diff options
Diffstat (limited to 'string')
-rw-r--r-- | string/bug-strcoll2.c | 93 | ||||
-rw-r--r-- | string/strcoll_l.c | 38 |
2 files changed, 94 insertions, 37 deletions
diff --git a/string/bug-strcoll2.c b/string/bug-strcoll2.c new file mode 100644 index 0000000000..5ce2f94f34 --- /dev/null +++ b/string/bug-strcoll2.c @@ -0,0 +1,93 @@ +/* Bug 18589: sort-test.sh fails at random. + Copyright (C) 1998-2015 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Ulrich Drepper <drepper@cygnus.com>, 1998. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with the GNU C Library; if not, see + <http://www.gnu.org/licenses/>. */ + +#include <stdio.h> +#include <string.h> +#include <locale.h> + +/* An incorrect strcoll optimization resulted in incorrect + results from strcoll for cs_CZ and da_DK. */ + +int +test_cs_CZ (void) +{ + const char t1[] = "config"; + const char t2[] = "choose"; + if (setlocale (LC_ALL, "cs_CZ.UTF-8") == NULL) + { + perror ("setlocale"); + return 1; + } + /* In Czech the digraph ch sorts after c, therefore we expect + config to sort before choose. */ + int a = strcoll (t1, t2); + int b = strcoll (t2, t1); + printf ("strcoll (\"%s\", \"%s\") = %d\n", t1, t2, a); + printf ("strcoll (\"%s\", \"%s\") = %d\n", t2, t1, b); + if (a < 0 && b > 0) + { + puts ("PASS: config < choose"); + return 0; + } + else + { + puts ("FAIL: Wrong sorting in cz_CZ.UTF-8."); + return 1; + } +} + +int +test_da_DK (void) +{ + const char t1[] = "AS"; + const char t2[] = "AA"; + if (setlocale (LC_ALL, "da_DK.ISO-8859-1") == NULL) + { + perror ("setlocale"); + return 1; + } + /* AA should be treated as the last letter of the Danish alphabet, + hence sorting after AS. */ + int a = strcoll (t1, t2); + int b = strcoll (t2, t1); + printf ("strcoll (\"%s\", \"%s\") = %d\n", t1, t2, a); + printf ("strcoll (\"%s\", \"%s\") = %d\n", t2, t1, b); + if (a < 0 && b > 0) + { + puts ("PASS: AS < AA"); + return 0; + } + else + { + puts ("FAIL: Wrong sorting in da_DK.ISO-8859-1"); + return 1; + } +} + +static int +do_test (void) +{ + int err = 0; + err |= test_cs_CZ (); + err |= test_da_DK (); + return err; +} + +#define TEST_FUNCTION do_test () +#include "../test-skeleton.c" diff --git a/string/strcoll_l.c b/string/strcoll_l.c index 8f1225fc7a..35bc0e42a2 100644 --- a/string/strcoll_l.c +++ b/string/strcoll_l.c @@ -29,7 +29,6 @@ # define STRING_TYPE char # define USTRING_TYPE unsigned char # define STRCOLL __strcoll_l -# define STRDIFF __strdiff # define STRCMP strcmp # define WEIGHT_H "../locale/weight.h" # define SUFFIX MB @@ -42,20 +41,6 @@ #include "../locale/localeinfo.h" #include WEIGHT_H -#define MASK_UTF8_7BIT (1 << 7) -#define MASK_UTF8_START (3 << 6) - -size_t -STRDIFF (const STRING_TYPE *s, const STRING_TYPE *t) -{ - size_t n; - - for (n = 0; *s != '\0' && *s++ == *t++; ++n) - continue; - - return n; -} - /* Track status while looking for sequences in a string. */ typedef struct { @@ -269,29 +254,9 @@ STRCOLL (const STRING_TYPE *s1, const STRING_TYPE *s2, __locale_t l) const USTRING_TYPE *extra; const int32_t *indirect; - /* In case there is no locale specific sort order (C / POSIX). */ if (nrules == 0) return STRCMP (s1, s2); - /* Fast forward to the position of the first difference. Needs to be - encoding aware as the byte-by-byte comparison can stop in the middle - of a char sequence for multibyte encodings like UTF-8. */ - uint_fast32_t encoding = - current->values[_NL_ITEM_INDEX (_NL_COLLATE_ENCODING_TYPE)].word; - if (encoding != __cet_other) - { - size_t diff = STRDIFF (s1, s2); - if (diff > 0) - { - if (encoding == __cet_utf8 && (*(s1 + diff) & MASK_UTF8_7BIT) != 0) - do - diff--; - while (diff > 0 && (*(s1 + diff) & MASK_UTF8_START) != MASK_UTF8_START); - s1 += diff; - s2 += diff; - } - } - /* Catch empty strings. */ if (__glibc_unlikely (*s1 == '\0') || __glibc_unlikely (*s2 == '\0')) return (*s1 != '\0') - (*s2 != '\0'); @@ -358,8 +323,7 @@ STRCOLL (const STRING_TYPE *s1, const STRING_TYPE *s2, __locale_t l) byte-level comparison to ensure that we don't waste time going through multiple passes for totally equal strings before proceeding to subsequent passes. */ - if (pass == 0 && encoding == __cet_other && - STRCMP (s1, s2) == 0) + if (pass == 0 && STRCMP (s1, s2) == 0) return result; else break; |