Re: [PATCH] git-mv: Keep moved index entries inact
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Jul 19, 2008, 23:54 UTC
- Message-ID
- <7vej5pwhub.fsf@gitster.siamese.dyndns.org>
- In-Reply-To
- <20080717223036.20680.9672.stgit@localhost>
Petr Baudis <pasky@suse.cz> writes:
Show 5 quoted lines
> A new test has been added to the testsuite to reflect this change. > Also, based on suggestion by Junio about desired symlink behaviour > of git mv, I have added two tests for that; however, I do not have > need or desire to spend time fixing this, so they are expected > to fail for now until someone gets around to fixing that.
Well, somebody would eventually come to help, then ;-).
Show 5 quoted lines
> builtin-mv.c | 62 ++++++++------------------------------------------------- > cache.h | 2 ++ > read-cache.c | 15 ++++++++++++++ > t/t7001-mv.sh | 57 ++++++++++++++++++++++++++++++++++++++++++++++++++++ > 4 files changed, 83 insertions(+), 53 deletions(-)
Very nice code reduction, isn't it?
Show 22 quoted lines
> diff --git a/read-cache.c b/read-cache.c
> index 1648428..70e5f57 100644
> --- a/read-cache.c
> +++ b/read-cache.c
> @@ -38,6 +38,21 @@ static void replace_index_entry(struct index_state *istate, int nr, struct cache
> istate->cache_changed = 1;
> }
>
> +void rename_index_entry_at(struct index_state *istate, int nr, const char *new_name)
> +{
> + struct cache_entry *old = istate->cache[nr], *new;
> + int namelen = strlen(new_name);
> +
> + new = xmalloc(cache_entry_size(namelen));
> + copy_cache_entry(new, old);
> + new->ce_flags = (new->ce_flags & ~CE_NAMEMASK) | namelen;
> + memcpy(new->name, new_name, namelen);
> +
> + cache_tree_invalidate_path(istate->cache_tree, old->name);
> + remove_index_entry_at(istate, nr);
> + add_index_entry(istate, new, ADD_CACHE_OK_TO_ADD|ADD_CACHE_OK_TO_REPLACE);
> +}Hmm, would this use of copy_cache_entry() kosher, I have to wonder. This new copy of cache entry begins its life unhashed, doesn't it? Shouldn't we be not copying its hashed/unhashed bits from the old one?
Also setting of that ce_flags looks wrong when namelen does not fit within the width of CE_NAMEMASK. Shouldn't it be doing the same thing as create_ce_flags()?
Show 9 quoted lines
> diff --git a/t/t7001-mv.sh b/t/t7001-mv.sh > index 336cfaa..6b615f8 100755 > --- a/t/t7001-mv.sh > +++ b/t/t7001-mv.sh > @@ -156,4 +156,61 @@ test_expect_success 'absolute pathname outside should fail' '( > > )' > > +# git mv meets angry Git maintainer
What's this comment about?
Show 7 quoted lines
> +test_expect_success 'git mv should not change sha1 of moved cache entry' ' > + > + rm -fr .git && > + git init && > + echo 1 >dirty && > + git add dirty && > + entry="$(git ls-files --stage dirty | cut -f 1)"
"rev-parse :dirty"?