Since commit f269048 (fetch: opportunistically update tracking refs,
2013-05-11) we update tracking refs opportunistically when fetching
remote branches. However, if there is a configured non-pattern refspec
that does not match any of the refspecs given on the command line then a
fatal error occurs.
Fix this by setting the "missing_ok" flag when calling get_fetch_map.
Test-added-by: Jeff King <peff@peff.net>
Signed-off-by: John Keeping <john@keeping.me.uk>
Acked-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In order to make sure the cloned repository is good, we run "rev-list
--objects --not --all $new_refs" on the repository. This is expensive
on large repositories. This patch attempts to mitigate the impact in
this special case.
In the "good" clone case, we only have one pack. If all of the
following are met, we can be sure that all objects reachable from the
new refs exist, which is the intention of running "rev-list ...":
- all refs point to an object in the pack
- there are no dangling pointers in any object in the pack
- no objects in the pack point to objects outside the pack
The second and third checks can be done with the help of index-pack as
a slight variation of --strict check (which introduces a new condition
for the shortcut: pack transfer must be used and the number of objects
large enough to call index-pack). The first is checked in
check_everything_connected after we get an "ok" from index-pack.
"index-pack + new checks" is still faster than the current "index-pack
+ rev-list", which is the whole point of this patch. If any of the
conditions fail, we fall back to the good old but expensive "rev-list
..". In that case it's even more expensive because we have to pay for
the new checks in index-pack. But that should only happen when the
other side is either buggy or malicious.
Cloning linux-2.6 over file://
before after
real 3m25.693s 2m53.050s
user 5m2.037s 4m42.396s
sys 0m13.750s 0m16.574s
A more realistic test with ssh:// over wireless
before after
real 11m26.629s 10m4.213s
user 5m43.196s 5m19.444s
sys 0m35.812s 0m37.630s
This shortcut is not applied to shallow clones, partly because shallow
clones should have no more objects than a usual fetch and the cost of
rev-list is acceptable, partly to avoid dealing with corner cases when
grafting is involved.
This shortcut does not apply to unpack-objects code path either
because the number of objects must be small in order to trigger that
code path.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
index-pack --strict looks up and follows parent commits. If shallow
information is not ready by the time index-pack is run, index-pack may
be led to non-existent objects. Make fetch-pack save shallow file to
disk before invoking index-pack.
git learns new global option --shallow-file to pass on the alternate
shallow file path. Undocumented (and not even support --shallow-file=
syntax) because it's unlikely to be used again elsewhere.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
And get rid of the remote-hg.force-push option hack.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This needs a specific patch from Git not applied yet.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It's better to catch the exception later on.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Always convert to strings (they are unicode because they come from JSON).
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Otherwise replacing a file with a directory doesn't work.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We want to show the remote heads, not the internal ones, which might
have garbage.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It simply picks the last head that is not closed, but we have a stored
list of open heads.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
So that when a diverge is detected, we know which ref to report an error
for.
Also, since we are not throwing an exception, return a proper error
code.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The version from Mercurial is extremely inefficient and convoluted, this
version achieves basically the same, at least for our purposes.
No functional changes.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The one from mercurial does a ton of things we are not interested in,
and we need some special modifications which are impossible otherwise.
Most of the code is borrowed from Mercurial, and cleaned up, but should
be functionally the same for our purposes, except that multiple heads
are not detected.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We don't care about the rest, and in fact, we shouldn't try to push
everything, as there might be garbage from previous failed pushes.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If we update the 'old' node, we might be updating the remote bookmark
even when our 'new' node is not related at all to what the remote has,
effectively forcing an update.
Let's do that only when forced push is configured.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We don't need to update both internal and remote bookmarks, so let's do
one or the other, and move the shared code earlier, so it's simpler.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
With lots branches and bookmarks, non-ff, updated and new.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
From the point of view of Mercurial, this creates a new branch head,
and requires a forced push.
Ideally, however, we would want it to work just like in git; new
branches can be pushed without problems.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Neither mercurial nor git allows pushing to a remote when it's a
non-fast-forward push. We should be able to detect these errors and
report them properly, as opposed to throwing an exception
stack-trace.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The logic when working with a local repository is totally different from
the one where we work with a remote repository; we need to pull and push
from it.
All this logic is currently not tested at all, so let's introduce a
variable to force the remote behavior.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We want to make sure everything works correctly, even if there's a
'master' bookmark in Mercurial.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Mercurial always checks out the 'default' branch, so there's no point in
complicating our lives trying to do something fancier, which causes
different behavior depending on whether the repository is local or
remote.
So let's always use 'default' (which we translate to 'master'), unless
we are in hg-git mode, which expects us to use the 'master' bookmark
instead.
Also, update the tests that used to check for different checkout
behaviors to simply check that the refs are there, remove unnecessary
ones, and fix the ones that expect something different.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
No need to manually keep track of the revision count.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As suggested by Jed Brown; there's no need to re-import all the commits.
Cc: Jed Brown <jed@59a2.org>
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Otherwise we won't know if revisions are replaced.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We won't be able to count the unmarked commits, but we are not going to
be able to do that anyway when we switch to SHA-1 ids.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This way we don't have to have duplicated Mercurial objects.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The user might have then configured differently, plus, all of them will
be loaded anyway later on.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
So that we don't need a temporary directory.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
So it's more standardized between all the tests.
Signed-off-by: Felipe Contreras <felipe.contreras@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
