glibc/posix/bug-regex1.c
DJ Delorie 69623c0db0 posix: remove some iso-8859-encoded characters
With the increasing adoption of UTF-8, modern editors may (will?)
replace iso-8859-encoded characters in the range 0x80..0xff with
their UTF-8 equivalent, as will mailers and other tools.  This breaks
our testsuite and corrupts patches.

So, this patch starts replacing these problematic characters with
\OCTal sequences instead (adding support for those in tst-fnmatch.c)
or with plain ASCII characters (PTESTS).

Reviewed-by: Carlos O'Donell <carlos@redhat.com>
2021-09-01 14:29:59 -04:00

65 lines
1.3 KiB
C

/* Test case by Jim Meyering <jim@meyering.net>. */
#include <locale.h>
#include <stdio.h>
#include <string.h>
#include <regex.h>
#include <wchar.h>
int
main (void)
{
struct re_pattern_buffer regex;
struct re_registers regs;
const char *s;
int match;
int result = 0;
memset (&regex, '\0', sizeof (regex));
setlocale (LC_ALL, "de_DE.ISO-8859-1");
fwide (stdout, -1);
re_set_syntax (RE_SYNTAX_POSIX_EGREP | RE_DEBUG);
puts ("in C locale");
setlocale (LC_ALL, "C");
s = re_compile_pattern ("[an\371]*n", 7, &regex);
if (s != NULL)
{
puts ("re_compile_pattern return non-NULL value");
result = 1;
}
else
{
match = re_match (&regex, "an", 2, 0, &regs);
if (match != 2)
{
printf ("re_match returned %d, expected 2\n", match);
result = 1;
}
else
puts (" -> OK");
}
puts ("in de_DE.ISO-8859-1 locale");
setlocale (LC_ALL, "de_DE.ISO-8859-1");
s = re_compile_pattern ("[an\371]*n", 7, &regex);
if (s != NULL)
{
puts ("re_compile_pattern return non-NULL value");
result = 1;
}
else
{
match = re_match (&regex, "an", 2, 0, &regs);
if (match != 2)
{
printf ("re_match returned %d, expected 2\n", match);
result = 1;
}
else
puts (" -> OK");
}
return result;
}