Jonathan Tan <jonathantanmy@xxxxxxxxxx> writes:
>> From: Heba Waly <heba.waly@xxxxxxxxx>
>>
>> Display hints to the user when trying to commit without staging the modified
>> files first (when advice.statusHints is set to true). Change the output of the
>> unsuccessful commit from e.g:
>
> Wrap your commit messages at 72 characters.
>
>> # [...]
>> # Changes not staged for commit:
>> # modified: builtin/commit.c
>> #
>> # no changes added to commit
>>
>> to:
>>
>> # [...]
>> # Changes not staged for commit:
>> # (use "git add <file>..." to update what will be committed)
>> # (use "git checkout -- <file>..." to discard changes in working directory)
>> #
>> # modified: ../builtin/commit.c
>
> For tidiness, can this line also be "builtin/commit.c" (that is, without
> the "../" at the beginning) to match what's before "to:"?
>
>> In ea9882bfc4 (commit: disable status hints when writing to COMMIT_EDITMSG,
>> 2013-09-12) the intent was to disable status hints when writing to
>> COMMIT_EDITMSG, but in fact the implementation disabled status messages in
>> more locations, e.g in case the commit wasn't successful, status hints
>> will still be disabled and no hints will be displayed to the user although
>> advice.statusHints is set to true.
>>
>> Signed-off-by: Heba Waly <heba.waly@xxxxxxxxx>
>> ---
>> builtin/commit.c | 1 +
>> t/t7500-commit-template-squash-signoff.sh | 9 +++++++++
>
> I wondered if there was a better place to put the test, but I couldn't
> find one, so this is fine.
>
>> @@ -961,6 +961,7 @@ static int prepare_to_commit(const char *index_file, const char *prefix,
>> */
>> if (!committable && whence != FROM_MERGE && !allow_empty &&
>> !(amend && is_a_merge(current_head))) {
>> + s->hints = advice_status_hints;
>> s->display_comment_prefix = old_display_comment_prefix;
>> run_status(stdout, index_file, prefix, 0, s);
>> if (amend)
>
> I checked that this undoing of "s->hints = 0" is safe, because s is no
> longer used in this function nor in the calling function cmd_commit()
> (which is the one that declared s locally).
>
> Still probably worth a comment, though. For example:
>
> This status is to be printed to stdout, so hints will be useful to the
> user. Reset s->hints to what the user configured.
>
> The corresponding comment on "s->hints = 0" might need to be tweaked,
> too, but I can't think of anything at the moment.
>
>> +test_expect_success 'commit without staging files fails and displays hints' '
>> + echo "initial" >>file &&
>> + git add file &&
>> + git commit -m initial &&
>> + echo "changes" >>file &&
>> + test_must_fail git commit -m initial >actual &&
>
> Use another commit message for this, since this is no longer "initial".
> (Maybe "after initial" or something like that.)
Thanks for a careful review.
