Git development
 help / color / mirror / Atom feed
From: Jonathan Nieder <jrnieder@gmail.com>
To: "Jean-Philippe Gariépy" <jean-philippe.gariepy@nuecho.com>
Cc: git@vger.kernel.org, Shawn Bohrer <shawn.bohrer@gmail.com>
Subject: Re: [BUG] git clean -X behaviour when .gitignore has sub-directory entries
Date: Mon, 27 Sep 2010 15:36:52 -0500	[thread overview]
Message-ID: <20100927203652.GB11957@burratino> (raw)
In-Reply-To: <4CA0AB9A.4050002@nuecho.com>

Hi,

Jean-Philippe Gariépy wrote:

> When using sub-directory entries in .gitignore, while the entry is
> ignored as expected, "git clean -X" doesn't clean the ignored
> sub-directory.

Thanks for reporting.

 $ cat test.sh
 rm -fr test &&
 git init test &&
 (
	cd test &&
	mkdir -p a/b/c &&
	>a/b/c/f &&
	echo '/a/b/' >.gitignore &&
	git status -s &&
	git clean -X -d -n &&
	git ls-files -o -i --exclude-standard &&
	git ls-files -o --directory -i --exclude-standard &&
	git clean -X -d -f &&
	echo ... drumroll ... &&
	! test -e a/b/c/f
 )
 $ sh test.sh || echo fail
 Initialized empty Git repository in /tmp/test/.git/
 ?? .gitignore
 ... drumroll ...
 fail

Variations:

 1) echo '/a/' >.gitignore
 2) echo '/a/b/c/f' >.gitignore
 3) >a/b/f

(1) does not fail; (2) and (3) still do.

Okay, so why does this happen?  Tracing:

-- 8< --
diff --git a/dir.c b/dir.c
index 133f472..5707ad0 100644
--- a/dir.c
+++ b/dir.c
@@ -693,6 +693,9 @@ static enum path_treatment treat_one_path(struct dir_struct *dir,
 					  int dtype, struct dirent *de)
 {
 	int exclude = excluded(dir, path, &dtype);
+
+	trace_printf("treat_one_path: path=%s exclude=%d\n", path, exclude);
+
 	if (exclude && (dir->flags & DIR_COLLECT_IGNORED)
 	    && exclude_matches_pathspec(path, *len, simplify))
 		dir_add_ignored(dir, path, *len);
-- >8 --

 $ GIT_TRACE=true PATH=/home/jrn/src/git/bin-wrappers:$PATH sh test.sh 
 trace: built-in: git 'init' 'test'
 Initialized empty Git repository in /tmp/test/.git/
 trace: built-in: git 'status' '-s'
 treat_one_path: path=.gitignore exclude=0
 treat_one_path: path=a exclude=0
 treat_one_path: path=a/b exclude=1
 ?? .gitignore
 trace: built-in: git 'clean' '-X' '-d' '-n'
 treat_one_path: path=.gitignore exclude=0
 treat_one_path: path=a exclude=0
 trace: built-in: git 'ls-files' '-o' '-i' '--exclude-standard'
 treat_one_path: path=.gitignore exclude=0
 treat_one_path: path=a exclude=0
 treat_one_path: path=a/b exclude=1
 treat_one_path: path=a/b/c exclude=0
 treat_one_path: path=a/b/c/f exclude=0
 trace: built-in: git 'ls-files' '-o' '--directory' '-i' '--exclude-standard'
 treat_one_path: path=.gitignore exclude=0
 treat_one_path: path=a exclude=0
 trace: built-in: git 'clean' '-X' '-d' '-f'
 treat_one_path: path=.gitignore exclude=0
 treat_one_path: path=a exclude=0
 ... drumroll ...
 $

 1) a/b/f is not actually considered excluded; only its containing
    directory is.

 2) git clean does not even examine a/b to consider whether to remove
    it: since a/ does not contain any tracked files, it stopped there.

The following would cause "git clean -ndx" to print more than it
should, but hopefully it illustrates the idea.

diff --git a/builtin/clean.c b/builtin/clean.c
index b508d2c..91624c2 100644
--- a/builtin/clean.c
+++ b/builtin/clean.c
@@ -84,8 +84,6 @@ int cmd_clean(int argc, const char **argv, const char *prefix)
 	if (force > 1)
 		rm_flags = 0;
 
-	dir.flags |= DIR_SHOW_OTHER_DIRECTORIES;
-
 	if (read_cache() < 0)
 		die("index file corrupt");
 
-- 

  reply	other threads:[~2010-09-27 20:40 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-09-27 14:35 [BUG] git clean -X behaviour when .gitignore has sub-directory entries Jean-Philippe Gariépy
2010-09-27 20:36 ` Jonathan Nieder [this message]
2010-10-03 18:55   ` Jean-Philippe Gariépy
2010-10-03 19:12     ` Jonathan Nieder

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=20100927203652.GB11957@burratino \
    --to=jrnieder@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=jean-philippe.gariepy@nuecho.com \
    --cc=shawn.bohrer@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