From: Jonathan Nieder <jrnieder@gmail.com>
To: "Nguyễn Thái Ngọc Duy" <pclouds@gmail.com>
Cc: git@vger.kernel.org,
Ilari Liusvaara <ilari.liusvaara@elisanet.fi>,
Jakub Narebski <jnareb@gmail.com>,
"Dmitry S. Kravtsov" <idkravitz@gmail.com>,
Shawn Pearce <spearce@spearce.org>
Subject: Re: [PATCH] cache-tree: do not cache empty trees
Date: Mon, 7 Feb 2011 03:17:40 -0600 [thread overview]
Message-ID: <20110207091740.GA5391@elie> (raw)
In-Reply-To: <1296914835-808-1-git-send-email-pclouds@gmail.com>
Nguyễn Thái Ngọc Duy wrote:
> Let's do it in a consistent way, always disregard empty trees in
> index. If users choose to create empty trees their own way, they
> should not use index at all.
While this violates some seeming invariants, like
1.
git reset --hard
git commit --allow-empty
git rev-parse HEAD^^{tree} >expect
git rev-parse HEAD^{tree} >actual
test_cmp expect actual
2.
git reset --hard
git revert HEAD
if git rev-parse HEAD~2
then
git rev-parse HEAD~2^{tree} >expect
git rev-parse HEAD^{tree} >actual
test_cmp expect actual
fi
, I think it's a good change. Malformed modes in trees already break
those false invariants iiuc.
Thanks.
> --- a/cache-tree.c
> +++ b/cache-tree.c
> @@ -621,9 +621,18 @@ static void prime_cache_tree_rec(struct cache_tree *it, struct tree *tree)
> struct tree *subtree = lookup_tree(entry.sha1);
> if (!subtree->object.parsed)
> parse_tree(subtree);
> + if (!hashcmp(entry.sha1, (unsigned char *)EMPTY_TREE_SHA1_BIN)) {
> + warning("empty tree detected! Will be removed in new commits");
> + cnt = -1;
> + break;
> + }
Aside from the warning, this part is an optimization, right?
> sub = cache_tree_sub(it, entry.path);
> sub->cache_tree = cache_tree();
> prime_cache_tree_rec(sub->cache_tree, subtree);
> + if (sub->cache_tree->entry_count == -1) {
> + cnt = -1;
> + break;
> + }
Would be nice to include a test for this, like so:
subdir/
empty1/
subsubdir/
empty2/
empty3/
> --- /dev/null
> +++ b/t/t1013-read-tree-empty.sh
> @@ -0,0 +1,20 @@
> +#!/bin/sh
> +
> +test_description='read-tree with empty trees'
> +
> +. ./test-lib.sh
> +
> +EMPTY_TREE=4b825dc642cb6eb9a060e54bf8d69288fbee4904
If we _have_ to hard-code this (why?) then I'd prefer to do so
in test-lib.sh.
[...]
> +test_expect_success 'write-tree removes empty tree' '
> + git read-tree `cat tree` &&
> + git write-tree >actual
> + echo $EMPTY_TREE >expected
> + test_cmp expected actual
Broken &&-chain. Not sure why this test relies on virtual objects
and another (independently nice) patch instead of adding the empty
tree to the object db --- is there something subtle I am missing?
The test does not distinguish between success due to git read-tree
omitting empty trees and success due to git mktree omitting empty
trees.
Hope that helps,
Jonathan
next prev parent reply other threads:[~2011-02-07 9:19 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-02-05 8:30 [PATCH] cache-tree: do not cache empty trees Nguyễn Thái Ngọc Duy
2011-02-05 9:50 ` Nguyễn Thái Ngọc Duy
2011-02-05 10:14 ` Jonathan Nieder
2011-02-05 10:32 ` Nguyen Thai Ngoc Duy
2011-02-05 14:07 ` Nguyễn Thái Ngọc Duy
2011-02-07 2:09 ` Junio C Hamano
2011-02-07 2:36 ` Nguyen Thai Ngoc Duy
2011-02-07 8:17 ` [PATCH] correct type of EMPTY_TREE_SHA1_BIN Jonathan Nieder
2011-02-09 23:33 ` Junio C Hamano
2011-02-07 9:17 ` Jonathan Nieder [this message]
2011-02-07 9:57 ` [PATCH] cache-tree: do not cache empty trees Nguyen Thai Ngoc Duy
2011-02-07 12:18 ` Ilari Liusvaara
2011-02-07 12:29 ` Nguyen Thai Ngoc Duy
2011-02-07 12:32 ` Jonathan Nieder
2011-02-07 20:48 ` Junio C Hamano
2011-02-08 4:11 ` Nguyen Thai Ngoc Duy
2011-02-08 4:30 ` Jonathan Nieder
2011-02-15 10:19 ` Yann Dirson
2011-02-16 14:29 ` Jakub Narebski
2011-02-08 10:40 ` Ilari Liusvaara
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
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=20110207091740.GA5391@elie \
--to=jrnieder@gmail.com \
--cc=git@vger.kernel.org \
--cc=idkravitz@gmail.com \
--cc=ilari.liusvaara@elisanet.fi \
--cc=jnareb@gmail.com \
--cc=pclouds@gmail.com \
--cc=spearce@spearce.org \
/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.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
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 read-only IMAP folder(s) and NNTP newsgroup(s).