C ++中的正则表达式问题 [英] Issue with regular expressions in C++
问题描述
我尝试使用以下正则表达式，该正则表达式已经在 C#中，在 C ++ 中工作，但是在 C ++ 中不工作.
I tried to use the following regular expression, which already works in C#, in C++ as well but it's not working in C++.
std::regex r = std::regex("([^%]*(%[.[0-9]*]?[a-z])*)*", std::regex::extended);
它设法匹配多个字符串并正确拒绝其他字符串，但被卡在字符串％d小于可用的pbn％f％d"上(确实被卡住-没有错误)，应该将其拒绝(因为存在％并没有紧随法律后缀).
it manages to match several strings and reject others correctly, but gets stuck (really stuck - no error) on the string "%d smaller than available pbn % f %d", which it should reject (since there is a % that's not immediately preceded by a legal suffix).
使用 std :: regex r = std :: regex((([[^％] *)(％(\\.([0-9] *))?[az])*)*); 表现出与我之前描述的行为完全相同的行为.(我假设这两个正则表达式是等效的-只是一个正则表达式(如C#使用)，第二个是ECMAScript(如c ++默认值)
Using std::regex r = std::regex("(([^%]*)(%(\\.([0-9]*))?[a-z])*)*"); exhibits exactly the same behavior i described before. (i assume those two regexes are equivalent - just one is in canonical form like C# uses and the second is ECMAScript like the c++ default)
我不确定是什么问题.我也想将整个字符串匹配到该模式，以便仅在整个字符串整体匹配时才匹配.所以我想为此目的使用 regex_match .我在C ++中使用以下代码:
I am not sure what's the problem.
Also i want to match the entire string to that pattern so it matches only if the entire string matches as a whole. So i want to use regex_match for that purpose. I use the following code in C++:
if (std::regex_match(str, r))
此外，在C#中，我使用以下代码执行该检查(整个字符串整体匹配):
Also, in C# i use the following code to perform that check (that the entire string matches as a whole):
Regex^ r = gcnew Regex("([^%]*(%[.[0-9]*]?[a-z])*)*", RegexOptions::IgnoreCase);
Match^ m = r->Match(str);
if (m->Success && m->Groups[0]->Length== str->Length)
仅举一个我想要正则表达式匹配的示例:
Just to give an example of what i want the regular expression to match is:
状态为％s (％d )的事件％s (％d )，移至状态％s (％d )...
Got event %s (%d) in state %s (%d), moving to state %s (%d) ...
或
一些％.34x 事件
并且正则表达式应该与以下内容不匹配:
And the regular expression is supposed not to match the following:
一些％东西.
用语言解释正则表达式应做的事情-它应接受唯一的字符串，其中所有出现的％(如果有的话)立即紧跟一个字母或一个.46456x(也就是一些数字和一个字母))并拒绝所有其他人.
To explain in words what the regex should do - it should accept the only string that all the occurrences (if any) of % in them is immediately preceded by a letter or by a .46456x (aka . some numbers and a letter) and reject all others.
更新:有效的正则表达式为 ^([^％] |％((\\.)?[0-9] +)?[a-zA-Z])* $ .问题在于，与C#正则表达式不同，它确实很慢，并且会拖慢ALOTTT的应用程序.所以我在想也许最好使用std :: regex_search来查找是否出现％的情况，而不是立即跟着后者，或者是.NUMBERS，然后是字母，或者是数字，然后是字母.将不胜感激正则表达式的帮助.
UPDATE:
The regex that works is ^([^%]|%((\\.)?[0-9]+)?[a-zA-Z])*$ . The problem is that unlike the C# regex this one is really slow and slows down the application by ALOTTT. So i was thinking maybe it's better to maybe use std::regex_search in order to find if there is an occurrence of % that's not immediately followed either by a latter or by a .NUMBERS and then a letter or by NUMBERS and then a letter. Will appreciate help with a regex that does that.
更新2:
我正在使用正则表达式 ^.*％(?！([.]?[0-9] +)?[a-zA-Z]).* $ 起作用，并且我将其与std :: regex_search结合使用.它比以前的解决方案要快得多，但比C#版本要慢得多(43秒，而在C#中不到6秒).有没有办法进一步优化它?
I am using the regex ^.*%(?!([.]?[0-9]+)?[a-zA-Z]).*$ which works, and i use it with std::regex_search. It's much faster then the previous solution but still much slower then the C# version (43 seconds vs less then 6 seconds in C#). Is there a way to optimize it even farther?
推荐答案
在这里，字符串中所有％都必须兼容.
如果是，则匹配整个字符串，如果不匹配，则不匹配
字符串.
Here you go, all % in string must be compliant.
If so, match the entire string, if not, don't match
the string.
我建议您使用 if(regex_search(sTarget，sRx，sMatch，flags))
但是 regex_match()会做同样的事情.
I suggest you do this with i.e. if ( regex_search( sTarget, sRx, sMatch, flags ) )
but regex_match() would do the same thing.
^(?:[^％] *％(?:\.[0-9] *)?[az])+ [^％] * $
扩展
^ # BOS
(?: # Cluster begin
[^%]* # Not % characters
% # % found
(?: \. [0-9]* )? # optional .###
[a-z] # single a-z required
)+ # Cluster end, 1 to many times
[^%]* # Not % characters
$ # EOS
这篇关于C ++中的正则表达式问题的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
