Re: [PATCH v2 0/3] am: let command-line options override saved options

From: Junio C Hamano <gitster@pobox.com>
Date: Aug 4, 2015, 21:12 UTC
Message-ID: <xmqqbnem69mm.fsf@gitster.dls.corp.google.com>
In-Reply-To: <1438697116-27799-1-git-send-email-pyokagan@gmail.com>

Paul Tan <pyokagan@gmail.com> writes:

Show 31 quoted lines

> Let command-line options override saved options in git-am when resuming
>
> This is a re-roll of [v1]. Previous versions:
>
> [v1] http://thread.gmane.org/gmane.comp.version-control.git/274789
>
> When resuming, git-am mistakenly ignores command-line options.
>
> For instance, when a patch fails to apply with "git am patch", subsequently
> running "git am --3way" would not cause git-am to fall back on attempting a
> threeway merge.  This occurs because by default the --3way option is saved as
> "false", and the saved am options are loaded after the command-line options are
> parsed, thus overwriting the command-line options when resuming.
>
> [PATCH 1/3] tweaks test-terminal.perl to redirect the stdin of the child
> process to a pty. This is to support the tests in [PATCH 2/3].
>
> [PATCH 2/3] fixes builtin/am.c, enabling command-line options to override saved
> options. However, even with this patch, the following command-line options have
> no effect when resuming:
>
> * --signoff overriding --no-signoff
>
> * --no-keep overriding --keep
>
> * --message-id overriding --no-message-id
>
> * --scissors overriding --no-scissors
>
> This is because they are only taken into account during the mail-parsing stage,
> which is skipped over when resuming.

It is more like "which has already happened", so I would tend to think that these are the right things to ignore. Otherwise, a pretty common sequence would not work well ...

    $ git am mbox
    ... conflicted ...
    $ edit .git/rebase-apply/patch
    $ git am

... if the "resuming" invocation re-split the message by running mailinfo again, the edit by the user will be lost.