From: Konstantin Kharlamov <hi-angel@yandex.ru>
To: Eli Zaretskii <eliz@gnu.org>
Cc: 37395@debbugs.gnu.org
Subject: bug#37395: [PATCH] diff-mode.el: take into account patch separators
Date: Mon, 16 Sep 2019 23:31:10 +0300 [thread overview]
Message-ID: <1568665870.19637.0@yandex.ru> (raw)
In-Reply-To: <1568357907.5095.0@yandex.ru>
On Пт, сен 13, 2019 at 09:58, Konstantin Kharlamov
<hi-angel@yandex.ru> wrote:
>
>
> On Пт, сен 13, 2019 at 09:14, Eli Zaretskii <eliz@gnu.org> wrote:
>>> From: Konstantin Kharlamov <Hi-Angel@yandex.ru>
>>> Date: Fri, 13 Sep 2019 00:34:45 +0300
>>>
>>> * lisp/vc/diff-mode.el (diff-goto-line-before-patch-separator): an
>>> inline function to check if prev. line was git-format-patch
>>> \x7f\x7fseparator,
>>> in which case go there.
>>> (diff-end-of-hunk): make use of
>>> \x7f\x7f(diff-goto-line-before-patch-separator)
>>
>> The descriptions of changes should start with a capital letter.
>> Also,
>> your lines in the commit log message are too long, they should not
>> exceed 61 characters (because in the release tarball we create a
>> ChangeLog file from them, which prepends a TAB character to each
>> line).
>>
>>> +(defsubst diff-goto-line-before-patch-separator ()
>>> + "Go to prev. line, then if it has patch separator as produced
>>> +by git-format-patch, stay there. Otherwise go back."
>>
>> The first line of a doc string should be a complete sentence. I
>> suggest to rephrase as follows:
>>
>> Return buffer position before patch separator produced by
>> \x7fgit-format-patch.
>
> Thank you, I'll address the comments a bit later this day!
>
>>> + (previous-line)
>>> + (when (not (looking-at "-- "))
>>> + (next-line))
>>> + (point))
>>
>> Btw, Diff mode is more general than just Git-produced diffs. Is
>> there
>> any possibility that this change will misfire in diffs produced by
>> other tools? If so, perhaps we should also verify the buffer is
>> under
>> Git control.
>
> Oh, while writing this, I figured I should change the regexp from `--
> ` to `^-- $`. Will do.
>
> With that addressed: very unlikely. A miscalculation could happen if
> all of the following holds true *simultaneously*:
>
> * The diff removes a line with the exact text `- `, i.e. two
> characters {dash,space}. That would result in {dash,dash,space} diff,
> same as patch separator.
> * The line was the last line removed in the hunk.
> * The hunk has no context after the removed line
>
> Note, these need to hold true simultaneously. Give the low
> probability of this compared to git-format-patch that is used
> literally everywhere (and that without this patch diff-mode can't
> handle it), I think this is a reasonable trade-off.
While still on it: I figured I could detect whether diff has git-diff
format by searching the `diff --git` text that seems to be always
present when starting diff-mode, and only then try to find the
patch-separator. Done. I'm not sure if the `setq-local foo` and then
modification of the `foo` in `(define-derived-mode)` is the preferred
way to do this, but otherwise it seems to be fine, and worked in my
tests.
next prev parent reply other threads:[~2019-09-16 20:31 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-12 21:33 bug#37395: Diff-mode doesn't take into account patch-separators as produced by git-format-patch Konstantin Kharlamov
2019-09-12 21:34 ` bug#37395: [PATCH] diff-mode.el: take into account patch separators Konstantin Kharlamov
2019-09-13 6:14 ` Eli Zaretskii
2019-09-13 6:58 ` Konstantin Kharlamov
2019-09-16 20:31 ` Konstantin Kharlamov [this message]
2019-09-13 9:19 ` bug#37395: [PATCH v2] " Konstantin Kharlamov
2019-09-16 20:26 ` bug#37395: [PATCH v3] " Konstantin Kharlamov
2019-10-07 4:39 ` Lars Ingebrigtsen
2019-10-07 23:04 ` Konstantin Kharlamov
2019-10-07 23:12 ` Konstantin Kharlamov
2019-10-08 15:59 ` Lars Ingebrigtsen
2019-10-08 20:08 ` Konstantin Kharlamov
2019-10-08 19:34 ` bug#37395: [PATCH v4] " Konstantin Kharlamov
2019-10-09 19:36 ` Lars Ingebrigtsen
2019-10-09 20:08 ` Konstantin Kharlamov
2019-10-09 20:07 ` bug#37395: [PATCH v5] " Konstantin Kharlamov
2019-10-13 3:53 ` Lars Ingebrigtsen
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=1568665870.19637.0@yandex.ru \
--to=hi-angel@yandex.ru \
--cc=37395@debbugs.gnu.org \
--cc=eliz@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).