git@vger.kernel.org list mirror (unofficial, one of many)
 help / color / Atom feed
From: Andreas Heiduk <asheiduk@gmail.com>
To: Duy Nguyen <pclouds@gmail.com>
Cc: Eric Sunshine <sunshine@sunshineco.com>, Git List <git@vger.kernel.org>
Subject: Re: [PATCH/RFC] completion: complete all possible -no-<options>
Date: Mon, 14 May 2018 21:58:21 +0200
Message-ID: <bda62b25-a569-9199-5973-2ea7b223ed5d@gmail.com> (raw)
In-Reply-To: <CACsJy8CwwseqL9M=-ML6hwgHG_GSK6pfeQU+MGWsx6=WCCZtmQ@mail.gmail.com>

Am 14.05.2018 um 19:26 schrieb Duy Nguyen:
> On Mon, May 14, 2018 at 7:03 PM, Andreas Heiduk <asheiduk@gmail.com> wrote:
>> Am 08.05.2018 um 17:24 schrieb Duy Nguyen:
>>> On Mon, Apr 23, 2018 at 7:36 AM, Eric Sunshine <sunshine@sunshineco.com> wrote:
>>>> I haven't looked at the implementation, so this may be an entirely
>>>> stupid suggestion, but would it be possible to instead render the
>>>> completions as?
>>>>
>>>>     % git checkout --<tab>
>>>>     --[no-]conflict=                   --[no-]patch
>>>>     --[no-]detach                      --[no-]progress
>>>>     --[no-]ignore-other-worktrees      --[no-]quiet
>>>>     --[no-]ignore-skip-worktree-bits   --[no-]recurse-submodules
>>>>     --[no-]merge                       --theirs
>>>>     --[no-]orphan=                     --[no-]track
>>>>     --ours
>>>>
>>>> This would address the problem of the --no-* options taking double the
>>>> screen space.
>>>
>>> It took me so long to reply partly because I remember seeing some guy
>>> doing clever trick with tab completion that also shows a short help
>>> text in addition to the complete words. I could not find that again
>>> and from my reading (also internet searching) it's probably not
>>> possible to do this without trickery.
>>
>> The fish-shell does something like that.
>>
>>     > git status --<tab here>
>>     --branch  (Show the branch and tracking info even in short-format)
>>     --help                       (Display the manual of a git command)
>>     --ignore-submodules                 (Ignore changes to submodules)
>>     --porcelain    (Give the output in a stable, easy-to-parse format)
>>     --short                      (Give the output in the short-format)
>>     --untracked-files              (The untracked files handling mode)
>>
>> Another tab will put a selection-cursor on the displayed list - you can
>> navigate that list with Cursor-Up/Cursor-Down, select an entry and that
>> entry will be inserted into the commandline. That selection process
>> would be useless if the options are presented as "--[no-]x" because THAT
>> cannot be inserted into the commandline without manual editing. And
>> that's the point of the fast option selection process.
> 
> Good to know.
> 
> BTW I looked at the git.fish completion script [1] and see that recent
> effort to help automate more in git-completion.bash might help there
> too. I notice a lot of options and help text hard coded there, if
> someone can explain to me how git.fish uses those, maybe I can change
> git to export something suitable for git.fish to use too [2].

I'm no expert, but some additional things required by fish (and I
suppose zsh too) but not by bash:

- grouping of long and short options
- help text
- argument types for options

Help text and long/short option grouping look like this:

    > git rebase -<tab>
    --force-rebase  -f                                (Force the rebase)
    --merge  -m                       (Use merging strategies to rebase)

All these infos seem to be available in `struct option` (for C stuff
at least). So I guess It would be easiest for Fish & Co if git just
exports the complete info in some stable format.

> 
> For example with latest git (in 'master') doing this
> 
>     ./git add --git-completion-helper
> 
> gives you the list of all options of "git add". Giving the help text
> for each option is definitely possible (I just didn't see any use for
> it until I looked at zsh/fish completion scripts) and maybe more in
> the future.
> 
> [1] https://github.com/fish-shell/fish-shell/blob/master/share/completions/git.fish
> [2] But then if your script has to work with old git versions too then
> this is a moot point.

Well, sooner or later those old git versions might not be supported by
those shells exactly due to the involved maintenance overhead. So
providing some helper is a step in the right direction. Not providing
only fossilizes the current state.

  reply index

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-17 18:13 Nguyễn Thái Ngọc Duy
2018-04-18  3:43 ` Junio C Hamano
2018-04-18 15:08   ` Duy Nguyen
2018-04-23  5:36 ` Eric Sunshine
2018-05-08 15:24   ` Duy Nguyen
2018-05-08 16:39     ` Stefan Beller
2018-05-09  3:20     ` Aaron Schrab
2018-05-14 17:14       ` Duy Nguyen
2018-05-14  3:33     ` Eric Sunshine
2018-05-14 16:39       ` Duy Nguyen
2018-05-14 17:03     ` Andreas Heiduk
2018-05-14 17:26       ` Duy Nguyen
2018-05-14 19:58         ` Andreas Heiduk [this message]
2018-05-27  8:38 ` [PATCH v2 0/3] " Nguyễn Thái Ngọc Duy
2018-05-27  8:38   ` [PATCH v2 1/3] parse-options: option to let --git-completion-helper show negative form Nguyễn Thái Ngọc Duy
2018-05-27  8:38   ` [PATCH v2 2/3] completion: suppress some -no- options Nguyễn Thái Ngọc Duy
2018-05-27  8:38   ` [PATCH v2 3/3] completion: collapse extra --no-.. options Nguyễn Thái Ngọc Duy
2018-05-29 18:48     ` Stefan Beller
2018-05-29 19:04       ` Duy Nguyen
2018-06-06  9:41   ` [PATCH v3 0/3] ompletion: complete all possible -no-<options> Nguyễn Thái Ngọc Duy
2018-06-06  9:41     ` [PATCH v3 1/3] parse-options: option to let --git-completion-helper show negative form Nguyễn Thái Ngọc Duy
2018-06-06  9:41     ` [PATCH v3 2/3] completion: suppress some -no- options Nguyễn Thái Ngọc Duy
2018-06-06  9:41     ` [PATCH v3 3/3] completion: collapse extra --no-.. options Nguyễn Thái Ngọc Duy

Reply instructions:

You may reply publically to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: http://vger.kernel.org/majordomo-info.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bda62b25-a569-9199-5973-2ea7b223ed5d@gmail.com \
    --to=asheiduk@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=pclouds@gmail.com \
    --cc=sunshine@sunshineco.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

git@vger.kernel.org list mirror (unofficial, one of many)

Archives are clonable:
	git clone --mirror https://public-inbox.org/git
	git clone --mirror http://ou63pmih66umazou.onion/git
	git clone --mirror http://czquwvybam4bgbro.onion/git
	git clone --mirror http://hjrcffqmbrq6wope.onion/git

Newsgroups are available over NNTP:
	nntp://news.public-inbox.org/inbox.comp.version-control.git
	nntp://ou63pmih66umazou.onion/inbox.comp.version-control.git
	nntp://czquwvybam4bgbro.onion/inbox.comp.version-control.git
	nntp://hjrcffqmbrq6wope.onion/inbox.comp.version-control.git
	nntp://news.gmane.org/gmane.comp.version-control.git

 note: .onion URLs require Tor: https://www.torproject.org/

AGPL code for this site: git clone https://public-inbox.org/ public-inbox