Re: [PATCH v5 15/15] fast-export: don't handle uninteresting refs

On Sun, Nov 11, 2012 at 2:59 PM, Felipe Contreras
[off-list ref] wrote:

They have been marked as UNINTERESTING for a reason, lets respect that.

Currently the first ref is handled properly, but not the rest, so:

 % git fast-export master ^master

Would currently throw a reset for master (2nd ref), which is not what we
want.

 % git fast-export master ^foo ^bar ^roo
 % git fast-export master salsa..tacos

Even if all these refs point to the same object; foo, bar, roo, salsa,
and tacos would all get a reset, and to a non-existing object (invalid
mark :0).

And even more, it would only happen if the ref is pointing to exactly
the same commit, but not otherwise:

 % git fast-export ^next next
 reset refs/heads/next
 from :0

 % git fast-export ^next next^{commit}
 # nothing
 % git fast-export ^next next~0
 # nothing
 % git fast-export ^next next~1
 # nothing
 % git fast-export ^next next~2
 # nothing

The reason this happens is that before traversing the commits,
fast-export checks if any of the refs point to the same object, and any
duplicated ref gets added to a list in order to issue 'reset' commands
after the traversing. Unfortunately, it's not even checking if the
commit is flagged as UNINTERESTING. The fix of course, is to do
precisely that.

The current behavior is most certainly not what we want. After this
patch, nothing gets exported, because nothing was selected (everything
is UNINTERESTING).

Signed-off-by: Felipe Contreras <redacted>

And here's yet another reason why this is obviously correct that I just found:

% git fast-export --use-done-feature
--{import,export}-marks=.git/hg/origin/marks-git
^refs/hg/origin/branches/default ^refs/hg/origin/bookmarks/test6
refs/heads/test6 ^refs/hg/origin/bookmarks/master
^refs/hg/origin/bookmarks/test
feature done
reset refs/hg/origin/bookmarks/test
from :4

reset refs/heads/test6
from :14

done

What is that refs/hg/origin/bookmarks/test doing there?

transport-helper does use a fast-export command like that to specify
precisely what refs should be *IGNORED*, and yet fast-export will
throw a reset for a ref that has been marked as UNINTERESTING. So, the
receiving end in the helper will see a reset for a ref that it
explicitly said was marked as outside it's refspec realm:

refspec refs/heads/*:refs/hg/origin/bookmarks/*

What is remote-hg supposed to do with 'refs/hg/origin/bookmarks/test'?
There's nothing that can be done, it's a bug in fast-export that such
a thing was exported in the first place. And the reason it happens is
that another ref happens to be pointing to the same object, (in this
case refs/hg/origin/branches/default)

So yeah, the patch is good.

Of course, transport-helper shouldn't even be specifying the negative
(^) refs, but that's another story.

Cheers.

-- 
Felipe Contreras