Thread: How to apply Regex in C++?

Hello to all,

I'd like to extract some patterns from a string using regex. I found that regex in C++ are supported only in C++ 11 and
searching about this, I found that it seems only almost Microsoft Visual Studio 2012 C++ compiler supports this.

Can I use the MSVisual Studio C++ compiler in other IDE like netbeans, Dev C++, Code blocks etc or only in MS Visual Studio?

Or do you know about another compiler that supports Regex in C++?

Thanks in advance.

Originally Posted by tabstop

) I don't know how up-to-date mingw is with C++11 (I haven't bothered to update mine in a while).

Since mingw is based on GCC, the <regex> header will be there, but any attempt to use it will result in an exception being thrown.
Clang has it implemented though.
Another option is Boost.Regex.

I use pcre3 for all my needs. There is a nice wrapper for C++ too.
The nuwen distro MinGW Distro - nuwen.net has it as part of the distro.
I link with the static library so no dll to lug around.


James

This example parses a Windows resource dialog file (rc) in the format created with:
https://fbedit.svn.sourceforge.net/s...Ed22/ResEd.zip
This editor allows the style and exestyle to be saved as hex.


It was tested with the MINGW distro from nuwen as it has pcre3 included.
MinGW Distro - nuwen.net


I also tested with the TDM-GCC distro but I had to build the pcre3 library.
I am not too versed with the pcre3 library, just enough to use it a bit.
I tried to show here a bit of the C++ wrapper.


James

Code:

//=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
// tested with NUWEN and TDM-GCC compilers
// make  sure to use -lpcrecpp -lpcre -DPCRE_STATIC when compiling
//=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
#include <iostream>
#include <fstream>
#include <algorithm>
using namespace std;
#include "pcrecpp.h"
#define WAITKEY system("pause")
void cls (void)
{
  system("cls");
}
//==============================================================================
int main ()
{
  vector<string>  Linesvec;
  string  sFileIn="DLG2SRC.rc";
  string  sLine;
  string  sStyle;
  string  sExStyle;
  int     Start={0};
  int     rc={0};
  int     i={0};
  pcrecpp::RE*  re[3];
  string   S1[10];
  ifstream  f;


//id name and value 
  re[0]= new  pcrecpp::RE("(?i)#define\\s*(\\w+)\\s*(\\d+)");
//'Dialog info 
  re[1]= new  pcrecpp::RE("(?i)\\s*(\\w*)\\s*(DIALOGEX|DIALOG)\\s*(\\d+)\\s*,\\s*(\\d+)\\s*,\\s*(\\d+)\\s*,\\s*(\\d+)");
// Control Information  
  re[2]= new  pcrecpp::RE("(?i)\\s*CONTROL\\s+\\x22\\s*([^\\x22\\r\\n]*)\\s*\\x22\\s*,\\s*(\\w*)\\s*,\\x22(\\s*\\w+)\\x22\\s*,([a-fx0-9]+)\\s*,(\\d+)\\s*,\\s*(\\d+)\\s*,\\s*(\\d+)\\s*,+\\s*(\\d+)\\s*(,+([a-fx0-9]*)\\s*)?");


// Open File
  f.open(sFileIn);
// Is it open? if not time to exit
  if(!f.is_open())
    {
      cout  <<  "Could not locate "  <<  sFileIn  <<  endl;
      WAITKEY;
      return 0;
    }
    
// Get each line and stuff it in a vector
  while(getline(f,sLine))
    {
      Linesvec.push_back(sLine);
    }


//close the rc file
  f.close();


/*
------------------------------------------------------------------------------
Get Id names and values from rc file
------------------------------------------------------------------------------
*/
  cout  <<  " Control Dialog Names and Values"  <<  endl;
  cout  <<  "----------------------------------"  <<  endl;


// C++11 range based for with auto data type detection for iterator


 for(auto it : Linesvec)
    {
      rc= re[0]->FullMatch( it,  &S1[0],  &S1[1]);
      if(rc )
        {
          cout  <<  S1[0]  <<  " = "  <<  S1[1]  <<  endl;
        }
    }


  WAITKEY;
  cls();
 for( auto it : Linesvec)
    {
      rc= re[1]->FullMatch( it,  &S1[0],  &S1[1],  &S1[2],  &S1[3],  &S1[4],  &S1[5]);
      if(rc )
        {
          cout << "----------------------------------" << endl;
          cout << "Dialog Information" << endl;
          cout << "----------------------------------" << endl;
          cout << "Dialog Name   " << S1[0] << endl;
          cout << "Dialog Type   " << S1[1] << endl;
          cout << "Dialog Left   " << S1[2] << endl;
          cout << "Dialog Top    " << S1[3] << endl;
          cout << "Dialog Width  " << S1[4] << endl;
          cout << "Dialog Height " << S1[5] << endl;
        }
    }


  WAITKEY;
  cls();


// Control Information




 for( auto it : Linesvec)
    {
      rc= re[2]->FullMatch( it,  &S1[0],  &S1[1],  &S1[2],  &S1[3],  &S1[4],  &S1[5],  &S1[6],  &S1[7],  &S1[8],  &S1[9]);
      if(rc )
        {
          cout << "----------------------------------" << endl;
          cout << "Control Information" << endl;
          cout << "----------------------------------" << endl;
          cout << "Id:        " << S1[1] << endl;
          cout << "Text:      " << S1[0] << endl;
          cout << "Class:     " << S1[2] << endl;
          cout << "Style:     " << S1[3] << endl;
          cout << "Left:      " << S1[4] << endl;
          cout << "Top:       " << S1[5] << endl;
          cout << "Width:     " << S1[6] << endl;
          cout << "Height:    " << S1[7] << endl;
          if(S1[9].length())
            {
              cout << "Exe Style: " << S1[9] << endl;
            }
          WAITKEY;
          cls();
        }
    }


  for(i=0; i<=2; i+=1)
    {
      delete re[i];
    }


  return 0;
}

Hello jcfuller,

Thanks for the help and example shared. I'll check your suggestions as soon as possible. I hope it works for me .

Best regards

Hello Mestertik,

Regex was introduced with TR1.
Microsoft Visual Studio 2008 support regular expressions.
TR1 Regular Expressions

Hello kasparov,

Thank you for your help, I'll check the link you share. I hope the regex used by TR1 are standard like regex in Perl.

Hello jcfuller,

I've been trying to install the MinGW distro you suggest me, all goes fine until I try "g++ --version".

I get the error:

Code:

C:\MinGW\bin\g++.exe is not a valid Win32 application.

I'm testing on windows XP.

Thanks again.

For XP you will need to scroll down to the History item and download the 10.4 release. This was his last 32bit version.

James

The Boost Library has Regex functionality:

Boost C++ Libraries

Hello jcfuller,

Thank you for your help. I've downloaded MinGW, but I don't know how to build pcre3 files. How did you do it? Since it is marked with error the line "#include prcecpp.h".

Hello Vespasian,

Thank you for the answer and share the link. I've unzipped but I still seeing how to make it work with netbeans and see how to use the regex library.

One question, boost libraries are portable, I mean that once compiled a project I won't need to have boost in another machine to run a program?

Thanks for the help.

Personally I find the boost libraries beyond my novice c++ skills.
I have used pcre3 for years in a number of languages and feel more comfortable with it.


If you are using the nuwen distro the boost libraries are included so there is no downloading needed if you choose this route.
For pcre3:
If you extracted the nuwen distro to mingw change the name. For the 32bit (10.4) I use MinGw104 or Nuwen32.
In code blocks: Menu -> Settings -> Toolchain set the executables Compiler's installation directory to the nuwen directory.
In code blocks: Menu -> Settings linker settings add pcrecpp and pcre to the link libraries.

Hello Vespasian,

Thank you for the detailed explanation. I'll follow to try to set boost on codeblocks. I hope will work fine since with jcfuller Mingw distro pcre worked.

Hello jcfuller,

Thank you. I was able to run your code doing as you said in codeblocks. I was not able to do it in netbeans. Anyway I need to see why.

Hello guys,

I was able to apply the procedure to use those libraries in codeblocks, but I was tryin in Visual Studio and it seems the options are very different.

Do you know how to add boost and pcre o use it in Visual Studio (I tried wit VS 2012, 2013)

I just don't find the same options to add lib and include files in VS.

Thanks in advance