Malformed branch name in fast-export when specifying non-HEAD/branch revision

Hi,

fast-export gives an invalid branch name for commits, if the revision specified is not HEAD or a branch name.

$ git --version git version 1.7.6

$ # Create $ git init export_test && cd export_test $ touch a && git add a && git commit -m 'Add a' $ touch b && git add b && git commit -m 'Add b' $ git branch branch1

$ # ok $ git fast-export HEAD

$ # also ok $ git fast-export branch1

$ # uses HEAD~1 instead of refs/heads/master $ git fast-export HEAD~1

blob mark :1 data 0

reset HEAD~1 commit HEAD~1 mark :2 author Owen Stephens <git@owenstephens.co.uk> 1313597436 +0100 committer Owen Stephens <git@owenstephens.co.uk> 1313597436 +0100 data 6 Add a M 100644 :1 a

$ # Uses the commit hash rather than refs/heads/master $ git fast-export $(git rev-parse HEAD~1)

blob mark :1 data 0

reset 5a8a8abb1ab44890501c64f2d51f671ab2108d60 commit 5a8a8abb1ab44890501c64f2d51f671ab2108d60 mark :2 author Owen Stephens <git@owenstephens.co.uk> 1313597436 +0100 committer Owen Stephens <git@owenstephens.co.uk> 1313597436 +0100 data 6 Add a M 100644 :1 a

$ # fast-import cannot handle this stream: $ git init subgit $ git fast-export HEAD~1 | (cd subgit/ && git fast-import) fatal: Branch name doesn't conform to GIT standards: HEAD~1 fast-import: dumping crash report to .git/fast_import_crash_26863

Regards, Owen.

Hi,

On Wed, Aug 17, 2011 at 10:21 AM, Owen Stephens <git@owenstephens.co.uk> wrote:

Thanks for the report. It turns out this bug has been reported and is in the testsuite as t9350.19 -- currently marked as expected to fail. I looked at the problem a couple years ago for a little bit but never finished that particular patch and never got back around to it.

Elijah Newren <newren@gmail.com> writes:

What _should_ be the right behaviour to begin with, I have to wonder.

Even though it is very clear that the set of objects that are exported are defined by the "rev-list arguments" given to the command, I do not think fast-export's semantics is not clearly defined as to what "refs" are to be updated.

The easiest fix for this issue would be to forbid "git fast-export HEAD~1" (or any range whose positive endpoints are _not_ refs), and I think that would be in line with the original motivation of the command to export the whole repository in a format fast-import would understand. The original f2dc849 (Add 'git fast-export', the sister of 'git fast-import', 2007-12-02) says "This program dumps (parts of) a git repository...", implying that partial export is within the scope of the command, but I do not think it was designed carefully enough to deal with ranges more complex than just "a set of branches".

I however have a feeling that people would want to say:

 - I want to export up to that commit, and have that commit on this branch
   on the importing side; or even better

 - I want to export up to that commit, but what refs points at the commits
   contained in the output stream will be decided when the output is
   imported.

I do not think the latter meshes well with how "fast-import" works, though. But fast-export should be fixable to allow the former without breaking the semantics of fast-import.

You can think of "fast-export" an off-line "push" command [*1*]; instead of giving a random commit object, e.g. "git fast-export HEAD~1", that can not be used as a ref, you can use the refspec notation to tell where the result should go, e.g. "git fast-export HEAD~1:refs/heads/a-bit-older", from the command line of fast-export.

I suspect that also may clarify what Sverre was trying to do in his recent series. The root cause of both this and the issue Sverre wanted to fix is the design mistake of fast-export that tries to reuse the notation of object range specification for a different purpose of telling which "ref" to update, I think.

[Footnote]

*1* In a similar sense, unpacking "git bundle" output is an off-line "fetch"; the bundle creator gave anchor points for tip objects, and allows the unpacker to map them into its own namespace.

On Wed, Aug 17, 2011 at 03:30:14PM -0700, Junio C Hamano wrote:

Yes, this was the conclusion I came to when I looked at this a month or so ago. You really need to give fast-export a mapping of objects to refnames, and it should output ref names _only_ for the mapping. That would handle this "not a ref" case, but would also let you push "refs/heads/foo" when it is equivalent to "refs/heads/master", without fast-export mentioning "refs/heads/master" at all.

-Peff

Heya,

On Wed, Aug 17, 2011 at 16:19, Jeff King <peff@peff.net> wrote:

Does this bring any new insights into how the problem I was pointing out (trying to push next if master points at the same commit does nothing) could/should be solved?

On Sun, Aug 21, 2011 at 03:29:38PM -0700, Sverre Rabbelier wrote:

Hmm. Maybe I am misremembering the problem, but I thought that worked already. If you say:

  git fast-export refs/heads/foo

you should get only reset/commit lines in the output for refs/heads/foo, no?

Now I can't seem to replicate the case where refs/heads/master is mentioned, but you didn't want it to be. I may have to go back and re-read the thread from a month or two ago when we discussed these issues.

-Peff

Heya,

On Mon, Aug 22, 2011 at 09:19, Jeff King <peff@peff.net> wrote:

Do you agree that this is expected behavior?

$ git init test
Initialized empty Git repository in /home/sverre/code/test/.git/
$ cd test/
$ echo content >> foo
sverre@laptop-sverre:~/code/test
$ git add foo
$ git commit -m first
[master (root-commit) 821176f] first
 1 files changed, 1 insertions(+), 0 deletions(-)
 create mode 100644 foo
$ echo content >> foo
$ git commit -am second
[master 1934282] second
 1 files changed, 1 insertions(+), 0 deletions(-)
$ git branch other
$ git fast-export ^master other
reset refs/heads/other
from 1934282469e3a83a5ef827fd31e074cfb4f3eadf

Because in current git.git, this doesn't work (the above is generated using a git that has the patch series Dscho and I sent out). Current git will instead do the following:

$ git fast-export ^master other reset refs/heads/other from :0

The 'from :0' here is obviously a bug (which is fixed by our patch series).

You might wonder, 'why would anyone do that', well, for example, they might be using marks:

$ git fast-export --export-marks=marksfile master > /dev/null $ git fast-export --import-marks=marksfile other reset refs/heads/other from :4

Again, the above is generated with my patched git, current git.git simply outputs nothing.

$ git fast-export --import-marks=marksfile other

On Mon, Aug 22, 2011 at 09:54:47AM -0700, Sverre Rabbelier wrote:

Yeah, that seems reasonable to me.

Yep, the current behavior is definitely wrong. But I thought your question was about accidentally mentioning refs/heads/master, which this doesn't do (nor should it).

I just read through the remote-helper threads from early June, and the only mention of triggering that is when you actually have a rename (i.e., your "refs/heads/foo" becomes remote's "refs/heads/bar", but we mention "refs/heads/foo" in the export stream). I was thinking there was another case, but I couldn't find mention of it.

Yeah, the behavior of your patch looks fine to me. I thought the point in contention was that having export understand refspecs would fix a lot of _other_ cases, too.

-Peff

Heya,

On Mon, Aug 22, 2011 at 10:57, Jeff King <peff@peff.net> wrote:

The patches Dscho and I sent are in response to the RFC patches (that are currently "stalled" in whats-cooking) I added on top of the series that rerolled yours.

Right. Sadly it doesn't look like I'll have time to try and fix 'git bundle' anytime soon, so this'll remain broken.

Jeff King <peff@peff.net> writes:

Perhaps if we added refspecs we could cover other cases, too, but I do not think we need to require the patch to cover additional cases. The more problematic was that the particular implementation in the patch smelled bad.

On Mon, Aug 22, 2011 at 02:27:00PM -0700, Junio C Hamano wrote:

OK. I didn't look at that at all. I just wanted to express my support for "refspecs would solve other problems, too". :)

-Peff