Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 11:18:57 -0500] rev 11276
revset: add a debugrevspec command
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 11:18:57 -0500] rev 11275
revset: introduce revset core
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 11:18:57 -0500] rev 11274
revset: introduce basic parser
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 11:18:57 -0500] rev 11273
remoteui: move from cmdutil to hg
Henrik Stuart <hg@hstuart.dk> [Thu, 03 Jun 2010 21:38:30 +0200] rev 11272
mq: better qqueue test for troublesome names
Cédric Duval <cedricduval@free.fr> [Thu, 03 Jun 2010 20:40:23 +0200] rev 11271
mq: prevent the creation of a queue whose name is already taken
Each check is moved under the code handling the relevant option, and
a new one is added for --create. This fixes duplicated entries being
added to the queues list.
Henrik Stuart <hg@hstuart.dk> [Wed, 02 Jun 2010 19:39:45 +0200] rev 11270
mq: fix naming issues for qqueue directories
Martin Geisler <mg@aragost.com> [Wed, 02 Jun 2010 14:54:25 +0200] rev 11269
win32text: mark this extension as deprecated
Martin Geisler <mg@aragost.com> [Wed, 02 Jun 2010 14:40:31 +0200] rev 11268
Merge with stable
Greg Ward <greg-hg@gerg.ca> [Tue, 01 Jun 2010 18:29:52 -0400] rev 11267
shrink-revlog: preserve mode of the shrunken index and data file.
Otherwise, the shrunken index file always has mode 0600 thanks to
mkstemp(). This is annoying on a server, where multiple users may need
to read/write the manifest. chmod()ing the data file is not strictly
necessary, but it's nice for consistency.
Javi Merino <cibervicho@gmail.com> [Wed, 19 May 2010 18:06:35 +0200] rev 11266
Fixed a bashism with the use of $RANDOM in hgeditor.
The variable $RANDOM is not POSIX so a portable /bin/sh may not define
it. When creating a directory with a random name it's better to use
mktemp, which, even though is not POSIX, exists in common Unixes
including Linux, OpenBSD, FreeBSD and MacOS X.
Alexander Solovyov <piranha@piranha.org.ua> [Thu, 27 May 2010 22:50:23 +0300] rev 11265
churn: add possibility to include/exclude paths
Alexander Solovyov <piranha@piranha.org.ua> [Wed, 02 Jun 2010 14:28:45 +0200] rev 11264
churn: support spaces in aliases (
issue2222)
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 10:55:11 -0500] rev 11263
Merge with stable
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 10:54:57 -0500] rev 11262
Merge with i18n
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 10:54:03 -0500] rev 11261
Added signature for changeset
24fe2629c6fd
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 10:54:03 -0500] rev 11260
Added tag 1.5.4 for changeset
24fe2629c6fd
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 10:44:03 -0500] rev 11259
Merge with i18n
Wagner Bruna <wbruna@softwareexpress.com.br> [Tue, 01 Jun 2010 12:25:43 -0300] rev 11258
merge with i18n stable
Wagner Bruna <wbruna@softwareexpress.com.br> [Tue, 01 Jun 2010 12:23:44 -0300] rev 11257
i18n-pt_BR: synchronized with
0bb67503ad4b
Matt Mackall <mpm@selenic.com> [Tue, 01 Jun 2010 10:12:07 -0500] rev 11256
Merge with stable
Adrian Buehlmann <adrian@cadifra.com> [Fri, 28 May 2010 17:28:34 +0200] rev 11255
clone: save hardlink state of util.copyfiles()
When trying to do hardlink-cloning, the os_link() call of the
first file tried already fails on Windows, if the source is on a
UNC path.
This change avoids calling os_link() for the rest of files, leaving
us with a *single* failed os_link() call per clone operation, if the
source can't do hardlinks.
Adrian Buehlmann <adrian@cadifra.com> [Fri, 28 May 2010 17:28:34 +0200] rev 11254
util.copyfiles: don't try os_link() again if it failed before
If the os_link() call on the first file in the directory fails [1],
we switch mode to using shutil.copy() for all remaining files.
[1] happens for example on Windows for every file when cloning from a UNC
path without specifying --pull.
Christian Ebert <blacktrash@gmx.net> [Tue, 01 Jun 2010 03:45:10 +0200] rev 11253
keyword: retrieve added and modified files from commitctx
Christian Ebert <blacktrash@gmx.net> [Tue, 01 Jun 2010 01:05:45 +0200] rev 11252
keyword: force dirstate normal when all changes in a file are recorded
Before this bugfix a file whose changes were entirely recorded was still
considered modified by "hg status".
Note: the test must use hg record -l/--logfile, because this is not
reproducible with hg record -m/--message.
Adrian Buehlmann <adrian@cadifra.com> [Mon, 31 May 2010 13:47:51 +0200] rev 11251
clone: print number of linked/copied files on --debug
Martin Geisler <mg@lazybytes.net> [Mon, 31 May 2010 21:43:03 +0200] rev 11250
Merge with stable
Martin Geisler <mg@lazybytes.net> [Mon, 31 May 2010 21:37:01 +0200] rev 11249
eol: extension for managing file EOLs
Matt Mackall <mpm@selenic.com> [Mon, 31 May 2010 12:57:47 -0500] rev 11248
Merge with stable
Matt Mackall <mpm@selenic.com> [Mon, 31 May 2010 12:57:34 -0500] rev 11247
Merge with i18n
Matt Mackall <mpm@selenic.com> [Mon, 31 May 2010 12:57:24 -0500] rev 11246
Merge with i18n
Wagner Bruna <wbruna@softwareexpress.com.br> [Tue, 25 May 2010 13:24:49 -0300] rev 11245
merge with i18n stable
Wagner Bruna <wbruna@softwareexpress.com.br> [Tue, 25 May 2010 13:20:33 -0300] rev 11244
i18n-pt_BR: added message from
1b8aa9ffa7dc
Wagner Bruna <wbruna@yahoo.com> [Sat, 22 May 2010 11:23:47 -0300] rev 11243
merge with i18n stable
Wagner Bruna <wbruna@yahoo.com> [Sat, 22 May 2010 11:22:36 -0300] rev 11242
i18n-pt_BR: synchronized with
ed71cb07d7b2
Matt Mackall <mpm@selenic.com> [Mon, 31 May 2010 12:34:47 -0500] rev 11241
Merge with stable
Brodie Rao <brodie@bitheap.org> [Wed, 19 May 2010 09:54:12 -0500] rev 11240
pager: exec pager using /bin/sh -c
This mimics the behavior of popen() and ensures setting the pager to something
like "LESS=FSRX less" will work.
Matt Mackall <mpm@selenic.com> [Fri, 14 May 2010 12:57:24 -0500] rev 11239
hgweb: fix race in refreshing repo list (
issue2188)
Martin Geisler <mg@aragost.com> [Mon, 31 May 2010 13:55:47 +0200] rev 11238
Merge with stable
Nicolas Dumazet <nicdumz.commits@gmail.com> [Mon, 31 May 2010 17:13:15 +0900] rev 11237
record: check that we are not committing a merge before patch selection
It might sound like a good idea to use record to filter changes when merging.
If someone attemps this, it's better to tell her "no" right ahead, before the
patch generation/line selection, so she does not spend time doing it just to
receive a red light after that (sometimes rather long) process.
Martin Geisler <mg@aragost.com> [Mon, 31 May 2010 13:43:03 +0200] rev 11236
record: better way to find help in docstring
This outputs only the lines in the verbatim block without relying on
the translator to have translated the little 'y - record this change'
string.
Nicolas Dumazet <nicdumz.commits@gmail.com> [Mon, 31 May 2010 14:48:26 +0900] rev 11235
record: docstring typo
Martin Geisler <mg@lazybytes.net> [Sun, 30 May 2010 23:24:10 +0200] rev 11234
mq: mention qqueue in module docstring
Adrian Buehlmann <adrian@cadifra.com> [Fri, 28 May 2010 17:28:34 +0200] rev 11233
clone: save hardlink state of util.copyfiles()
When trying to do hardlink-cloning, the os_link() call of the
first file tried already fails on Windows, if the source is on a
UNC path.
This change avoids calling os_link() for the rest of files, leaving
us with a *single* failed os_link() call per clone operation, if the
source can't do hardlinks.
Adrian Buehlmann <adrian@cadifra.com> [Fri, 28 May 2010 17:28:34 +0200] rev 11232
util.copyfiles: don't try os_link() again if it failed before
If the os_link() call on the first file in the directory fails [1],
we switch mode to using shutil.copy() for all remaining files.
[1] happens for example on Windows for every file when cloning from a UNC
path without specifying --pull.
Alexander Solovyov <piranha@piranha.org.ua> [Thu, 27 May 2010 22:50:04 +0300] rev 11231
cmdutil: cleanup imports
Ronny Pfannschmidt <Ronny.Pfannschmidt@gmx.de> [Thu, 27 May 2010 17:47:40 +0200] rev 11230
make transactions work on non-refcounted python implementations
Henrik Stuart <hg@hstuart.dk> [Sat, 29 May 2010 20:32:39 +0200] rev 11229
mq: support multiple patch queues using qqueue