  nlp.parseMatch = function (str) {
+    return parseMatch(str)
+  }
 
   /** current version of the library */
   nlp.version = version

tests/constructor.test.js

@@ -1,18 +1,18 @@
 const test = require('tape')
 const nlp = require('./_lib')
 
-test('extra exports:', function(t) {
+test('extra exports:', function (t) {
   t.ok(nlp.version, 'version number exported')
 
-  t.doesNotThrow(function() {
+  t.doesNotThrow(function () {
     nlp.verbose(true)
     nlp.verbose(false)
   }, 'can set verbosity')
 
   t.end()
 })
 
-test('tokenize() runs without pos-tagging', function(t) {
+test('tokenize() runs without pos-tagging', function (t) {
   const str = 'Miss Hoover, I glued my head to my shoulder.'
   const r = nlp.tokenize(str)
   t.equal(r.out('text'), str, 'tokenize output is same')
@@ -25,7 +25,7 @@ test('tokenize() runs without pos-tagging', function(t) {
   t.end()
 })
 
-test('tokenize() accepts lexicon param', function(t) {
+test('tokenize() accepts lexicon param', function (t) {
   let doc = nlp.tokenize('spencer kelly is working here', {
     'spencer kelly': 'Person',
     working: 'NotFun',
@@ -36,3 +36,28 @@ test('tokenize() accepts lexicon param', function(t) {
   t.equal(doc.has('#Verb'), false, 'not a full tag')
   t.end()
 })
+
+test('parseMatch() results are symmetric', function (t) {
+  const doc = nlp(`Why doesnt ross, the largest friend, simply eat the other 5?`)
+  let matches = [
+    '#MaleName the #Adjective friend',
+    '^why',
+    '#Value$',
+    null,
+    ['simply', 'eat'],
+    'why',
+    '.',
+    '. simply?',
+    'simply eat',
+    'tornado alley #Hoover',
+  ]
+  matches.forEach(str => {
+    let match = nlp.parseMatch(str)
+    let a = doc.match(match).json()
+    let b = doc.match(str).json()
+    a = JSON.stringify(a)
+    b = JSON.stringify(b)
+    t.equal(a, b, str)
+  })
+  t.end()
+})

types/index.d.ts

@@ -41,6 +41,10 @@ declare interface nlp<D extends object, W extends object, Ph extends Object> {
   clone(): nlp<D, W, Ph>
   /**  current semver version of the library */
   version: nlp.ExtendedDocument<D, W, Ph>
+  /** grab the document's context data */
+  world(): W
+  /** pre-parse a match statement, for faster lookups*/
+  parseMatch(str: string): nlp<D, W, Ph>
 }
 
 declare function nlp(text?: string, lexicon?: Lexicon): nlp.DefaultDocument