Re: [PATCH v2 0/4] In PS1 prompt, make upstream state indicators consistent with other state indicators

[Junio C Hamano <gitster@pobox.com>]

Ævar Arnfjörð Bjarmason <avarab@gmail.com> writes:

> On Sun, Feb 27 2022, Justin Donnelly via GitGitGadget wrote:
>
>> These patches are about the characters and words that can be configured to
>> display in the PS1 prompt after the branch name. I've been unable to find a
>> consistent terminology. I refer to them as follows: [short | long] [type]
>> state indicator where short is for characters (e.g. ?), long is for words
>> (e.g. |SPARSE), and type is the type of indicator (e.g. sparse or upstream).
>> I'd be happy to change the commit messages to a different terminology if
>> that's preferred.
>>
>> There are a few inconsistencies with the PS1 prompt upstream state indicator
>> (GIT_PS1_SHOWUPSTREAM).
>>
>>  * With GIT_PS1_SHOWUPSTREAM="auto", if there are no other short state
>>    indicators (e.g. + for staged changes, $ for stashed changes, etc.), the
>>    upstream state indicator appears adjacent to the branch name (e.g.
>>    (main=)) instead of being separated by SP or GIT_PS1_STATESEPARATOR (e.g.
>>    (main =)).
>>  * If there are long state indicators (e.g. |SPARSE), a short upstream state
>>    indicator (i.e. GIT_PS1_SHOWUPSTREAM="auto") is to the right of the long
>>    state indicator (e.g. (main +|SPARSE=)) instead of with the other short
>>    state indicators (e.g. (main +=|SPARSE)).
>>  * The long upstream state indicator (e.g. GIT_PS1_SHOWUPSTREAM="verbose")
>>    is separated from other (short or long) state indicators by a hard-coded
>>    SP. Other long state indicators are separated by a hard-coded pipe (|).
>>
>> These patches are to make the upstream state indicators more consistent with
>> other state indicators.
>>
>> ----------------------------------------------------------------------------
>>
>> Changes since v1:
>>
>>  * Added __git_ps1 examples and before/after tables to commit messages where
>>    applicable. This should make it clearer what the behavior is for other
>>    (not upstream) state indicators, and how the patches make the upstream
>>    state indicator more consistent.
>>  * Removed some extraneous information about long state indicators from
>>    patch 2 commit message. This wasn't really helpful, and was a
>>    distraction.
>
> Since this was all in reponse to my review: I've looked this over again
> and this all LGTM now:
>
> Reviewed-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>

Thanks, both.