[SCM] GNU M4 source repository branch, branch-1.4, updated. v1.4.13-23-ga6c565c

[Eric Blake]

This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU M4 source repository".

http://git.sv.gnu.org/gitweb/?p=m4.git;a=commitdiff;h=a6c565c6f4b934be334dd9ada97b215595d10103

The branch, branch-1.4 has been updated
       via  a6c565c6f4b934be334dd9ada97b215595d10103 (commit)
       via  cde8ed621040381a2e9584b3d72940ef8c125115 (commit)
      from  30cd2068dcb8b51f9fe29741732027bb00b522a5 (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
commit a6c565c6f4b934be334dd9ada97b215595d10103
Author: Eric Blake <address@hidden>
Date:   Fri Nov 27 21:45:33 2009 -0700

    Work around BSD getopt bug.
    
    * gnulib: Update.
    * doc/m4.texinfo (Command line files): Add test.
    
    Signed-off-by: Eric Blake <address@hidden>

commit cde8ed621040381a2e9584b3d72940ef8c125115
Author: Eric Blake <address@hidden>
Date:   Fri Nov 27 21:21:39 2009 -0700

    Use fastmap for better regex performance.
    
    * src/input.c (word_start): Delete.
    (pop_wrapup): Free memory on exit.
    (set_word_regexp): Compile a fastmap instead.
    (peek_token, next_token): Use the fastmap.
    
    Signed-off-by: Eric Blake <address@hidden>
    (cherry picked from commit 950807234e3ffd376ad51b2a21cd276dd8b4e59c)

-----------------------------------------------------------------------

Summary of changes:
 ChangeLog      |   12 ++++++++++++
 doc/m4.texinfo |   16 ++++++++++++++++
 gnulib         |    2 +-
 src/input.c    |   42 +++++++++++++++---------------------------
 4 files changed, 44 insertions(+), 28 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 19cf20d..0c9b193 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,15 @@
+2009-11-28  Eric Blake  <address@hidden>
+
+       Work around BSD getopt bug.
+       * gnulib: Update.
+       * doc/m4.texinfo (Command line files): Add test.
+
+       Use fastmap for better regex performance.
+       * src/input.c (word_start): Delete.
+       (pop_wrapup): Free memory on exit.
+       (set_word_regexp): Compile a fastmap instead.
+       (peek_token, next_token): Use the fastmap.
+
 2009-11-26  Eric Blake  <address@hidden>
 
        Ignore write failures before stack overflow exit.
diff --git a/doc/m4.texinfo b/doc/m4.texinfo
index 2cf86d4..550fa4a 100644
--- a/doc/m4.texinfo
+++ b/doc/m4.texinfo
@@ -989,6 +989,22 @@ syscmd([echo 'esyscmd(echo hi >&2 && echo err"print(bye
 @error{}bye
 @result{}0
 @end example
+
address@hidden Test that we obey POSIX semantics with -D interspersed with
address@hidden files, even with POSIXLY_CORRECT (BSD getopt gets it wrong).
+
+$ @kbd{m4 }
address@hidden
+ifdef(`__unix__', ,
+      `errprint(` skipping: syscmd does not have unix semantics
+')m4exit(`77')')dnl
+changequote(`[', `]')dnl
+syscmd([POSIXLY_CORRECT=1 ']__program__[' -Dbar=hello foo -Dbar=world foo])dnl
address@hidden
address@hidden
+sysval
address@hidden
address@hidden example
 @end ignore
 
 @node Syntax
diff --git a/gnulib b/gnulib
index 23cd741..c0c5acf 160000
--- a/gnulib
+++ b/gnulib
@@ -1 +1 @@
-Subproject commit 23cd74133c81b80d335c701848b71d85a53402c6
+Subproject commit c0c5acfbe255f5542bc1c81c7aec223d95e504a6
diff --git a/src/input.c b/src/input.c
index 740df78..7336c53 100644
--- a/src/input.c
+++ b/src/input.c
@@ -154,7 +154,6 @@ STRING ecomm;
 
 # define DEFAULT_WORD_REGEXP "[_a-zA-Z][_a-zA-Z0-9]*"
 
-static char *word_start;
 static struct re_pattern_buffer word_regexp;
 static int default_word_regexp;
 static struct re_registers regs;
@@ -396,6 +395,9 @@ pop_wrapup (void)
       obstack_free (&file_names, NULL);
       obstack_free (wrapup_stack, NULL);
       free (wrapup_stack);
+#ifdef ENABLE_CHANGEWORD
+      regfree (&word_regexp);
+#endif /* ENABLE_CHANGEWORD */
       return false;
     }
 
@@ -762,8 +764,6 @@ set_comment (const char *bc, const char *ec)
 void
 set_word_regexp (const char *regexp)
 {
-  int i;
-  char test[2];
   const char *msg;
   struct re_pattern_buffer new_word_regexp;
 
@@ -785,31 +785,19 @@ set_word_regexp (const char *regexp)
       return;
     }
 
-  /* If compilation worked, retry using the word_regexp struct.
-     Can't rely on struct assigns working, so redo the compilation.  */
-  regfree (&word_regexp);
+  /* If compilation worked, retry using the word_regexp struct.  We
+     can't rely on struct assigns working, so redo the compilation.
+     The fastmap can be reused between compilations, and will be freed
+     by the final regfree.  */
+  if (!word_regexp.fastmap)
+    word_regexp.fastmap = xcharalloc (UCHAR_MAX + 1);
   msg = re_compile_pattern (regexp, strlen (regexp), &word_regexp);
+  assert (!msg);
   re_set_registers (&word_regexp, &regs, regs.num_regs, regs.start, regs.end);
-
-  if (msg != NULL)
-    {
-      M4ERROR ((EXIT_FAILURE, 0,
-               "INTERNAL ERROR: expression recompilation `%s': %s",
-               regexp, msg));
-    }
+  if (re_compile_fastmap (&word_regexp))
+    assert (false);
 
   default_word_regexp = false;
-
-  if (word_start == NULL)
-    word_start = (char *) xmalloc (256);
-
-  word_start[0] = '\0';
-  test[1] = '\0';
-  for (i = 1; i < 256; i++)
-    {
-      test[0] = i;
-      word_start[i] = re_search (&word_regexp, test, 1, 0, 0, NULL) >= 0;
-    }
 }
 
 #endif /* ENABLE_CHANGEWORD */
@@ -900,7 +888,7 @@ next_token (token_data *td, int *line)
 
 #ifdef ENABLE_CHANGEWORD
 
-  else if (!default_word_regexp && word_start[ch])
+  else if (!default_word_regexp && word_regexp.fastmap[ch])
     {
       obstack_1grow (&token_stack, ch);
       while (1)
@@ -1062,9 +1050,9 @@ peek_token (void)
     }
   else if ((default_word_regexp && (isalpha (ch) || ch == '_'))
 #ifdef ENABLE_CHANGEWORD
-      || (! default_word_regexp && word_start[ch])
+           || (! default_word_regexp && word_regexp.fastmap[ch])
 #endif /* ENABLE_CHANGEWORD */
-      )
+           )
     {
       result = TOKEN_WORD;
     }


hooks/post-receive
-- 
GNU M4 source repository