git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Duy Nguyen <pclouds@gmail.com>
To: Jonathan Nieder <jrnieder@gmail.com>
Cc: "Git Mailing List" <git@vger.kernel.org>,
	"Kirill Müller" <kirill.mueller@ivt.baug.ethz.ch>,
	"Junio C Hamano" <gitster@pobox.com>
Subject: Re: [PATCH v2] checkout: add --no-widen for restoring files in sparse checkout mode
Date: Mon, 1 Apr 2013 12:13:53 +0700	[thread overview]
Message-ID: <CACsJy8BVcu2VwxgQ9URViHMx2+6436npcYSHyZ5Mj0aVhVFtug@mail.gmail.com> (raw)
In-Reply-To: <20130401044816.GI11704@elie.Belkin>

On Mon, Apr 1, 2013 at 11:48 AM, Jonathan Nieder <jrnieder@gmail.com> wrote:
> Hi,
>
> Nguyễn Thái Ngọc Duy wrote:
>
>> "git checkout -- <paths>" is usually used to restore all modified
>> files in <paths>. In sparse checkout mode, this command is overloaded
>> with another meaning: to add back all files in <paths> that are
>> excluded by sparse patterns.
>>
>> Add "--no-widen" option to do what normal mode does: restore all
>> modified files and nothing else.
>
> In an ideal world, I would like "git checkout --widen" to modify the
> .git/info/sparse-checkout file, to be able to do:
>
>         git clone --sparse-checkout=Documentation git://repo.or.cz/git.git
>         cd git
>         git checkout --widen -- README COPYING INSTALL
>
> and hack on a tree with Documentation/, README, COPYING, and INSTALL
> present with no actual code to distract.  And "git checkout --no-widen"
> could be a way to ask to respect the existing sparse pattern.

In an ideal world, I would spend more time on this and add
--edit-sparse, which opens up $EDITOR, lets you edit the patterns and
reapplies the patterns after $EDITOR exits (catching faults if
possible). Unfortunately I don't use sparse checkout as much as before
and therefore have little motivation to do it. I would really like
narrow clone to replace sparse checkout, but I haven't made much
progress on that front either. I'll try to get back on that once
pathspec-magics topic is settled.

> This patch isn't about tweaking the sparse-checkout pattern; instead,
> it's about how "git checkout" interacts with the skip-worktree bit.
> Maybe a good name would be --respect-skip-worktree?

I'm bad at naming. If nobody objects, I'll take this as the new option name.

>> --- a/Documentation/git-checkout.txt
>> +++ b/Documentation/git-checkout.txt
>> @@ -180,6 +180,17 @@ branch by running "git rm -rf ." from the top level of the working tree.
>>  Afterwards you will be ready to prepare your new files, repopulating the
>>  working tree, by copying them from elsewhere, extracting a tarball, etc.
>>
>> +--no-widen::
>> +     In sparse checkout mode, `git checkout -- <paths>` would
>> +     update all entries matched by <paths> regardless of sparse
>> +     patterns. This option only updates entries matched by <paths>
>> +     and sparse patterns.
>> +
>> +--widen::
>> +     Revert the effect of `--no-widen` if specified and make
>> +     `git checkout -- <paths>` update all entries matched by
>> +     <paths> regardless of sparse patterns.
>
> Perhaps, combining the descriptions of the positive and negative forms:
>
>         --respect-skip-worktree::
>                 By default, `git checkout -- <paths>` creates or updates files
>                 matching <paths> regardless of the skip-worktree bit.  This
>                 option makes 'git checkout' skips entries with the
>                 skip-worktree bit set, which can be useful in sparse checkout
>                 mode.

OK

> I'm afraid I can't imagine when --no-respect-skip-worktree would be
> useful.  That can easily be a failure of my imagination, though.

There may be scripts that expect "git checkout -- foo" to reset
everything in "foo". Or maybe you just want to check out a single file
and do not bother to edit sparse patterns as you won't need it for
long.

  parent reply	other threads:[~2013-04-01  5:14 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-22 10:28 Hard reset of a subdirectory in a sparse checkout setting Kirill Müller
2013-03-23  9:59 ` Duy Nguyen
2013-03-23 13:02   ` Kirill Müller
2013-03-24  5:06 ` [PATCH] checkout: add --sparse for restoring files in sparse checkout mode Nguyễn Thái Ngọc Duy
2013-03-24  6:12   ` Eric Sunshine
2013-03-24 18:17   ` Jonathan Nieder
2013-03-24 19:50     ` Kirill Müller
2013-03-25  1:34     ` Duy Nguyen
2013-03-25  1:57       ` Jonathan Nieder
2013-03-30 10:02   ` [PATCH v2] checkout: add --no-widen " Nguyễn Thái Ngọc Duy
2013-04-01  4:48     ` Jonathan Nieder
2013-04-01  5:12       ` Junio C Hamano
2013-04-01  5:13       ` Duy Nguyen [this message]
2013-04-01  6:18         ` Jonathan Nieder
2013-04-12 23:12     ` [PATCH v3] checkout: add --ignore-skip-worktree-bits " Nguyễn Thái Ngọc Duy
2013-05-14  9:27       ` Müller  Kirill
2013-05-14 10:21         ` Duy Nguyen

Reply instructions:

You may reply publicly 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

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

  git send-email \
    --in-reply-to=CACsJy8BVcu2VwxgQ9URViHMx2+6436npcYSHyZ5Mj0aVhVFtug@mail.gmail.com \
    --to=pclouds@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jrnieder@gmail.com \
    --cc=kirill.mueller@ivt.baug.ethz.ch \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).