git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Igor Djordjevic <igor.d.djordjevic@gmail.com>
To: Junio C Hamano <gitster@pobox.com>, Jake Stine <jake.stine@gmail.com>
Cc: git@vger.kernel.org
Subject: Re: [bug] git stash push {dir-pathspec} wipes untracked files
Date: Thu, 15 Mar 2018 21:39:25 +0100	[thread overview]
Message-ID: <780a67cd-7e94-f155-d04d-72799e2997fa@gmail.com> (raw)
In-Reply-To: <xmqq1sgl703v.fsf@gitster-ct.c.googlers.com>

On 15/03/2018 17:52, Junio C Hamano wrote:
> 
> > Hi, I ran into what I believe is a bug today.  I’m using primarily Git
> > for Windows 2.16.2 and also reproduced the behavior on Git for Windows
> > 2.15.1 and Git 2.14.1 on Ubuntu:
> >
> > Given any repository with at least one subdirectory:
> >
> > 1.       Create some untracked files in the subdir
> > 2.       Modify a tracked file in the subdir
> > 3.       Execute `git stash push subdir`
> > 4.       The untracked files will be removed, without warning.
> >
> > `git stash push` behaves as-expcted and does not touch untracked
> > files.  It’s only when a directory tree is specified as [pathspec]
> > that the problem occurs.
> 
> I wonder if this is the same as the topic on this thread.
> 
>   https://public-inbox.org/git/CA+HNv10i7AvWXjrQjxxy1LNJTmhr7LE4TwxhHUYBiWtmJCOf_A@mail.gmail.com/
> 
> What is curious is that the fix bba067d2 ("stash: don't delete
> untracked files that match pathspec", 2018-01-06) appeared first in
> 2.16.2, on which Windows 2.16.2 is supposed to be built upon.
> 
> > Here's the precise reproduction case executed on a linux box:
> 
> This does not reproduce for me with v2.16.2-17-g38e79b1fda (the tip
> of 'maint'); I do not have an  install of vanilla v2.16.2 handy, but
> I suspect v2.16.2 would work just fine, too.
> 
> > jake@jake-VirtualBox:~/woot$ git --version
> > git version 2.14.1
> > ...
> > The expected result is that when I do `ls subdir` the file
> > "untracked.txt" still exists.  Alternatively, git stash should warn me
> > before destroying my untracked files, and require I specify --force or
> > similar to invoke destructive behavior.

I can't seem to reproduce this on 2.16.2.windows.1, either:

+ git --version
git version 2.16.2.windows.1
+ git init woot
Initialized empty Git repository in /woot/.git/
+ cd woot
+ mkdir subdir
+ echo test
+ echo test
+ git add meh.txt subdir/meh2.txt
+ git commit '--message=stash bug testing'
[master (root-commit) afec47d] stash bug testing
 2 files changed, 2 insertions(+)
 create mode 100644 meh.txt
 create mode 100644 subdir/meh2.txt
+ git commit '--message=stash bug testing'
On branch master
nothing to commit, working tree clean
+ echo test
+ echo append
+ git stash push subdir
Saved working directory and index state WIP on master: afec47d stash bug testing
+ ls subdir
meh2.txt  untracked.txt

Regards, Buga

      parent reply	other threads:[~2018-03-15 20:39 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-15 16:40 [bug] git stash push {dir-pathspec} wipes untracked files Jake Stine
2018-03-15 16:52 ` Junio C Hamano
2018-03-15 17:02   ` Randall S. Becker
2018-03-15 20:39   ` Igor Djordjevic [this message]

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=780a67cd-7e94-f155-d04d-72799e2997fa@gmail.com \
    --to=igor.d.djordjevic@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jake.stine@gmail.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
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).