Hmm. I tried it and I still only get 3 matches. The fourth has a start
position of -1 every time. I wonder if it's because I'm using pcre instead
of regex? Do you have a url for regex c source code (I hope it is standard
c code and not dependent on unix library functions)?
If not, I can get by now with removing the = sign from the third match.
Thanks
Bill
To: Plucker Development List
Michael <[EMAIL PROTECTED]>
Nordstr�m cc: (bcc: Bill Nalen/Towers Perrin)
<micke@sslug. Subject: Regex (was: Windows Palm
conduit)
dk>
06/07/2001
03:43 PM
Please
respond to
Plucker
Development
List
On Thu, Jun 07, 2001, Bill Nalen/Towers Perrin wrote:
> Okay, let me try again. I need to end up with a string that C++ can
> swallow. The way I have it coded now it will find
> match(0): href="http://someurl/page.html";
> match(1): href
> match(2): ="http://someurl/page.html";
Hi Bill,
I believe you should get one more regex match. I hacked together a
quick test program to try it out,
#include <stdio.h>
#include <string.h>
#include <regex.h>
#define REGEX "[ \r\t\n]*([a-zA-Z_][-.a-zA-Z_0-9]*)([ \r\t\n]*=[ \r\t\n]
*('[^']*'|\"[^\"]*\"|[-a-zA-Z0-9./:+*%?!()_#=~]*))?"
#define NMATCH 4
int main(int argc, char* argv[])
{
regex_t compiled_pattern;
regmatch_t match[NMATCH];
char* string;
int flag;
int i;
if (argc != 2)
exit(1);
string = (char*) malloc(strlen(argv[1]) + 1);
if (string == NULL)
exit(1);
flag = REG_EXTENDED;
regcomp(&compiled_pattern, REGEX, flag);
flag = 0;
regexec(&compiled_pattern, argv[1], NMATCH, match, flag);
for (i = 0; i < NMATCH; i++) {
strncpy(string, &argv[1][match[i].rm_so], match[i].rm_eo -
match[i].rm_so);
string[match[i].rm_eo - match[i].rm_so] = '\0';
fprintf(stdout, "Match(%d): %s\n", i, string);
}
regfree(&compiled_pattern);
return 0;
}
'href="http://someurl/page.html";' will give the following output,
Match(0): href="http://someurl/page.html";
Match(1): href
Match(2): ="http://someurl/page.html";
Match(3): "http://someurl/page.html";
> Can someone tell me what the r does?
'r' is used for raw strings. It turns off the interpretation of the
backslash characters.
/Mike
