emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Bugfix


From: Chong Yidong
Subject: Bugfix
Date: Thu, 28 Feb 2008 22:24:48 -0500

Joe Wells <address@hidden> wrote:

> BUG #1:  An overlay's after-string property that would appear at the
> end of the buffer is not displayed, if the same overlay also has a
> display property and an immediately preceding overlay also has an
> after-string property.  (Putting extra characters at the end of the
> buffer works around this bug.)

To reproduce:

  (progn
    (if (get-buffer "*test*")
        (kill-buffer "*test*"))
    (pop-to-buffer "*test*")
    (erase-buffer)
    (insert "ABCD")
    (let ((o1 (make-overlay 2 3))
          (o2 (make-overlay 3 4))
          (s " "))
      (overlay-put o1 'after-string "1")
      (overlay-put o1 'display s)
      (overlay-put o2 'display s)))

I believe is a pretty safe and straightforward fix, but since it's in
the redisplay engine, it'd be good if someone double-checks before I
check it into the branch.  (It's already in the trunk.)

The bug is due to an optimization in next_overlay_string that tells the
iterator to stop scanning if we're at the end of the buffer and there
are no more overlay strings to process.  This fails when there's a
display string, because it->end_charpos gives the length of the string
instead of the length of the buffer.  In this case, we simply shouldn't
set overlay_strings_at_end_processed_p.


*** emacs/src/xdisp.c.~1.1149.2.23.~    2008-02-28 20:59:38.000000000 -0500
--- emacs/src/xdisp.c   2008-02-28 22:02:16.000000000 -0500
***************
*** 4677,4683 ****
        /* If we're at the end of the buffer, record that we have
         processed the overlay strings there already, so that
         next_element_from_buffer doesn't try it again.  */
!       if (IT_CHARPOS (*it) >= it->end_charpos)
        it->overlay_strings_at_end_processed_p = 1;
  
        /* If we have to display `...' for invisible text, set
--- 4677,4683 ----
        /* If we're at the end of the buffer, record that we have
         processed the overlay strings there already, so that
         next_element_from_buffer doesn't try it again.  */
!       if (NILP (it->string) && IT_CHARPOS (*it) >= it->end_charpos)
        it->overlay_strings_at_end_processed_p = 1;
  
        /* If we have to display `...' for invisible text, set




reply via email to

[Prev in Thread] Current Thread [Next in Thread]