#include "regexp.h"
#include <assert.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
namespace KJS {
RegExp::RegExp(const UString &p, int flags)
: _flags(flags), _numSubPatterns(0)
{
#ifdef HAVE_PCREPOSIX
int options = PCRE_UTF8;
if (flags & IgnoreCase)
options |= PCRE_CASELESS;
if (flags & Multiline)
options |= PCRE_MULTILINE;
const char *errorMessage;
int errorOffset;
_regex = pcre_compile(p.UTF8String().c_str(), options, &errorMessage, &errorOffset, NULL);
if (!_regex) {
#ifndef NDEBUG
fprintf(stderr, "KJS: pcre_compile() failed with '%s'\n", errorMessage);
#endif
return;
}
#ifdef PCRE_INFO_CAPTURECOUNT
pcre_fullinfo(_regex, NULL, PCRE_INFO_CAPTURECOUNT, &_numSubPatterns);
#endif
#else
int regflags = 0;
#ifdef REG_EXTENDED
regflags |= REG_EXTENDED;
#endif
#ifdef REG_ICASE
if ( f & IgnoreCase )
regflags |= REG_ICASE;
#endif
regcomp(&_regex, p.ascii(), regflags);
#endif
}
RegExp::~RegExp()
{
#ifdef HAVE_PCREPOSIX
pcre_free(_regex);
#else
regfree(&_regex);
#endif
}
UString RegExp::match(const UString &s, int i, int *pos, int **ovector)
{
if (i < 0)
i = 0;
int dummyPos;
if (!pos)
pos = &dummyPos;
*pos = -1;
if (ovector)
*ovector = 0;
if (i > s.size() || s.isNull())
return UString::null();
#ifdef HAVE_PCREPOSIX
if (!_regex)
return UString::null();
int *offsetVector;
int offsetVectorSize;
int fixedSizeOffsetVector[3];
if (!ovector) {
offsetVectorSize = 3;
offsetVector = fixedSizeOffsetVector;
} else {
offsetVectorSize = (_numSubPatterns + 1) * 3;
offsetVector = new int [offsetVectorSize];
}
const CString buffer(s.UTF8String());
convertUTF16OffsetsToUTF8Offsets(buffer.c_str(), &i, 1);
const int numMatches = pcre_exec(_regex, NULL, buffer.c_str(), buffer.size(), i, 0, offsetVector, offsetVectorSize);
if (numMatches < 0) {
#ifndef NDEBUG
if (numMatches != PCRE_ERROR_NOMATCH)
fprintf(stderr, "KJS: pcre_exec() failed with result %d\n", numMatches);
#endif
if (offsetVector != fixedSizeOffsetVector)
delete [] offsetVector;
return UString::null();
}
convertUTF8OffsetsToUTF16Offsets(buffer.c_str(), offsetVector, (numMatches == 0 ? 1 : numMatches) * 2);
*pos = offsetVector[0];
if (ovector)
*ovector = offsetVector;
return s.substr(offsetVector[0], offsetVector[1] - offsetVector[0]);
#else
const uint maxMatch = 10;
regmatch_t rmatch[maxMatch];
char *str = strdup(s.ascii()); if (regexec(&_regex, str + i, maxMatch, rmatch, 0)) {
free(str);
return UString::null();
}
free(str);
if (!ovector) {
*pos = rmatch[0].rm_so + i;
return s.substr(rmatch[0].rm_so + i, rmatch[0].rm_eo - rmatch[0].rm_so);
}
_numSubPatterns = 0;
for(uint j = 1; j < maxMatch && rmatch[j].rm_so >= 0 ; j++)
_numSubPatterns++;
int ovecsize = (_numSubPatterns+1)*3; *ovector = new int[ovecsize];
for (uint j = 0; j < _numSubPatterns + 1; j++) {
if (j>maxMatch)
break;
(*ovector)[2*j] = rmatch[j].rm_so + i;
(*ovector)[2*j+1] = rmatch[j].rm_eo + i;
}
*pos = (*ovector)[0];
return s.substr((*ovector)[0], (*ovector)[1] - (*ovector)[0]);
#endif
}
}