该程序位于 下载 package 中的 bin 目录，要运行该程序，您需要将 下载 package 中的 lib 目录加入环境变量 LD_LIBRARY_PATH 。
- 支持 并、交、差 的正则表达式引擎
- 规则引擎建库工具
- 多正则表达式匹配（Multiple Regular Expression Matching）
- 多正则表达式匹配（Multiple Regular Expression Matching）中的动态 DFA 算法
- 有多个初始状态的 DFA
- 多正则表达式匹配的应用
- 一个很难的字符串问题
Usage:
regex_build.exe Options [Input-Regex-File]
Description:
Compile multiple regular expressions into one DFA
Options:
-h : Show this help infomation
-q : Be quiet, don't print progress info
-I : Ignore case for regex
-L : Use Latin1 charset, not utf8(the default charset)
-a Add-DotStar(Optional): can be 0, 1, 2
0: Do not add DotStar, treat all regular expression as head-anchored
The result DFA will be a "RegexMatch" DFA, not a "RegexSearch" DFA.
1: Respect the regex's head anchor
2: Prepend DotStar on the unioned DFA, this is just for **DEBUG TEST**
Note:
* Head anchored regex example: "^Hello\s+World!"
* If this option is omitted, Add-Dot-Star is 0
* If the option argument is omitted, Add-Dot-Star is 1
-O Large-DFA-File: Large, but pretty fast
-o Small-DFA-File: Small(maybe ~10x smaller than Large-DFA)
Small-DFA is essentially a kind of Virtual Machine DFA.
Small-DFA is not only small, but also fast, in most cases, it is as
faster as Large-DFA. But it may be much faster than Large-DFA in some
special/ideal cases.
-b Bin-Meta-File: The meta file used for capturing submatch
-s Optional-Arg
Build dfa with submatch capture, but the dfa algorithm can only capture
one-pass DFA.
If Optional-Arg is 's', the algorithm will ignore the regex's one-pass
property, and try to use one-pass capturing algorithm to capture all
submatches, this may produce bad result. But when using utf8 encoding,
some unicode-one-pass regex is NOT byte-one-pass, for example:
"从([^到]+)到([^怎]+)怎么走" is unicode-one-pass but not byte-one-pass.
By using option "-ss", MultiRegexSubmatch::match_utf8() will successfully
capture the submatches.
Prepend a '*' at the start of a line has the same effect of '-ss' for the
current regex.
-E C-Program-HeaderFile
Generate enum constant definitions in C-Program-HeaderFile
constant name is specified by second column of Input-Regex-File
-D : Build a dynamic matching DFA
In many cases, the full unioned dfa can not be built, in this situation,
dynamic matching DFA is a compromised solution between full-nfa matching
and full-dfa matching.
For speed up online dynamic matching, this program use a heuristic algorithm
to clustering(partial union) SubDFAs offline.
-z MinZpathLen, default is 2, only effective for full-unioned-dfa
-g : Write DFA and NFA dot graph file for every regex
-G : Write dot graph file for unioned DFA
-c Conflict-Report-File
-t DFA-Type: can be 'd', '1', '2', default is 'd'
a: use DFA class which optimized for adfa_build
d: use DenseDFA
2: use DenseDFA_V2
-T Timeout1[:Timeout2]: Timout2 is optional
Timeout1: Timeout for compiling one regex
Timeout2: Timeout for union all regex
-P : DO NOT Limit cluster_union_power_size, default is true
Input-Regex-File format
This is a tab separated text file, column 1 is the regex, other columns are
optinal user defined contents.
The regex is advanced regex: http://terark.cc/p/?p=1280
Lines can be continued by backslash(\), same as which in C/C++ language.
Bin-Meta-File format
This is a tab separated text file, column description:
Column 1: Integer regex-id, start with 0
Column 2: Integer number-of-submaches(including the full match)
Column 3: Boolean do-capturing, 1 or 0
Column 4: Boolean is-one-pass, 1 or 0
Column 5: String copy of full Input-Regex-File line
规则库的 DFA 创建好以后，可以使用下面的测试程序进行性能测试，测试工具的具体用法，可以参考下表中的代码链接。
编译好的测试程序包含在 package 的 sample/bin 目录，下面的代码同时也包含在 package 中的 sample/src 目录。
| 代码 | package 中可执行文件的路径 |
|---|---|
| regex_allmatch.cpp | sample/bin/regex_allmatch.exe |
| regex_maxmatch.cpp | sample/bin/regex_maxmatch.exe |
| regex_submatch.cpp | sample/bin/regex_submatch.exe |