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
Matt Mackall <mpm@selenic.com> [Fri, 28 May 2010 16:01:57 -0500] rev 11228
Merge with stable
Matt Mackall <mpm@selenic.com> [Fri, 28 May 2010 14:41:11 -0500] rev 11227
status: avoid performance regression when no .hgsub is present
(introduced by
24ce8f0c0a39)
Martin Geisler <mg@aragost.com> [Thu, 27 May 2010 10:37:14 +0200] rev 11226
doc: document expansion of %include and ui.username