Re: [RFC/PATCH v4 6/6] completion: add public _GIT_complete helper

[Felipe Contreras <felipe.contreras@gmail.com>]

On Mon, May 7, 2012 at 8:56 PM, Junio C Hamano <gitster@pobox.com> wrote:
> Felipe Contreras <felipe.contreras@gmail.com> writes:
>
>> So that users can easily define aliases, such as:
>>
>>  _GIT_complete gf git_fetch
>>
>> Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
>> ---
>>  contrib/completion/git-completion.bash |   10 +++++-----
>>  t/t9902-completion.sh                  |    9 ++++++++-
>>  2 files changed, 13 insertions(+), 6 deletions(-)
>>
>> diff --git a/contrib/completion/git-completion.bash b/contrib/completion/git-completion.bash
>> index 049110e..2b7ef02 100755
>> --- a/contrib/completion/git-completion.bash
>> +++ b/contrib/completion/git-completion.bash
>> @@ -2678,6 +2678,7 @@ _gitk ()
>>
>>  __git_func_wrap ()
>>  {
>> +     local cmd="${1#git_}" cmd_pos=1
>>       if [[ -n ${ZSH_VERSION-} ]]; then
>>               emulate -L bash
>>               setopt KSH_TYPESET
>> @@ -2695,8 +2696,7 @@ __git_func_wrap ()
>>       _$1
>>  }
>>
>> -# this is NOT a public function; use at your own risk
>> -__git_complete ()
>> +_GIT_complete ()
>
> If it is now a public function, please have some description as to how to
> use it for people who find this in the tarball extract.

This is RFC, I'm not planning to push for this, I truly have given up on it.

> I am guessing that
>
>        _GIT_complete frotz git_fetch
>
> is a way to declare that 'git frotz' wants the same kind of completion as
> 'git fetch' command, but I am not sure, as if it were the case it strikes
> me somewhat odd that it is not "_GIT_complete frotz fetch".

Yeah, I think that would be nicer, but one would need to take into
consideration the git/gitk cases.

Hopefully somebody else would pick this up and go through the eternal
nitpicking.

Cheers.

-- 
Felipe Contreras