unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Juri Linkov <juri@linkov.net>
To: Dmitry Gutov <dgutov@yandex.ru>
Cc: 52349@debbugs.gnu.org, Sean Whitton <spwhitton@spwhitton.name>
Subject: bug#52349: 29.0.50; vc-git and diff-mode: stage hunks
Date: Mon, 19 Sep 2022 09:50:18 +0300	[thread overview]
Message-ID: <86leqf6bwl.fsf@mail.linkov.net> (raw)
In-Reply-To: <78b1100c-89a6-5daf-548f-778807105307@yandex.ru> (Dmitry Gutov's message of "Mon, 19 Sep 2022 05:09:19 +0300")

> Do you know if Git use a stable ordering for files?

It seems the ordering is stable unless it's changed explicitly
by command line options.

> If not, here's where the proposed implementation might fail:
>
> Suppose we have the staging area with contents
>
> a/bar b/bar
> +bbb
> a/foo b/foo
> +aaa
> +ccc
>
> and the diff to check in with contents
>
> a/foo b/foo
> +aaa
> a/bar b/bar
> +bbb
> +ccc
>
> ...then the check will succeed, I think.

This check is intended to detect only added/deleted files
that get into the staging area.

> Even if Git always sorts them the same, I suppose an externally-produced
> diff could trigger this scenario. It should be a pretty rare case, though,
> so it's probably fine.
>
> A tweak like the following could fix it, though: instead of replacing the
> chunks with "", maybe keep the file headers. Then the remaining contents
> for the same file in vc-git-patch-string wouldn't be able to stick to the
> previous file's chunks.

This looks too complicated.  And indeed, this is a rare case, so maybe
something like this could be added when this will became a real problem.

>>> Perhaps if we called diff-file-next in addition to (move-beginning-of-line
>>> 1)?
>> diff-file-next doesn't work: it stops in the middle of the diff file
>> header.
>> Therefore the patch navigates git diff file headers explicitly.
>>
>>> But we need to save both strings: and ensure that if there is a match
>>> for that file header, then all hunks are contained in the patch as well
>>> (and nothing extra, for that file).
>> This is implemented by the patch.  Maybe it could help you to see how it
>> works
>> by running it under debugger and stepping through it.
>>
>>> It's complex logic, so if you manage to write a test as well, that would be
>>> excellent.
>> A test could written when someone will create infrastructure for testing
>> git commands with helpers to create a git repository and checking its content.
>
> test/lisp/vc/vc-tests.el actually contains a helper like this.
>
> Every scenario starts with calling vc-test--create-repo-function, and there
> are tests for 'version-diff' at the very end of the file. It's somewhat
> convoluted, so I don't blame you for missing it.

Actually, I looked into test/lisp/vc/vc-git-tests.el that is almost empty.
I expected that since this check is git-specific, it should be in
vc-git-tests.el.





  reply	other threads:[~2022-09-19  6:50 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-12-07  8:59 bug#52349: 29.0.50; vc-git and diff-mode: stage hunks Manuel Uberti via Bug reports for GNU Emacs, the Swiss army knife of text editors
2021-12-07 17:03 ` Juri Linkov
2021-12-07 17:08   ` Manuel Uberti via Bug reports for GNU Emacs, the Swiss army knife of text editors
2021-12-07 19:04     ` Juri Linkov
2021-12-07 17:10 ` Dmitry Gutov
2021-12-07 19:06   ` Juri Linkov
2021-12-08  2:29     ` Dmitry Gutov
2021-12-08 18:57       ` Juri Linkov
2021-12-09 23:17         ` Dmitry Gutov
2022-02-08 19:57           ` Juri Linkov
2022-02-12  1:43             ` Dmitry Gutov
2022-02-12 19:42               ` Juri Linkov
2022-02-13  1:32                 ` Dmitry Gutov
2022-02-13 19:48                   ` Juri Linkov
2022-02-13 22:51                     ` Dmitry Gutov
2022-02-13 19:56                   ` Juri Linkov
2022-02-13 22:41                     ` Dmitry Gutov
2022-08-21 16:07                       ` Juri Linkov
2022-02-13 22:37                   ` Dmitry Gutov
2022-08-21 18:53                     ` Juri Linkov
2022-08-24  2:06                       ` Dmitry Gutov
2022-08-24 18:20                         ` Juri Linkov
2022-08-24 20:20                           ` Dmitry Gutov
2022-08-27 20:07                             ` Juri Linkov
2022-08-28  0:40                               ` Dmitry Gutov
2022-08-28 19:45                                 ` Juri Linkov
2022-08-28 23:34                                   ` Dmitry Gutov
2022-09-08 19:29                             ` Juri Linkov
2022-09-08 22:48                               ` Sean Whitton
2022-09-10  1:36                                 ` Dmitry Gutov
2022-09-11 15:05                                   ` Juri Linkov
2022-09-11 21:57                                     ` Dmitry Gutov
2022-09-12 18:19                                       ` Juri Linkov
2022-09-19  2:09                                         ` Dmitry Gutov
2022-09-19  6:50                                           ` Juri Linkov [this message]
2022-09-19 12:57                                             ` Dmitry Gutov
2022-10-02 18:51                                               ` Juri Linkov
2022-11-04  1:32                                                 ` Dmitry Gutov
2022-08-26 17:11                       ` Filipp Gunbin
2022-08-27 19:56                         ` Juri Linkov
2021-12-07 20:07 ` 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=86leqf6bwl.fsf@mail.linkov.net \
    --to=juri@linkov.net \
    --cc=52349@debbugs.gnu.org \
    --cc=dgutov@yandex.ru \
    --cc=spwhitton@spwhitton.name \
    /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).