unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Gergely Risko <gergely@risko.hu>
To: 11773@debbugs.gnu.org
Subject: bug#11773: 24.1; [PATCH] Handle first change undo marker correctly in case of insert-file-contents
Date: Sun, 24 Jun 2012 19:43:16 +0200	[thread overview]
Message-ID: <87pq8od2rv.fsf@gergely.risko.hu> (raw)

Tags: patch

I've run into the following problem the other day:
-=-
(defun common ()
  (save-buffer)
  (setq buffer-undo-list nil))

(defun insert-bad ()
  (interactive)
  (common)
  (insert-file-contents "/etc/fstab"))

(defun insert-good ()
  (interactive)
  (common)
  (insert "foobar"))
-=-

If I open a file and run insert-good and then I undo, the file goes back
to unmodified state.

If I open a file and run insert-bad and then I undo, the file doesn't
go back to unmodified state.

Investigating, it turned out that if insert-file-contents is ran for
an unmodified buffer, the first change marker of (t HIGH . LOW) is not
inserted although it should've been.

This is because of a bug in decode_string at coding.c, where by the time
the record_insert is ran at the end, MODIFF is already too high.
MODIFF gets increased in insert_from_gap (used by produce_chars).  But
at the point of increase there is a record_insert before:
-=-
  record_insert (GPT, nchars);
  MODIFF++;
-=-

So why it's not having the effect of recording first change?  Because
decode_coding disables undo list handling at the beginning of the
function.  Therefore, my solution is to record the first change (if
MODIFF <= SAVE_MODIFF) before the undo list handling gets disabled.

I attached the patch, please consider applying it.

Thanks,
Gergely Risko

diff --git a/src/coding.c b/src/coding.c
index eb89563..212e73e 100644
--- a/src/coding.c
+++ b/src/coding.c
@@ -7083,6 +7083,15 @@ decode_coding (struct coding_system *coding)
        set_buffer_internal (XBUFFER (coding->dst_object));
       if (GPT != PT)
        move_gap_both (PT, PT_BYTE);
+
+      /* We have to disable undo_list, so we can record the whole
+        insert transaction via record_insert at the end.
+        Unfortunately, disabling the undo recording also disables the
+        recording of the first change in the undo_list.  Therefore
+        we check for first change here and record it via
+        record_first_change if needed.  */
+      if (MODIFF <= SAVE_MODIFF)
+       record_first_change ();
       undo_list = BVAR (current_buffer, undo_list);
       BVAR (current_buffer, undo_list) = Qt;
     }





             reply	other threads:[~2012-06-24 17:43 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-06-24 17:43 Gergely Risko [this message]
2012-08-14  5:09 ` bug#11773: 24.1; [PATCH] Handle first change undo marker correctly in case of insert-file-contents Chong Yidong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/emacs/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87pq8od2rv.fsf@gergely.risko.hu \
    --to=gergely@risko.hu \
    --cc=11773@debbugs.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).