Re: [RFC/PATCH 2/2] Teach commit to handle CHERRY_HEAD automatically

[Jay Soffian <jaysoffian@gmail.com>]

On Tue, Feb 15, 2011 at 6:00 PM, Jonathan Nieder <jrnieder@gmail.com> wrote:
>  - "git commit --amend" to say "I'm done fixing the broken thing".
>
>  - "git commit --fixup=HEAD/--squash=HEAD" to say "done fixing;
>   let's look at this again later and squash it when I am less
>   confused".
>
> Both are mistakes if HEAD is the previous and good commit rather than
> the broken one.  Maybe there is some simple safety that could protect
> against them?

As you see below, this is already protected against.

>>  --reset-author::
>> -     When used with -C/-c/--amend options, declare that the
>> -     authorship of the resulting commit now belongs of the committer.
>> -     This also renews the author timestamp.
>> +     When used with -C/-c/--amend options, or when committing after a
>> +     a conflicting cherry-pick,
>
> or when committing after a conflicted merge, no?

No. The person committing a merge is already the author of the merge,
why would they use --reset-author?

>
>>                                  declare that the authorship of the
>> +     resulting commit now belongs of the committer. This also renews
>> +     the author timestamp.
>
> How does it interact with --author?

No change from before, --author forces the author of the new commit.

>> +++ b/builtin/commit.c
> [...]
>> @@ -609,7 +609,7 @@ static int prepare_to_commit(const char *index_file, const char *prefix,
>>                       die_errno("could not read log file '%s'",
>>                                 logfile);
>>               hook_arg1 = "message";
>> -     } else if (use_message) {
>> +     } else if (use_message && !cherry_pick) {
>
> Wouldn't this make
>
>        git commit -C foo
>
> after a failed cherry-pick use MERGE_MSG instead of the message the
> user requested?

No, because the -C will force the cherry_pick flag to 0 in
parse_and_validate_options:

		/* Let message-specifying options override CHERRY_HEAD */

I'll make this logic clearer though as I need to address Junio's
earlier message.

>> +             if (cherry_pick)
>> +                     fprintf(fp,
>> +                             "#\n"
>> +                             "# It looks like you may be committing a cherry-pick.\n"
>
> Aside: shouldn't we suggest some porcelain-ish command (git merge
> --clear?  git commit --no-merge?) to remove .git/MERGE_HEAD instead of
> asking the committer to do it?

We have git merge --reset as an alias for reset --merge. Since reset
--merge takes care of this case too (I think, I'll check) we can
suggest that for both.

> This section is used to give a preview and sanity check for the
> commit.
>
>  - if we are on the wrong branch, that might be a mistake.
>  - if the author is someone else, that might be a mistake.
>  - if there are multiple parents, that might be a mistake.
>  - if there are changes included, some might be mistakes.
>  - if there are changes excluded, some might be mistakes,
>  - if there are untracked files, some might be mistakes.
>
> Where does committing a cherry-pick fall into that picture?  Maybe
>
>        # Author:    A U Thor <author@example.com>
>        # Date:      Tue Beb 9 01:23:45 1911 -0500
>        #
>        # If this is not correct, please try again using the
>        # --author and --date or --reset-author options.

Okay.

>> @@ -898,6 +907,7 @@ static void handle_untracked_files_arg(struct wt_status *s)
>>               die("Invalid untracked files mode '%s'", untracked_files_arg);
>>  }
>>
>> +
>
> Stray whitespace change?

Indeed.

> How can I get out of this state if I really do want to amend?

git reset, same as it ever was?

>> @@ -943,11 +955,19 @@ static int parse_and_validate_options(int argc, const char *argv[],
>>               die("Only one of -c/-C/-F/--fixup can be used.");
>>       if (message.len && f > 0)
>>               die("Option -m cannot be combined with -c/-C/-F/--fixup.");
>> +     if (cherry_pick) {
>> +             /* Let message-specifying options override CHERRY_HEAD */
>> +             if (f > 0 || message.len)
>> +                     cherry_pick = 0;
>> +             else
>> +                     /* used for authorship side-effect only */
>> +                     use_message = "CHERRY_HEAD";
>> +     }
>
> Hmm, what if I have commits F and F' and after trying to cherry-pick F
> I decide I want the message from F'?

I don't think I understand. commit -c F', or just commit (w/o options)
and you get MERGE_MSG generated by cherry-pick.

>> @@ -1118,6 +1138,7 @@ int cmd_status(int argc, const char **argv, const char *prefix)
>>       gitmodules_config();
>>       git_config(git_status_config, &s);
>>       in_merge = file_exists(git_path("MERGE_HEAD"));
>> +     cherry_pick = file_exists(git_path("CHERRY_HEAD"));
>
> Is it possible to be both in_merge and cherry_pick?

No, I need to rework this into an enum, maybe enum { CONFLICT_NONE,
CONFLICT_MERGE, CONFLICT_CHERRY_PICK } conflict_type.

>> -                     reflog_msg = "commit";
>> +                     reflog_msg = cherry_pick ? "commit (cherry-pick)"
>> +                                              : "commit";
>
> Nice.  Probably worth mentioning in the commit message.
>
>> +                     elif [ -f "$g/CHERRY_HEAD" ]; then
>> +                             r="|CHERRY-PICKING"
>
> Likewise.

Right.

Thanks for the feedback!

j.