Jeff King wrote:
> You cc'd me, which I guess means you git-blame'd the line in question.
> But you really need to parent-blame about five steps back
Worse, I used ‘git log -- builtin-revert.c’ and stopped reading when
it seemed you must know be current maintainer for that line. Sorry
about that.
> we could actually make it stash the
> original authorship information somewhere (in addition to the commit
> message template) and then pull it out automatically.
I think that can wait for cherry-pick --continue.
Here’s an ugly patch to use --author. I call it ugly because the
relevant part of t3507-cherry-pick-conflict.sh output looks like
this:
| Automatic cherry-pick failed. After resolving the conflicts,
| mark the corrected paths with 'git add <paths>' or 'git rm <paths>'
| and commit the result with:
|
| git commit --author='A U Thor <author@xxxxxxxxxxx>' \
| --date='1112912113 -0700'
In other words, the command is long and the date human-unfriendly.
Anyway, given that cherry-pick --continue is just around the corner, I
withdraw my complaint about the ‘commit -c’ version.
Thanks for the pointer.
diff --git a/builtin/revert.c b/builtin/revert.c
index bbafc41..9b8e7d6 100644
--- a/builtin/revert.c
+++ b/builtin/revert.c
@@ -252,9 +252,16 @@ static char *help_msg(void)
"and commit the result");
if (action == CHERRY_PICK) {
+ const char *ident = git_author_info(IDENT_ERROR_ON_NO_NAME);
+ const char *ident_end = ident ? strchr(ident, '>') : NULL;
+ if (!ident_end || ident_end[1] != ' ')
+ die("git_author_info returned nonsense");
strbuf_addf(&helpbuf, " with: \n"
"\n"
- " git commit\n");
+ " git commit --author='%.*s' \\\n"
+ " --date='%s'\n",
+ (int) (ident_end + 1 - ident), ident,
+ ident_end + 2);
}
else
strbuf_addch(&helpbuf, '.');
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html