Re: [kbrint@hamachi.rufus.net: cperl-mode problem under emacs 22.1]

[Richard Stallman <rms@gnu.org>]

Does anyone see any problem with Bob's patch?


Date: Thu, 9 Aug 2007 19:57:02 -0400
From: Bob Rogers <rogers-emacs@rgrjr.dyndns.org>
To: rms@gnu.org, emacs-devel@gnu.org
Subject: [kbrint@hamachi.rufus.net: cperl-mode problem under emacs 22.1]
In-Reply-To: <E1IIdZL-00040W-U4@fencepost.gnu.org>

   From: Richard Stallman <rms@gnu.org>
   Date: Wed, 08 Aug 2007 00:54:51 -0400

   Would someone please try to DTRT, then ack?  Sometimes these problems
   are impossible to fix, or impossible without creating worse problems.
   But it would be good to try it and see, before giving up.

   ------- Start of forwarded message -------
   X-Spam-Status: No, score=0.0 required=5.0 tests=UNPARSEABLE_RELAY 
	   autolearn=failed version=3.1.0
   From: kevin brintnall <kbrint@hamachi.rufus.net>
   To: bug-gnu-emacs@gnu.org
   Date: Tue,  7 Aug 2007 10:13:55 -0500 (CDT)
   Subject: cperl-mode problem under emacs 22.1

   emacs -q
   C-x C-f file.pl
   M-x cperl-mode

   Type into buffer:

   $k =~ s{a}{b};
	     ^ the second open brace will fail to insert, generating message:
	     Args out of range: 14, 15

   This is with the built-in cperl-mode 5.22 in Emacs 22.1.

   This does not happen with emacs 21.x or 20.x.

The second hunk of the following patch seems to fix (or at least paper
over) this problem.  While testing, I found a similar problem:  If you
put

	{
	}

into an empty cperl-mode buffer and then try to change it to

	{
	  $k =~ s{a}{b};
	}

cperl-find-pods-heres dies in a *different* place when you type the
*first* "{".  This is addressed by the first hunk.

   Needless to say, I do not have a whole lot of confidence in these
band-aids; cperl-find-pods-heres is 1122 lines long, and I do not even
begin to understand it.  Take it FWIW.

   ###################################

   With cperl-mode 5.23 form ilya, I can type the {}'s, but the following
   error is generated instead:

   End of `s{ ... }{ ... }' string/RE not found: (scan-error Unbalanced parentheses 11 13)

   This causes fontification to fail.

I can confirm this, but have no further insights.

					-- Bob Rogers
					   http://rgrjr.dyndns.org/

------------------------------------------------------------------------
Index: lisp/progmodes/cperl-mode.el
===================================================================
RCS file: /sources/emacs/emacs/lisp/progmodes/cperl-mode.el,v
retrieving revision 1.89
diff -c -r1.89 cperl-mode.el
*** lisp/progmodes/cperl-mode.el	16 Apr 2007 07:05:28 -0000	1.89
--- lisp/progmodes/cperl-mode.el	9 Aug 2007 23:46:22 -0000
***************
*** 3729,3736 ****
  	(set-syntax-table reset-st))))
  
  (defsubst cperl-look-at-leading-count (is-x-REx e)
!   (if (re-search-forward (concat "\\=" (if is-x-REx "[ \t\n]*" "") "[{?+*]")
! 			 (1- e) t)	; return nil on failure, no moving
        (if (eq ?\{ (preceding-char)) nil
  	(cperl-postpone-fontification
  	 (1- (point)) (point)
--- 3729,3739 ----
  	(set-syntax-table reset-st))))
  
  (defsubst cperl-look-at-leading-count (is-x-REx e)
!   (if (and (> (point) e)
! 	   ;; return nil on failure, no moving
! 	   (re-search-forward (concat "\\="
! 				      (if is-x-REx "[ \t\n]*" "") "[{?+*]")
! 			      (1- e) t))
        (if (eq ?\{ (preceding-char)) nil
  	(cperl-postpone-fontification
  	 (1- (point)) (point)
***************
*** 4778,4784 ****
  		      (progn
  			(cperl-postpone-fontification
  			 (1- e1) e1 'face my-cperl-delimiters-face)
! 			(if (assoc (char-after b) cperl-starters)
  			    (progn
  			      (cperl-postpone-fontification
  			       b1 (1+ b1) 'face my-cperl-delimiters-face)
--- 4781,4788 ----
  		      (progn
  			(cperl-postpone-fontification
  			 (1- e1) e1 'face my-cperl-delimiters-face)
! 			(if (and (not (eobp))
! 				 (assoc (char-after b) cperl-starters))
  			    (progn
  			      (cperl-postpone-fontification
  			       b1 (1+ b1) 'face my-cperl-delimiters-face)