Fix ranges with multibyte characters as endpoints.

This is another bug in computing the fastmap.  It was reported by a user
of sed because it usually does not happen with !_LIBC.  However, it is
there in that case too.

The bug is that whenever we have a range at the beginning of the regex,
the regex must be tested on any possible multibyte character.  The reason
why _LIBC masks it, is that in general there is a collation symbol for
each possible multibyte-character lead byte, so all the lead bytes are
in general already part of the fastmap.

The tests use cyrillic characters as an example.  With _LIBC, they pass
without the patch too, but you can make them fail by removing collation
symbols handling.

3 files changed, 106 insertions, 2 deletions

diff --git a/posix/Makefile b/posix/Makefile
index c9ce18b..2a467a8 100644
--- a/posix/Makefile
+++ b/posix/Makefile
@@ -82,7 +82,7 @@ tests		:= tstgetopt testfnm runtests runptests	     \
 		   bug-regex17 bug-regex18 bug-regex19 bug-regex20 \
 		   bug-regex21 bug-regex22 bug-regex23 bug-regex24 \
 		   bug-regex25 bug-regex26 bug-regex27 bug-regex28 \
-		   bug-regex29 \
+		   bug-regex29 bug-regex30 \
 		   tst-nice tst-nanosleep tst-regex2 \
 		   transbug tst-rxspencer tst-pcre tst-boost \
 		   bug-ga1 tst-vfork1 tst-vfork2 tst-vfork3 tst-waitid \
@@ -195,6 +195,7 @@ bug-regex22-ENV = LOCPATH=$(common-objpfx)localedata
 bug-regex23-ENV = LOCPATH=$(common-objpfx)localedata
 bug-regex25-ENV = LOCPATH=$(common-objpfx)localedata
 bug-regex26-ENV = LOCPATH=$(common-objpfx)localedata
+bug-regex30-ENV = LOCPATH=$(common-objpfx)localedata
 tst-rxspencer-ARGS = --utf8 rxspencer/tests
 tst-rxspencer-ENV = LOCPATH=$(common-objpfx)localedata
 tst-pcre-ARGS = PCRE.tests
diff --git a/posix/bug-regex30.c b/posix/bug-regex30.c
new file mode 100644
index 0000000..ef2bd76
--- /dev/null
+++ b/posix/bug-regex30.c
@@ -0,0 +1,103 @@
+/* Russian regular expression tests.
+   Copyright (C) 2009 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+   Contributed by Paolo Bonzini <pbonzini@redhat.com>, 2009.
+
+   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, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#include <sys/types.h>
+#include <mcheck.h>
+#include <regex.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <locale.h>
+
+/* Tests supposed to match.  */
+struct
+{
+  const char *pattern;
+  const char *string;
+  int flags, nmatch;
+  regmatch_t rm[5];
+} tests[] = {
+  /* U+0413	\xd0\x93	CYRILLIC CAPITAL LETTER GHE
+     U+0420	\xd0\xa0        CYRILLIC CAPITAL LETTER ER 
+     U+0430	\xd0\xb0	CYRILLIC SMALL LETTER A
+     U+0433	\xd0\xb3	CYRILLIC SMALL LETTER GHE
+     U+0440	\xd1\x80	CYRILLIC SMALL LETTER ER
+     U+044F	\xd1\x8f	CYRILLIC SMALL LETTER YA */
+  { "[\xd0\xb0-\xd1\x8f]", "\xd0\xb3", 0, 1,
+    { { 0, 2 } } },
+  { "[\xd0\xb0-\xd1\x8f]", "\xd0\x93", REG_ICASE, 1,
+    { { 0, 2 } } },
+  { "[\xd1\x80-\xd1\x8f]", "\xd0\xa0", REG_ICASE, 1,
+    { { 0, 2 } } },
+};
+
+
+static int
+do_test (void)
+{
+  if (setlocale (LC_ALL, "de_DE.UTF-8") == NULL)
+    {
+      puts ("setlocale failed");
+      return 1;
+    }
+
+  int ret = 0;
+
+  for (size_t i = 0; i < sizeof (tests) / sizeof (tests[0]); ++i)
+    {
+      regex_t re;
+      regmatch_t rm[5];
+      int n = regcomp (&re, tests[i].pattern, tests[i].flags);
+      if (n != 0)
+	{
+	  char buf[500];
+	  regerror (n, &re, buf, sizeof (buf));
+	  printf ("regcomp %zd failed: %s\n", i, buf);
+	  ret = 1;
+	  continue;
+	}
+
+      if (regexec (&re, tests[i].string, tests[i].nmatch, rm, 0))
+	{
+	  printf ("regexec %zd failed\n", i);
+	  ret = 1;
+	  regfree (&re);
+	  continue;
+	}
+
+      for (n = 0; n < tests[i].nmatch; ++n)
+	if (rm[n].rm_so != tests[i].rm[n].rm_so
+              || rm[n].rm_eo != tests[i].rm[n].rm_eo)
+	  {
+	    if (tests[i].rm[n].rm_so == -1 && tests[i].rm[n].rm_eo == -1)
+	      break;
+	    printf ("regexec match failure rm[%d] %d..%d\n",
+		    n, rm[n].rm_so, rm[n].rm_eo);
+	    ret = 1;
+	    break;
+	  }
+
+      regfree (&re);
+    }
+
+  return ret;
+}
+
+#define TEST_FUNCTION do_test ()
+#include "../test-skeleton.c"
diff --git a/posix/regcomp.c b/posix/regcomp.c
index 446fed5..6966b5d 100644
--- a/posix/regcomp.c
+++ b/posix/regcomp.c
@@ -377,7 +377,7 @@ re_compile_fastmap_iter (regex_t *bufp, const re_dfastate_t *init_state,
 	     applies to multibyte character sets; for single byte character
 	     sets, the SIMPLE_BRACKET again suffices.  */
 	  if (dfa->mb_cur_max > 1
-	      && (cset->nchar_classes || cset->non_match
+	      && (cset->nchar_classes || cset->non_match || cset->nranges
 # ifdef _LIBC
 		  || cset->nequiv_classes
 # endif /* _LIBC */