正则引擎简单分析

阅读：评论：0

正则引擎简单分析

由于公司需要用到正则搜索，所以简单看了下正则表达式引擎，根据具体使用环境，选择一款最适合的正则引擎。
正则表达式引擎发展到现在已有十多种，由于只用在linux系统，根据网上推荐，确认PCRE、RE2、C REGEX、C++ REGEX、BOOST REGEX这五个引擎。PERE使用的是perl语言，也排除。C++ REGEX对C++ 11的支持不太好，也排除掉。RE2需要到google官网下载资源（国内的没找到），翻墙失败，也给PASS掉。
最终选取了C REGEX与BOOST REGEX两种正则表达式引擎进行性能测试。
此次测试是根据实际使用情况，使用JSON格式进行测试。

代码

c regex:

/*c regex test

at20150119*/

#include <regex.h>

#include <iostream>

#include <sys/types.h>

#include <stdio.h>

#include <cstring>

#include <sys/time.h>

using namespace std;

const int times = 1000000;

stringstrBuf("{"uid":"10011","userName":"sdy","name":"witto","sex":"man","identityCode":"123552036","telephone":"15211012232","qq":"888555666","weibo":"48sdfs@dsf","cityId":"0010","schoolId":"1534","subject":"hard","ruxuenianfen":"20140901","xuezhi":"4","dormAddress":"xinjiekouwaidajie28hao","roomInfo":"4floor506","familyAdress":"huoxing","familyPhone":"125362356211","contactName":"whoami","contactPhone":"15200133623","contactRelation":"dddddd"}");

string jsonBuf(

"head":

{

"flowId":"10001",

"subFlowId":"001",

"step":"04"

"content":

{

"key1":"abandenabandenabandenabandenabandenabandenabandenabandenabandenabandenabandenabanden",

"key2":"litterlitterlitterlitterlitterlitterlitterlitterlitterlitterlitterlitterlitterlitter",

"key3":"attheendattheendattheendattheendattheendattheendattheendattheendattheendattheend",

"key4":"rabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbitrabbit",

"key5":"111111111111111111111111111111111111111111111111111111111111111111111111111111111111",

"key6":"222222222222222222222222222222222222222222222222222222222222222222222222222222222222",

"key7":"333333333333333333333333333333333333333333333333333333333333333333333333333333333333",

"key8":"444444444444444444444444444444444444444444444444444444444444444444444444444444444444",

"key9":"555555555555555555555555555555555555555555555555555555555555555555555555555555555555"

}

}");

int main(int argc,char** argv)

{

//char pattern[64]=""step":"04"";

charpattern[64]=""familyPhone":"125362356211"";

const size_t nmatch = 10;

regmatch_t pm[10];

int z ;

regex_t reg;

timeval end,start;

gettimeofday(&start,NULL);

long count = 0;

for(int i = 0 ; i < times; ++i)

{

regcomp(&reg,pattern,REG_EXTENDED|REG_NOSUB);

if(!regexec(&reg,strBuf.c_str(),nmatch,pm,REG_NOTBOL))

count++;

regfree(&reg);

}

gettimeofday(&end,NULL);

uint time = (end.tv_sec-start.tv_sec)*1000000 + end.tv_usec -start.tv_usec;

cout<<"count:"<<count<<"."<<endl;

cout<<time/1000000<<" s and"<<time%1000000<<" us."<<endl;

return 0 ;

}

boost regex:

/*c regex test

at20150119*/

#include <iostream>

#include <sys/time.h>

#include <string>

#include <boost/regex.hpp>

using namespace std;

using namespace boost;

const int times = 1000000; //1 milliontimes