Attachment 952258 Details for Bug 799863 – Backported fix

[patch] Backported fix

grep-2.6.3-w-multibyte-fix.patch (text/plain), 3.30 KB, created by Jaroslav Škarvada on 2014-10-30 16:57:42 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Jaroslav Škarvada

Created: 2014-10-30 16:57:42 UTC

Size: 3.30 KB

patch

obsolete

>diff --git a/src/dfa.c b/src/dfa.c
>--- a/src/dfa.c
>+++ b/src/dfa.c
>@@ -770,6 +770,20 @@ parse_bracket_exp (void)
> /* Return non-zero if C is a `word-constituent' byte; zero otherwise.  */
> #define IS_WORD_CONSTITUENT(C) (ISALNUM(C) || (C) == '_')
> 
>+#define PUSH_LEX_STATE(s)			\
>+  do						\
>+    {						\
>+      char const *lexptr_saved = lexptr;	\
>+      size_t lexleft_saved = lexleft;		\
>+      lexptr = (s);				\
>+      lexleft = strlen (lexptr)
>+
>+#define POP_LEX_STATE()				\
>+      lexptr = lexptr_saved;			\
>+      lexleft = lexleft_saved;			\
>+    }						\
>+  while (0)
>+
> static token
> lex (void)
> {
>@@ -1057,14 +1071,33 @@ lex (void)
> 	case 'W':
> 	  if (!backslash || (syntax_bits & RE_NO_GNU_OPS))
> 	    goto normal_char;
>-	  zeroset(ccl);
>-	  for (c2 = 0; c2 < NOTCHAR; ++c2)
>-	    if (IS_WORD_CONSTITUENT(c2))
>-	      setbit(c2, ccl);
>-	  if (c == 'W')
>-	    notset(ccl);
>+
>+	  if (MB_CUR_MAX <= 1)
>+	    {
>+	      zeroset (ccl);
>+	      for (c2 = 0; c2 < NOTCHAR; ++c2)
>+		if (IS_WORD_CONSTITUENT (c2))
>+		  setbit (c2, ccl);
>+	      if (c == 'W')
>+		notset (ccl);
>+	      laststart = false;
>+	      return lasttok = CSET + charclass_index (ccl);
>+	    }
>+
>+	  /* FIXME: see if optimizing this, as is done with ANYCHAR and
>+	     add_utf8_anychar, makes sense.  */
>+
>+	  /* \w and \W are documented to be equivalent to [_[:alnum:]] and
>+	     [^_[:alnum:]] respectively, so tell the lexer to process those
>+	     strings, each minus its "already processed" '['.  */
>+	  PUSH_LEX_STATE (c == 'w' ? "_[:alnum:]]" : "^_[:alnum:]]");
>+
>+	  lasttok = parse_bracket_exp ();
>+
>+	  POP_LEX_STATE ();
>+
> 	  laststart = 0;
>-	  return lasttok = CSET + charclass_index(ccl);
>+	  return lasttok;
> 
> 	case '[':
> 	  if (backslash)
>diff --git a/tests/Makefile.am b/tests/Makefile.am
>--- a/tests/Makefile.am
>+++ b/tests/Makefile.am
>@@ -61,6 +61,7 @@ TESTS =						\
>   turkish-I					\
>   turkish-I-without-dot				\
>   word-multi-file				\
>+  word-multibyte				\
>   yesno.sh
> 
> EXTRA_DIST =					\
>diff --git a/tests/Makefile.in b/tests/Makefile.in
>--- a/tests/Makefile.in
>+++ b/tests/Makefile.in
>@@ -827,6 +827,7 @@ TESTS = \
>   turkish-I					\
>   turkish-I-without-dot				\
>   word-multi-file				\
>+  word-multibyte				\
>   yesno.sh
> 
> EXTRA_DIST = \
>@@ -1247,6 +1248,8 @@ turkish-I-without-dot.log: turkish-I-without-dot
> 	@p='turkish-I-without-dot'; $(am__check_pre) $(LOG_COMPILE) "$$tst" $(am__check_post)
> word-multi-file.log: word-multi-file
> 	@p='word-multi-file'; $(am__check_pre) $(LOG_COMPILE) "$$tst" $(am__check_post)
>+word-multibyte.log: word-multibyte
>+	@p='word-multibyte'; $(am__check_pre) $(LOG_COMPILE) "$$tst" $(am__check_post)
> yesno.sh.log: yesno.sh
> 	@p='yesno.sh'; $(am__check_pre) $(LOG_COMPILE) "$$tst" $(am__check_post)
> .test.log:
>diff --git a/tests/word-multibyte b/tests/word-multibyte
>new file mode 100644
>--- /dev/null
>+++ b/tests/word-multibyte
>@@ -0,0 +1,23 @@
>+#!/bin/sh
>+# This would fail for grep-2.20
>+. "${srcdir=.}/init.sh"; path_prepend_ ../src
>+
>+require_en_utf8_locale_
>+
>+printf '\xc3\xa1\n' > in || framework_failure_
>+LC_ALL=en_US.UTF-8
>+export LC_ALL
>+
>+fail=0
>+
>+for LOC in en_US.UTF-8 zh_CN $LOCALE_FR_UTF8; do
>+  out=out1-$LOC
>+  LC_ALL=$LOC grep '\w' in >$out || fail=1
>+  compare in $out || fail=1
>+
>+  out=out2-$LOC
>+  LC_ALL=$LOC grep '\W' in >$out && fail=1
>+  compare /dev/null $out || fail=1
>+done
>+
>+Exit $fail
>-- 
>1.9.3
>

Actions: View | Diff

Attachments on bug 799863: 952258 | 985631