Re: [AUCTeX-devel] Fwd: Bug in font-latex.el

[Mosè Giordano]

Hi Ralf,

I had committed the fix before your message, this is what I've done:
--8<---------------cut here---------------start------------->8---
--- a/font-latex.el
+++ b/font-latex.el
@@ -1220,8 +1220,6 @@ triggers Font Lock to recognize the change."

 (defun font-latex-jit-lock-force-redisplay (buf start end)
   "Compatibility for Emacsen not offering `jit-lock-force-redisplay'."
-  (if (fboundp 'jit-lock-force-redisplay)
-      (jit-lock-force-redisplay buf start end)
     ;; The following block is an expansion of `jit-lock-force-redisplay'
     ;; and involved macros taken from CVS Emacs on 2007-04-28.
     (with-current-buffer buf
@@ -1236,7 +1234,7 @@ triggers Font Lock to recognize the change."
                  buffer-file-truename)
              (put-text-property start end 'fontified t))
          (unless modified
-           (restore-buffer-modified-p nil)))))))
+           (restore-buffer-modified-p nil))))))

 (defun font-latex-fontify-region (beg end &optional loudly)
   "Fontify region from BEG to END.
--8<---------------cut here---------------end--------------->8---

IIUC you're suggesting to do something like
--8<---------------cut here---------------start------------->8---
--- a/font-latex.el
+++ b/font-latex.el
@@ -1221,7 +1221,10 @@ triggers Font Lock to recognize the change."
 (defun font-latex-jit-lock-force-redisplay (buf start end)
   "Compatibility for Emacsen not offering `jit-lock-force-redisplay'."
   (if (fboundp 'jit-lock-force-redisplay)
-      (jit-lock-force-redisplay buf start end)
+      (condition-case nil
+         (jit-lock-force-redisplay buf start end)
+       (wrong-number-of-arguments
+        (jit-lock-force-redisplay (copy-marker start) (copy-marker end))))
     ;; The following block is an expansion of `jit-lock-force-redisplay'
     ;; and involved macros taken from CVS Emacs on 2007-04-28.
     (with-current-buffer buf
--8<---------------cut here---------------end--------------->8---
Is it correct?

Thank you,
Mosè


2013/5/12 Ralf Angeli <address@hidden>:
> * Mosč Giordano (2013-05-12) writes:
>
>> you're right, `jit-lock-force-redisplay' has been recently changed,
>> see http://bzr.savannah.gnu.org/lh/emacs/trunk/revision/112357  Maybe
>> there is a good reason behind the change of arguments number, but I
>> simply don't know it.  I can see two solutions:
>> - report this to Emacs and make them restore backward compatibility;
>
> If this is done probably depends on how much the function is seen as
> being for internal use only.  Its use in AUCTeX is probably already a
> bit hackish but so is lots of compatibility code.
>
>> - implement `font-latex-jit-lock-force-redisplay' in AUCTeX without
>> relying on possibly existing `jit-lock-force-redisplay'.
>> As a temporary fix, I'm going to change
>> `font-latex-jit-lock-force-redisplay' following the second solution.
>
> Not sure what you have in mind, but we could also use something like the
> following.  Not that it is particulary nice.
>
>       (condition-case nil
>           (jit-lock-force-redisplay buf start end)
>         (error
>          (jit-lock-force-redisplay (copy-marker start) (copy-marker end))))
>
> --
> Ralf