1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/patches/glibc/2.9/140-regex-BZ9697.patch Sun Apr 19 16:17:11 2009 +0000
1.3 @@ -0,0 +1,111 @@
1.4 +Original patch from: gentoo/src/patchsets/glibc/2.9/0052_all_glibc-2.9-regex-BZ9697.patch
1.5 +
1.6 +-= BEGIN original header =-
1.7 +http://sourceware.org/ml/libc-alpha/2009-01/msg00005.html
1.8 +
1.9 +From ea8ca0dfcbf2721bcf2c08ce3c01d5764b827572 Mon Sep 17 00:00:00 2001
1.10 +From: Ulrich Drepper <drepper@redhat.com>
1.11 +Date: Thu, 8 Jan 2009 00:42:28 +0000
1.12 +Subject: [PATCH] (re_compile_fastmap_iter): Rewrite COMPLEX_BRACKET handling.
1.13 +
1.14 +-= END original header =-
1.15 +
1.16 +diff -durN glibc-2_9.orig/posix/regcomp.c glibc-2_9/posix/regcomp.c
1.17 +--- glibc-2_9.orig/posix/regcomp.c 2008-05-15 05:07:21.000000000 +0200
1.18 ++++ glibc-2_9/posix/regcomp.c 2009-02-02 22:00:41.000000000 +0100
1.19 +@@ -350,47 +350,67 @@
1.20 + #ifdef RE_ENABLE_I18N
1.21 + else if (type == COMPLEX_BRACKET)
1.22 + {
1.23 +- int i;
1.24 + re_charset_t *cset = dfa->nodes[node].opr.mbcset;
1.25 +- if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes
1.26 +- || cset->nranges || cset->nchar_classes)
1.27 +- {
1.28 ++ int i;
1.29 ++
1.30 + # ifdef _LIBC
1.31 +- if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0)
1.32 ++ /* See if we have to try all bytes which start multiple collation
1.33 ++ elements.
1.34 ++ e.g. In da_DK, we want to catch 'a' since "aa" is a valid
1.35 ++ collation element, and don't catch 'b' since 'b' is
1.36 ++ the only collation element which starts from 'b' (and
1.37 ++ it is caught by SIMPLE_BRACKET). */
1.38 ++ if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0
1.39 ++ && (cset->ncoll_syms || cset->nranges))
1.40 + {
1.41 +- /* In this case we want to catch the bytes which are
1.42 +- the first byte of any collation elements.
1.43 +- e.g. In da_DK, we want to catch 'a' since "aa"
1.44 +- is a valid collation element, and don't catch
1.45 +- 'b' since 'b' is the only collation element
1.46 +- which starts from 'b'. */
1.47 + const int32_t *table = (const int32_t *)
1.48 + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
1.49 + for (i = 0; i < SBC_MAX; ++i)
1.50 + if (table[i] < 0)
1.51 + re_set_fastmap (fastmap, icase, i);
1.52 + }
1.53 +-# else
1.54 +- if (dfa->mb_cur_max > 1)
1.55 +- for (i = 0; i < SBC_MAX; ++i)
1.56 +- if (__btowc (i) == WEOF)
1.57 +- re_set_fastmap (fastmap, icase, i);
1.58 +-# endif /* not _LIBC */
1.59 +- }
1.60 +- for (i = 0; i < cset->nmbchars; ++i)
1.61 ++# endif /* _LIBC */
1.62 ++
1.63 ++ /* See if we have to start the match at all multibyte characters,
1.64 ++ i.e. where we would not find an invalid sequence. This only
1.65 ++ applies to multibyte character sets; for single byte character
1.66 ++ sets, the SIMPLE_BRACKET again suffices. */
1.67 ++ if (dfa->mb_cur_max > 1
1.68 ++ && (cset->nchar_classes || cset->non_match
1.69 ++# ifdef _LIBC
1.70 ++ || cset->nequiv_classes
1.71 ++# endif /* _LIBC */
1.72 ++ ))
1.73 + {
1.74 +- char buf[256];
1.75 +- mbstate_t state;
1.76 +- memset (&state, '\0', sizeof (state));
1.77 +- if (__wcrtomb (buf, cset->mbchars[i], &state) != (size_t) -1)
1.78 +- re_set_fastmap (fastmap, icase, *(unsigned char *) buf);
1.79 +- if ((bufp->syntax & RE_ICASE) && dfa->mb_cur_max > 1)
1.80 ++ unsigned char c = 0;
1.81 ++ do
1.82 + {
1.83 +- if (__wcrtomb (buf, towlower (cset->mbchars[i]), &state)
1.84 +- != (size_t) -1)
1.85 +- re_set_fastmap (fastmap, 0, *(unsigned char *) buf);
1.86 ++ mbstate_t mbs;
1.87 ++ memset (&mbs, 0, sizeof (mbs));
1.88 ++ if (__mbrtowc (NULL, (char *) &c, 1, &mbs) == (size_t) -2)
1.89 ++ re_set_fastmap (fastmap, false, (int) c);
1.90 + }
1.91 ++ while (++c != 0);
1.92 + }
1.93 ++
1.94 ++ else
1.95 ++ {
1.96 ++ /* ... Else catch all bytes which can start the mbchars. */
1.97 ++ for (i = 0; i < cset->nmbchars; ++i)
1.98 ++ {
1.99 ++ char buf[256];
1.100 ++ mbstate_t state;
1.101 ++ memset (&state, '\0', sizeof (state));
1.102 ++ if (__wcrtomb (buf, cset->mbchars[i], &state) != (size_t) -1)
1.103 ++ re_set_fastmap (fastmap, icase, *(unsigned char *) buf);
1.104 ++ if ((bufp->syntax & RE_ICASE) && dfa->mb_cur_max > 1)
1.105 ++ {
1.106 ++ if (__wcrtomb (buf, towlower (cset->mbchars[i]), &state)
1.107 ++ != (size_t) -1)
1.108 ++ re_set_fastmap (fastmap, false, *(unsigned char *) buf);
1.109 ++ }
1.110 ++ }
1.111 ++ }
1.112 + }
1.113 + #endif /* RE_ENABLE_I18N */
1.114 + else if (type == OP_PERIOD