Re: [Boost-users] Mismatch and regex newbie problem

From: david v (danova_fr_at_[hidden])
Date: 2006-08-29 08:48:23

For example,
If the regex to search for is "test" and the string is "tast" , then 1 
mismatch is found position 2.

John, i think the code below you have submitted is doing exactly what i want 
right ??
std::string::const_iterator start = in.begin();
std::string::const_iterator end = in.end();
boost::sregex_token_iterator i(start, end, *regex, -1), j;
while(i != j)
{
   cout << *i << endl;
}

>Hello,
>I'm using the boost regex library to search short expressions in large
>strings. The code below works fine but it almost does what ever i
>want. The thing i'm missing and that i'm not sure to handle are the
>mismatches. Basically when i look for a regex in a string i want to
>have the possibility to get the number of mismatches and their
>respective positions. I read through the docs and saw that the
>boost_check may do the work. Well the think is that i'm not sure how
>to use it with my code and since the boost library is quite complex
>it would be great if somebody could help to move forward.
>
>How can i use the code below to get the mismatches and their
>positions with boost_check ???

What's boost_check and what's it got to do with regex?

>#####Setting the regex
>.....
>re = "test";
>boost::regex *regex = new boost::regex(re,boost::regbase::icase);
>.....
>
>#######Getting the matched
>.......
>std::string::const_iterator start = in.begin();
>std::string::const_iterator end = in.end();
>boost::smatch match;
>while
>(boost::regex_search(start,end,match,*regex,boost::match_extra))
>{
>cout << match.[0] << endl;
>start = match[0].second;
>}
>
>#############Getting the mismatches ???
>How ???
>
>Any help would be extemely helpful !!!

What do you mean by mismatches? The sections of the string that didn't
match? That would be:

std::string::const_iterator start = in.begin();
std::string::const_iterator end = in.end();
boost::smatch match;
while(boost::regex_search(start,end,match,*regex))
{
   cout << match.prefix() << endl;
   start = match[0].second;
}
// avoid a fencepost error:
cout << match.suffix() << endl;

Note that the match_extra flag is unnecessary unless you really need to
record repeated captures. Normal capture information - of the kind you get
from Perl and other regex engines don't need this - and it slows down
matching as well as requiring building the lib in a non-default mode.

You can also do this rather easier with regex_token_iterator:

std::string::const_iterator start = in.begin();
std::string::const_iterator end = in.end();
boost::sregex_token_iterator i(start, end, *regex, -1), j;
while(i != j)
{
   cout << *i << endl;
}

HTH, John.