Patrick Mezard <patrick@mezard.eu> [Sat, 12 May 2012 00:24:07 +0200] rev 16658
phases: make advance/retractboundary() atomic
Before this, if advanceboundary() failed after updating some roots but
before calling retractboundary(), the phase cache would be left in an
invalid state, marked dirty, and written as such. This patch approach is
to turn advance/retractboundary() into phasecache methods, then operate
on copies and merge them back on success.
With the same technique, we can ensure the atomicity of combinations of
advance/retractboundary() calls, like those performed in changegroup
handling code.
Patrick Mezard <patrick@mezard.eu> [Sat, 12 May 2012 00:24:07 +0200] rev 16657
phases: introduce phasecache
The original motivation was changectx.phase() had special logic to
correctly lookup in repo._phaserev, including invalidating it when
necessary. And at other places, repo._phaserev was accessed directly.
This led to the discovery that phases state including _phaseroots,
_phaserev and _dirtyphase was manipulated in localrepository.py,
phases.py, repair.py, etc. phasecache helps encapsulating that.
This patch replaces all phase state in localrepo with phasecache and
adjust related code except for advance/retractboundary() in phases.
These still access to phasecache internals directly. This will be
addressed in a followup.
Patrick Mezard <patrick@mezard.eu> [Sat, 12 May 2012 00:19:30 +0200] rev 16656
mq: introduce mq.check setting
When:
[mq]
check = True
is set, qpush, qpop and qgoto behave as if -c/--check were passed. If
incompatible options like -f/--force or --exact are set, this setting is
ignored.
This setting enables what many users expect mq default behaviour to be.
Patrick Mezard <patrick@mezard.eu> [Sat, 12 May 2012 00:19:30 +0200] rev 16655
mq: introduce qgoto --check
Patrick Mezard <patrick@mezard.eu> [Sat, 12 May 2012 00:19:30 +0200] rev 16654
mq: introduce qpush --check
qpush --check let you qpush with uncommitted files not overlapping
patched files.
Patrick Mezard <patrick@mezard.eu> [Sat, 12 May 2012 00:19:30 +0200] rev 16653
mq: introduce qpop --check
qpop --check let you qpop with uncommitted files if they do not
intersect with files touched by the popped patches.
David Soria Parra <dsp@php.net> [Sat, 12 May 2012 09:43:12 +0200] rev 16652
pager: check if signal.SIGPIPE exists
We have to check for signal.SIGPIPE before we attempt
to set it.
Matt Mackall <mpm@selenic.com> [Sat, 12 May 2012 10:02:47 +0200] rev 16651
diffhelpers: harden testhunk
Yuya Nishihara <yuya@tcha.org> [Sat, 12 May 2012 16:10:01 +0900] rev 16650
patch: fix segfault against unified diffs which start line is zero
Since
2b1ec74c961f, if a chunk starts with "@@ -0,1", oldstart turns into
a negative value. Because diffhelpers.testhunk() doesn't expect negative bstart,
it bypasses "alen > blen - bstart" condition and segfaults at
"PyList_GET_ITEM(b, i + bstart)".
Steven Stallion <sstallion@gmail.com> [Fri, 11 May 2012 22:48:19 -0700] rev 16649
bugzilla: fix transport initialization on python 2.4
Bryan O'Sullivan <bryano@fb.com> [Tue, 08 May 2012 15:31:00 -0700] rev 16648
bisect: set HG_NODE when runing a command
When running a command, set the environment variable HG_NODE to
tell the command which changeset is being visited.
Bryan O'Sullivan <bryano@fb.com> [Tue, 08 May 2012 15:29:09 -0700] rev 16647
bisect: track the current changeset (
issue3382)
Introduce a new revset feature, bisect(current), that identifies
the changeset currently being bisected.
Jesse Glick <jesse.glick@oracle.com> [Fri, 04 May 2012 15:56:45 -0400] rev 16646
localrepo: optimize internode status calls using withflags
Introduce manifestdict.withflags() to get a set of all files which have any
flags set, since these are likely to be a minority. Otherwise checking .flags()
for every file is a lot of dictionary lookups and is quite slow.
Jesse Glick <jesse.glick@oracle.com> [Fri, 04 May 2012 15:54:55 -0400] rev 16645
localrepo: optimize internode status calls using match.always
Introduce match.always() to check if a match object always says yes, i.e.
None was passed in. If so, mfmatches should not bother iterating every file in
the repository.
Na'Tosha Bard <natosha@unity3d.com> [Fri, 11 May 2012 23:11:43 +0200] rev 16644
largefiles: add --all-largefiles flag to clone (
issue3188)
Augie Fackler <raf@durin42.com> [Fri, 04 May 2012 16:00:33 -0500] rev 16643
httpclient: update to revision
892730fe7f46 of httpplus
Matt Mackall <mpm@selenic.com> [Sat, 12 May 2012 00:06:11 +0200] rev 16642
merge with stable
Bryan O'Sullivan <bryano@fb.com> [Tue, 08 May 2012 14:46:04 -0700] rev 16641
parsers: use the correct maximum radix tree depth
Previously, we would not use more than half of a SHA-1 hash when
constructing and searching the tree.
Patrick Mezard <patrick@mezard.eu> [Wed, 09 May 2012 18:45:14 +0200] rev 16640
revset: make matching() preserve input revision order
Jesse Glick <jesse.glick@oracle.com> [Thu, 10 May 2012 14:17:05 -0400] rev 16639
revset: documentation typo "metatadata"
Na'Tosha Bard <natosha@unity3d.com> [Fri, 11 May 2012 15:32:22 +0200] rev 16638
largefiles: fix deletion of multiple missing largefiles (
issue3329)
Levi Bard <levi@unity3d.com> [Fri, 11 May 2012 15:15:50 +0200] rev 16637
addremove: document default similarity behavior (
issue3429)
Na'Tosha Bard <natosha@unity3d.com> [Fri, 11 May 2012 14:42:26 +0200] rev 16636
largefiles: follow normal codepath for addremove if non-largefiles repo (
issue3249)
Patrick Mezard <patrick@mezard.eu> [Fri, 11 May 2012 16:57:26 +0200] rev 16635
mq: add --no-backup for qpush/qpop/qgoto