Vadim Gelfer <vadim.gelfer@gmail.com> [Sat, 12 Aug 2006 12:30:02 -0700] rev 2858
update copyrights.
Vadim Gelfer <vadim.gelfer@gmail.com> [Sat, 12 Aug 2006 12:12:23 -0700] rev 2857
merge.
Alexis S. L. Carvalho <alexis@cecm.usp.br> [Sat, 12 Aug 2006 08:53:23 -0300] rev 2856
Fix some bugs introduced during the manifest refactoring
Alexis S. L. Carvalho <alexis@cecm.usp.br> [Sat, 12 Aug 2006 09:24:04 -0300] rev 2855
Fix "hg qnew -f foo" without -m
The refresh method assumes it has a string.
Vadim Gelfer <vadim.gelfer@gmail.com> [Sat, 12 Aug 2006 11:53:12 -0700] rev 2854
merge with mpm.
Matt Mackall <mpm@selenic.com> [Fri, 11 Aug 2006 11:00:38 -0500] rev 2853
Remove manifest.readflags
Matt Mackall <mpm@selenic.com> [Fri, 11 Aug 2006 10:57:42 -0500] rev 2852
Change remaining users of manifest flags
Matt Mackall <mpm@selenic.com> [Thu, 10 Aug 2006 22:38:56 -0500] rev 2851
Switch to simpler manifestdict
Matt Mackall <mpm@selenic.com> [Wed, 09 Aug 2006 22:07:13 -0500] rev 2850
Merge: use single objects for tracking manifests
Matt Mackall <mpm@selenic.com> [Wed, 09 Aug 2006 17:11:01 -0500] rev 2849
Merge: save away mode bit so that we don't need manifest later
Matt Mackall <mpm@selenic.com> [Wed, 09 Aug 2006 15:03:46 -0500] rev 2848
Merge manifest refactor work
Matt Mackall <mpm@selenic.com> [Wed, 09 Aug 2006 14:53:03 -0500] rev 2847
Combine manifest dict and flags dict into a single object
This gets rid of the need to track two objects and might save memory.
This might be faster implemented as a subclassed dict with auxiliary
functions to access a sparse flags dict.
Matt Mackall <mpm@selenic.com> [Wed, 19 Jul 2006 19:11:50 -0500] rev 2846
manifestflags: eliminate remaining users of direct dict access
Matt Mackall <mpm@selenic.com> [Sun, 16 Jul 2006 03:22:05 -0500] rev 2845
Use strings for manifestflags to allow storing multiple flags
Matt Mackall <mpm@selenic.com> [Sun, 16 Jul 2006 03:14:17 -0500] rev 2844
Start using manifestflags methods
Matt Mackall <mpm@selenic.com> [Sun, 16 Jul 2006 02:59:20 -0500] rev 2843
Add manifestflags class
Vadim Gelfer <vadim.gelfer@gmail.com> [Fri, 11 Aug 2006 14:50:41 -0700] rev 2842
import: get lock and wlock one time
this makes import of many patches 6x faster in kernel repo.
Thomas Arendsen Hein <thomas@intevation.de> [Fri, 11 Aug 2006 09:11:54 +0200] rev 2841
Adjusted documentation for hgrc's "[smtp] host" for change in 9f745d3675d4.
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 17:37:57 -0700] rev 2840
fix test output
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 17:33:08 -0700] rev 2839
mq: only add mq attribute to local repo
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 17:18:55 -0700] rev 2838
mq: apply patch is any posative guard matches
this is like guards command from quilt package.
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 17:02:11 -0700] rev 2837
fetch: fix test fail
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 16:10:59 -0700] rev 2836
mq: do not allow to push from repo with patches applied
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 15:44:05 -0700] rev 2835
call reposetup functions of extension modules whenever repo created
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 15:02:18 -0700] rev 2834
mq: test commit in repo with patches applied
Brendan Cully <brendan@kublai.com> [Thu, 10 Aug 2006 11:18:48 -0700] rev 2833
Disallow commit over an applied mq patch.
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 10 Aug 2006 14:58:10 -0700] rev 2832
qselect: add --pop, --reapply options
Danek Duvall <danek.duvall@sun.com> [Wed, 09 Aug 2006 12:05:40 -0700] rev 2831
Fix some tests for portability.
Matt Mackall <mpm@selenic.com> [Wed, 09 Aug 2006 13:55:18 -0500] rev 2830
Merge with crew, fix most tests
Matt Mackall <mpm@selenic.com> [Tue, 08 Aug 2006 10:56:48 -0500] rev 2829
Merge: combine force and forcemerge arguments
Matt Mackall <mpm@selenic.com> [Tue, 08 Aug 2006 10:35:19 -0500] rev 2828
Merge: move most tests to the beginning
Matt Mackall <mpm@selenic.com> [Tue, 08 Aug 2006 09:26:00 -0500] rev 2827
Merge: refactor err and failedmerge -> unresolved
Matt Mackall <mpm@selenic.com> [Tue, 08 Aug 2006 09:19:20 -0500] rev 2826
Refactor update locking slightly
Matt Mackall <mpm@selenic.com> [Tue, 08 Aug 2006 09:11:16 -0500] rev 2825
Merge: combine choose and moddirstate to partial
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 23:53:04 -0500] rev 2824
Rename merge.allow -> merge.branchmerge
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 23:18:16 -0500] rev 2823
Use demandload @ syntax
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 22:54:33 -0500] rev 2822
Introduce update helper functions: update, merge, clean, and revert
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 22:35:36 -0500] rev 2821
Teach demandload how to import as
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 20:54:03 -0500] rev 2820
Factor doupdate into _lookup + hg.update
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 17:26:09 -0500] rev 2819
Fix extensions for merge change
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 09 Aug 2006 09:38:11 -0700] rev 2818
mq: make guards more strict, add tests
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 17:11:36 -0700] rev 2817
mq: drop qversion command
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 17:08:59 -0700] rev 2816
fetch: hold lock and wlock across all operations
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 09 Aug 2006 02:04:49 -0700] rev 2815
tests: forgot to add test-mq-guards.out
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 16:40:08 -0700] rev 2814
fetch: lock repo across pull and commit
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 16:37:41 -0700] rev 2813
fetch: drop commands.doupdate
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 16:23:02 -0700] rev 2812
test-fetch: make output stable
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 16:09:26 -0700] rev 2811
fetch: fix breakage from mpm.
add test so will not break again.
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 21:42:50 -0700] rev 2810
mq: new commands qselect, qguard
implement quilt-style guards for mq.
guards allow to control whether patch can be pushed.
if guard X is active and patch is guarded by +X (called "posative guard"),
patch can be pushed. if patch is guarded by -X (called "nagative guard"),
patch cannot be pushed and is skipped.
use qguard to set/list guards on patches. use qselect to set/list
active guards.
also "qseries -v" prints guarded patches with "G" now.
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 18:14:03 -0700] rev 2809
fix permission of test-mq-qsave
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 08 Aug 2006 18:12:48 -0700] rev 2808
mq: add join method
Brendan Cully <brendan@kublai.com> [Mon, 07 Aug 2006 14:59:32 -0700] rev 2807
Make mq camelcase consistent with the rest of hg.
Brendan Cully <brendan@kublai.com> [Mon, 07 Aug 2006 14:48:24 -0700] rev 2806
Add basic qsave/qrestore test.
Brendan Cully <brendan@kublai.com> [Mon, 07 Aug 2006 14:48:18 -0700] rev 2805
Update qsave to use StatusEntry; don't throw exception on bad status lines.
Brendan Cully <brendan@kublai.com> [Mon, 07 Aug 2006 14:12:03 -0700] rev 2804
Update mq to use new logmessage arglist from 2794:bd8a9a94139f
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 16:47:06 -0500] rev 2803
Merge with crew
Matt Mackall <mpm@selenic.com> [Mon, 07 Aug 2006 16:27:09 -0500] rev 2802
Move repo.verify
Matt Mackall <mpm@selenic.com> [Thu, 03 Aug 2006 21:24:19 -0500] rev 2801
revlog.lookup optimization
Matt Mackall <mpm@selenic.com> [Thu, 03 Aug 2006 20:06:05 -0500] rev 2800
Remove **opts arg from commands.doupdate
Matt Mackall <mpm@selenic.com> [Thu, 03 Aug 2006 15:24:41 -0500] rev 2799
Move merge code to its own module
Pull update and merge3 out of localrepo into merge.py
s/self/repo/
Add temporary API function in hg.py
Convert all users
Matt Mackall <mpm@selenic.com> [Thu, 03 Aug 2006 13:07:57 -0500] rev 2798
Add a doc string
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 07 Aug 2006 19:08:55 -0700] rev 2797
fetch: add missing import.
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 07 Aug 2006 17:33:14 -0700] rev 2796
fetch: do not fetch if working dir modified
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 07 Aug 2006 17:27:38 -0700] rev 2795
new extension: fetch -> combine pull and merge/update
pull changes from remote repo.
if changes add new head: merge with new head, commit result.
else: update working dir to new head.
this makes useful command like git 'pull' command.
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 07 Aug 2006 17:23:04 -0700] rev 2794
commands.logmessage: simplify calling.
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 07 Aug 2006 17:22:17 -0700] rev 2793
localrepo.update: allow to suppress commit reminder.
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 06 Aug 2006 17:27:05 +0200] rev 2792
mq: unused variables, improper usage of 'is [not]', undefined variable
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 06 Aug 2006 17:24:13 +0200] rev 2791
mq: codingstyle
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 06 Aug 2006 17:20:51 +0200] rev 2790
mq: remove unecessary test
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 06 Aug 2006 17:18:53 +0200] rev 2789
mq: fix variables shadowing builtin
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 06 Aug 2006 16:42:36 +0200] rev 2788
util.patch: use shellquote
Brendan Cully <brendan@kublai.com> [Sat, 05 Aug 2006 14:59:47 -0700] rev 2787
mq: use more portable util.parse_patch_output instead of handrolled version.
Brendan Cully <brendan@kublai.com> [Sat, 05 Aug 2006 15:23:26 -0700] rev 2786
Add portable shell-quoting function; teach mq to use it.
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 06 Aug 2006 16:03:02 +0200] rev 2785
mailbomb: add a comment and remove the bcc in a more pythonic way
Brendan Cully <brendan@kublai.com> [Fri, 04 Aug 2006 14:36:03 -0700] rev 2784
Add '.' as a shortcut revision name for the working directory parent.
Vadim Gelfer <vadim.gelfer@gmail.com> [Sat, 05 Aug 2006 02:00:09 -0700] rev 2783
make error better.
Vadim Gelfer <vadim.gelfer@gmail.com> [Fri, 04 Aug 2006 10:38:17 -0700] rev 2782
merge with self.
Vadim Gelfer <vadim.gelfer@gmail.com> [Fri, 04 Aug 2006 10:37:40 -0700] rev 2781
fix output of test-backout.
Brendan Cully <brendan@kublai.com> [Thu, 03 Aug 2006 15:12:15 -0700] rev 2780
Test suite for log --follow and --follow-first.
Brendan Cully <brendan@kublai.com> [Thu, 03 Aug 2006 11:48:56 -0700] rev 2779
Add log option --follow-first to follow only the first parent of
merges.
This can be useful to extract a single line of development.
Brendan Cully <brendan@kublai.com> [Thu, 03 Aug 2006 13:29:31 -0700] rev 2778
Make log --follow revision range start default to working dir parent.
Brendan Cully <brendan@kublai.com> [Thu, 03 Aug 2006 11:06:09 -0700] rev 2777
Make log --follow without a file list follow a single head.
This includes all the parents or children (depending on the direction
of the revision range).
Vadim Gelfer <vadim.gelfer@gmail.com> [Thu, 03 Aug 2006 11:12:02 -0700] rev 2776
merge with brendan.
Brendan Cully <brendan@kublai.com> [Thu, 03 Aug 2006 11:09:12 -0700] rev 2775
Use StatusEntry class instead of repeated status line parsing.
This small cleanup abstracts out repeated instances of
(rev, name) = applied[x].split(':'), making it easier to use
applied or change the status file format.
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 02 Aug 2006 21:06:12 -0700] rev 2774
mq: make queue.unapplied useful as api
"Mathieu Clabaut <mathieu.clabaut@gmail.com>" [Wed, 02 Aug 2006 09:29:17 +0200] rev 2773
Import say a file is not tracked after patching : test case
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 02 Aug 2006 20:33:57 -0700] rev 2772
mq: simplify save_dirty
Brendan Cully <brendan@kublai.com> [Wed, 02 Aug 2006 15:07:37 -0700] rev 2771
Add -f option to qfold; improve qfold documentation.
Brendan Cully <brendan@kublai.com> [Wed, 02 Aug 2006 14:48:59 -0700] rev 2770
Clean up qnew help text.
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 02 Aug 2006 13:04:36 -0700] rev 2769
tests: fix output for test-backout
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 02 Aug 2006 12:46:00 -0700] rev 2768
allow None for path of hg.repository
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 02 Aug 2006 10:48:34 -0700] rev 2767
mq: rename read_series as parse_series, make simpler and faster
Bryan O'Sullivan <bos@serpentine.com> [Wed, 02 Aug 2006 09:18:56 -0700] rev 2766
fix call to commands.setremoteconfig
Vadim Gelfer <vadim.gelfer@gmail.com> [Wed, 02 Aug 2006 06:36:27 -0700] rev 2765
mq: print matches if patch name not unique
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 01 Aug 2006 19:58:34 -0700] rev 2764
clone: replace long help string of inaky with doc comment.
Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com> [Tue, 01 Aug 2006 11:12:34 -0700] rev 2763
Breakup clone -r extra text help to fit in 80 columns
Inaky Perez-Gonzalez <inaky.perez-gonzalez@intel.com> [Tue, 01 Aug 2006 09:50:30 -0700] rev 2762
Expand clone help with hardlink information
Christian Ebert <blacktrash@gmx.net> [Wed, 02 Aug 2006 03:17:06 +0200] rev 2761
fix hg backup option recommendation for auto-merge
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 01 Aug 2006 15:51:13 -0700] rev 2760
import: make patch apply if run in subdir
fix is same as for mq patch. patch apply code should be merged.
Christian Ebert <blacktrash@gmx.net> [Wed, 02 Aug 2006 00:20:28 +0200] rev 2759
hgcommand.vim: requirements & doc self-install more consistent; cosmetics
Check for availability of builtin system().
Set 'compatible' while sourcing script.
File operation during doc self-install:
Escape spaces in ex-commands.
Use quotes in system calls.
Incremented revision number(s).
Size of patch is due to cosmetic cleanups:
Removal of spaces at line endings.
<SID>Prefix to functions when called.
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 01 Aug 2006 15:40:54 -0700] rev 2758
fix mq test output.
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 01 Aug 2006 15:40:28 -0700] rev 2757
merge patches from brendan cully that did not apply clean against tip.
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 15:24:18 -0700] rev 2756
Add -s option to qseries: display first line of patch header.
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 14:56:54 -0700] rev 2755
Check for parent/mod as well as parent.mod in findext.
Debian/Ubuntu ship a system-wide hgrc that loads extensions using
the syntax hgext/extension=, rather than hgext.extension=. Although
the extensions work correctly, hg help <extension> doesn't
understand this form without this patch.
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 15:24:12 -0700] rev 2754
Add more verbose help text to mq commands.
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 15:24:11 -0700] rev 2753
Add -m, -l, -e options to qfold.
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 12:21:55 -0700] rev 2752
Add -f option to qdelete, to remove patch file.
Vadim Gelfer <vadim.gelfer@gmail.com> [Tue, 01 Aug 2006 10:57:28 -0700] rev 2751
mq: add qmv as alias for qrename
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 10:55:06 -0700] rev 2750
New self-explanatory command qrename.
Sébastien Pierre <sebastien@xprima.com> [Tue, 01 Aug 2006 09:51:36 -0400] rev 2749
[darcs2hg] Windows compatibilty patct
Originally submitted by Daiju Kito
1. single quote didn't work in several cases
2. os.unlink tries to delete tmpfile when it's called. It won't wait like unix.
Brendan Cully <brendan@kublai.com> [Mon, 31 Jul 2006 20:33:56 -0700] rev 2748
New mq command qfold: Merge patches into the current patch.
Patches should be in the series file but not yet applied.
Brendan Cully <brendan@kublai.com> [Mon, 31 Jul 2006 18:39:31 -0700] rev 2747
Add command qheader to display the header of a given patch.
Brendan Cully <brendan@kublai.com> [Mon, 31 Jul 2006 17:55:43 -0700] rev 2746
Add option -e/--edit to qrefresh, to edit the existing header.
Brendan Cully <brendan@kublai.com> [Mon, 31 Jul 2006 18:47:43 -0700] rev 2745
Change patch header as well as commit message with qrefresh -m or -l.
Brendan Cully <brendan@kublai.com> [Tue, 01 Aug 2006 02:23:05 -0700] rev 2744
Remove pointless reposetup hook from hgk.py
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 31 Jul 2006 12:02:13 -0700] rev 2743
import: switch from application/x-patch to text/x-diff
christian ebert says his mailer was broken.
Alexis S. L. Carvalho <alexis@cecm.usp.br> [Mon, 31 Jul 2006 14:22:17 -0300] rev 2742
mq: correct the use of super
The first argument passed to super is supposed to be the class where
it's being used, not self.__class__ - the two can be different when
there's inheritance involved.
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 31 Jul 2006 07:31:31 -0700] rev 2741
log: add -f/--follow option, to follow rename/copy
Vadim Gelfer <vadim.gelfer@gmail.com> [Mon, 31 Jul 2006 07:11:12 -0700] rev 2740
clean up hg.py: move repo constructor code into each repo module
Vadim Gelfer <vadim.gelfer@gmail.com> [Sun, 30 Jul 2006 22:52:34 -0700] rev 2739
merge with mpm.
Matt Mackall <mpm@selenic.com> [Mon, 31 Jul 2006 00:47:43 -0500] rev 2738
Kill ui.setconfig_remoteopts
This brought too much knowledge about command line particulars into
the ui code. Moved to commands.py.
Vadim Gelfer <vadim.gelfer@gmail.com> [Sun, 30 Jul 2006 21:46:38 -0700] rev 2737
pull: allow to pull from bundle file without need for bundle: syntax
Giorgos Keramidas <keramida@ceid.upatras.gr> [Sun, 30 Jul 2006 09:37:08 +0300] rev 2736
avoid calling (cd ...) with `nil' as a directory name
When (hg-root) cannot determine the current Mercurial root, print an
error message instead of trying to evaluate (cd nil).
Christian Ebert <blacktrash@gmx.net> [Sat, 29 Jul 2006 18:15:08 +0200] rev 2735
fix output file format %r
Alexis S. L. Carvalho <alexis@cecm.usp.br> [Fri, 28 Jul 2006 21:20:41 -0300] rev 2734
hbisect.py: don't rely on __del__ to write the current state.
This is yet another page of the "Thou shalt not do too much inside
__del__ methods" book, in the "demandload and __del__ don't go well
together" chapter.
The bisect extension is broken in 0.9.1:
$ hg bisect init
$ hg bisect bad
Fatal Python error: Interpreter not initialized (version mismatch?)
Aborted
(yes, I tripled checked my instalation to make sure the problem is not
there)
It's been broken since revision fe1689273f84 moved the import of the
binascii module into a demandload.
(In details: the first time that "hg bisect bad" (or good) is called,
there are still no revisions saved in .hg/bisect/*, so bisect.__init__
doesn't call hg.bin on anything. So, when we reach __del__, the
binascii module still hasn't been imported and we get that "nice"
message above.)
Christian Ebert <blacktrash@gmx.net> [Sat, 29 Jul 2006 01:58:12 +0200] rev 2733
hgcommand.vim: cleanup of doc self-install code
Vadim Gelfer <vadim.gelfer@gmail.com> [Sun, 30 Jul 2006 21:24:06 -0700] rev 2732
import: allow application/x-patch for funny patch from christian ebert
Christian Ebert <blacktrash@gmx.net> [Sat, 29 Jul 2006 10:27:54 +0200] rev 2731
patchbomb: do not write Bcc