PCRE test related to https://github.com/merces/pev/issues/60

pcre_ucp_test.c

//
// Compile and run:
//   LDFLAGS="-lpcre" make pcre_ucp_test && ./pcre_ucp_test 
//

#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <pcre.h>

#define SIZEOF_ARRAY(array)		sizeof(array) / sizeof(*array)

bool regex_match(const char *pattern, const char *value, bool use_unicode) {
	const char *err;
	int errofs;
	int ovector[1000];
	pcre *re = pcre_compile(pattern, use_unicode ? PCRE_UCP : 0, &err, &errofs, NULL);
	if (re == NULL) {
		fprintf(stderr, "%s regex compilation failed\n", use_unicode ? "PCRE_UCP" : "ASCII");
		exit(EXIT_FAILURE);
	}
	const int rc = pcre_exec(re, NULL, value, sizeof(value), 0, 0, ovector, SIZEOF_ARRAY(ovector));
	pcre_free(re);
	return rc > 0;
}

int main() {
	const char pattern[] = "^[a-zA-Z]{3,}://.*$";
	const char value[] = "http://www.example.com/foo/bar";

	bool matches;

	matches = regex_match(pattern, value, false);
	printf("%s matches: %s\n", "ASCII", matches ? "yes" : "nope");

	matches = regex_match(pattern, value, true);
	printf("%s matches: %s\n", "PCRE_UCP", matches ? "yes" : "nope");

	return 0;
}

Mac output:

jweyrich@ptah$:pev [* master]$ uname -sv && LDFLAGS="-lpcre" make pcre_ucp_test && ./pcre_ucp_test
Darwin Darwin Kernel Version 13.1.0: Thu Jan 16 19:40:37 PST 2014; root:xnu-2422.90.20~2/RELEASE_X86_6
make: `pcre_ucp_test' is up to date.
ASCII matches: yes
PCRE_UCP regex compilation faile

Linux output:

jweyrich@devel:~/pev$ uname -sv && uname -sv && LDFLAGS="-lpcre" make pcre_ucp_test && ./pcre_ucp_test
Linux #1 SMP Debian 3.2.54-2
cc   -lpcre  pcre_ucp_test.c   -o pcre_ucp_test
ASCII matches: nope
PCRE_UCP matches: nope

Just found the same problem.