Martin Geisler <mg@lazybytes.net> [Tue, 21 Jul 2009 00:17:10 +0200] rev 9231
i18n-da: switch back to "rettelse" for "patch"
Martin Geisler <mg@lazybytes.net> [Mon, 20 Jul 2009 23:55:55 +0200] rev 9230
i18n-da: updates in gpg, highlight, inotify, and mq
Martin Geisler <mg@lazybytes.net> [Mon, 20 Jul 2009 23:24:12 +0200] rev 9229
i18n-da: synchronized with 78e54b9f3a62
Cédric Duval <cedricduval@free.fr> [Sat, 18 Jul 2009 22:49:49 +0200] rev 9228
i18n-fr: synchronized with 78e54b9f3a62
Cédric Duval <cedricduval@free.fr> [Sat, 18 Jul 2009 22:45:59 +0200] rev 9227
i18n-fr: translations for dispatch.py
Cédric Duval <cedricduval@free.fr> [Sat, 18 Jul 2009 22:45:31 +0200] rev 9226
i18n-fr: translation of various core messages
Cédric Duval <cedricduval@free.fr> [Sat, 18 Jul 2009 22:44:50 +0200] rev 9225
i18n-fr: translation for the archive command
Cédric Duval <cedricduval@free.fr> [Sat, 18 Jul 2009 22:43:34 +0200] rev 9224
i18n-fr: translation of bisect
Wagner Bruna <wbruna@softwareexpress.com.br> [Thu, 16 Jul 2009 12:13:23 -0300] rev 9223
i18n-pt_BR: synchronized with 78e54b9f3a62
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 16:55:17 +0200] rev 9222
merge with stable
Wagner Bruna <wbruna@softwareexpress.com.br> [Mon, 06 Jul 2009 10:39:46 -0300] rev 9221
i18n-pt_BR: synchronized with 196d40e7c885 (1.3)
Greg Ward <greg-hg@gerg.ca> [Tue, 21 Jul 2009 20:31:45 -0400] rev 9220
transaction: document close(), abort() methods
Dirkjan Ochtman <dirkjan@ochtman.nl> [Thu, 23 Jul 2009 20:44:26 +0200] rev 9219
merge with crew-stable
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:26:42 +0200] rev 9218
zeroconf: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:26:26 +0200] rev 9217
win32text: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:25:54 +0200] rev 9216
win32mbcs: fix formatting of lists with proper reST markup
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:25:34 +0200] rev 9215
purge: fix formatting of lists with proper reST markup
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:24:52 +0200] rev 9214
patchbomb: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:24:34 +0200] rev 9213
parentrevspec: change list to a literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:24:03 +0200] rev 9212
pager: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:23:53 +0200] rev 9211
interhg: use reST syntax for literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:23:39 +0200] rev 9210
highlight: use reST syntax for literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:23:24 +0200] rev 9209
hgk: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:23:08 +0200] rev 9208
hgcia: use reST syntax for literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:22:50 +0200] rev 9207
extdiff: use reST syntax for literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:22:35 +0200] rev 9206
color: use reST syntax for literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:22:05 +0200] rev 9205
churn: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:21:46 +0200] rev 9204
bugzilla: wrap example template line
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:20:18 +0200] rev 9203
bugzilla: reformat list of configuration options
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:02:06 +0200] rev 9202
bookmarks: use reST syntax for literal block
Martin Geisler <mg@lazybytes.net> [Thu, 23 Jul 2009 00:01:24 +0200] rev 9201
acl: use reST syntax for literal blocks
Martin Geisler <mg@lazybytes.net> [Wed, 22 Jul 2009 23:26:27 +0200] rev 9200
transplant: better reST formatting
Bryan O'Sullivan <bos@serpentine.com> [Wed, 22 Jul 2009 15:28:30 -0700] rev 9199
Merge with crew-stable
Martin Geisler <mg@lazybytes.net> [Wed, 22 Jul 2009 23:12:54 +0200] rev 9198
coding style: use a space after comma
I left a cases like 'lambda x,y:' alone -- the lack of a space does
not bother me as much when the variables are single letters.
Martin Geisler <mg@lazybytes.net> [Wed, 22 Jul 2009 23:08:32 +0200] rev 9197
commands: remove ineffective parenthesis
Martin Geisler <mg@lazybytes.net> [Wed, 22 Jul 2009 22:56:45 +0200] rev 9196
merge with crew-stable
Christian Ebert <blacktrash@gmx.net> [Wed, 22 Jul 2009 08:28:30 +0100] rev 9195
keyword: reformat kwfiles help for minirst parser
Martin Geisler <mg@lazybytes.net> [Tue, 21 Jul 2009 12:19:53 +0200] rev 9194
hgignore.5: mark file names and cmdline options as literal text
Martin Geisler <mg@lazybytes.net> [Tue, 21 Jul 2009 12:09:32 +0200] rev 9193
merge with crew-stable
Matt Mackall <mpm@selenic.com> [Thu, 23 Jul 2009 12:58:15 -0500] rev 9192
Added signature for changeset 3ef6c14a1e8e
Matt Mackall <mpm@selenic.com> [Thu, 23 Jul 2009 12:58:10 -0500] rev 9191
Added tag 1.3.1 for changeset 3ef6c14a1e8e
Christian Ebert <blacktrash@gmx.net> [Wed, 22 Jul 2009 19:12:34 +0100] rev 9190
keyword: argument to "kwdemo --rcfile" must be string (bugfix)
Matt Mackall <mpm@selenic.com> [Wed, 22 Jul 2009 18:27:23 -0500] rev 9189
heads: more clarity for --closed
Matt Mackall <mpm@selenic.com> [Wed, 22 Jul 2009 18:08:02 -0500] rev 9188
Merge with crew-stable
Matt Mackall <mpm@selenic.com> [Wed, 22 Jul 2009 17:59:11 -0500] rev 9187
update: allow -c to jump branches if clean
Matt Mackall <mpm@selenic.com> [Wed, 22 Jul 2009 17:53:23 -0500] rev 9186
subrepo: use '/' for joining non-local paths
Matt Mackall <mpm@selenic.com> [Wed, 22 Jul 2009 00:28:37 -0500] rev 9185
fetch: drop force arg for commit (issue1752)
This appears to be unneeded and forces fetch commits to examine clean files.
Bryan O'Sullivan <bos@serpentine.com> [Wed, 22 Jul 2009 15:26:27 -0700] rev 9184
Fix issue1679: path reconstruction in hgwebdir was mangling things badly
Martin Geisler <mg@lazybytes.net> [Wed, 22 Jul 2009 22:50:34 +0200] rev 9183
do not translate commit messages
It is not very helpful to have 'Added tag %s for changeset %s' and
similar messages translated into different languages when people work
together using different locales.
We now use English strings without support for translations. If
needed, the user can still supply a custom string for most commands.
Martin Geisler <mg@lazybytes.net> [Tue, 21 Jul 2009 12:06:32 +0200] rev 9182
hgignore.5: expanded description
- explain that .hgignore must be created manually
- hint that it is typically put under revision control
Matt Mackall <mpm@selenic.com> [Mon, 20 Jul 2009 19:18:38 -0500] rev 9181
Merge with stable
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 19 Jul 2009 21:15:40 +0200] rev 9180
merge with -stable
Patrick Mezard <pmezard@gmail.com> [Sun, 19 Jul 2009 18:44:20 +0200] rev 9179
Update win32 build instructions for rst2html
Patrick Mezard <pmezard@gmail.com> [Sun, 19 Jul 2009 18:28:43 +0200] rev 9178
Merge with crew-stable
Matt Mackall <mpm@selenic.com> [Mon, 20 Jul 2009 19:15:44 -0500] rev 9177
Merge with crew-stable
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Sun, 19 Jul 2009 18:59:19 +0200] rev 9176
graphlog: fix incoming with local repo (issue1731)
Patrick Mezard <pmezard@gmail.com> [Sun, 19 Jul 2009 18:23:37 +0200] rev 9175
Mention GnuWin32 gettext in win32 build instructions
James Abbatiello <abbeyj at gmail.com> [Tue, 14 Jul 2009 22:38:12 -0400] rev 9174
Fix test-demandimport and test-trusted under Windows
The Windows-only wrapper around stdout is causing both of these tests to fail.
test-demandimport fails because it tries to print repr(sys.stdout). Use
stderr instead since that is not wrapped.
test-trusted fails because the wrapper doesn't handle softspace and an
unexpected extra space gets printed.
Matt Mackall <mpm@selenic.com> [Fri, 17 Jul 2009 11:52:53 -0500] rev 9173
Merge with crew-stable
Martin Geisler <mg@lazybytes.net> [Sat, 18 Jul 2009 22:28:28 +0200] rev 9172
doc: update README with links to Docutils
Martin Geisler <mg@lazybytes.net> [Fri, 17 Jul 2009 20:52:21 +0200] rev 9171
hg: better wrapping of string literal
Matt Mackall <mpm@selenic.com> [Sat, 18 Jul 2009 12:41:49 -0500] rev 9170
Merge with crew
Patrick Mezard <pmezard@gmail.com> [Fri, 17 Jul 2009 16:01:19 +0200] rev 9169
Merge with crew-stable
Patrick Mezard <pmezard@gmail.com> [Fri, 17 Jul 2009 13:31:12 +0200] rev 9168
convert/svn: fix backtrace in tags discovery (issue1742)
Brendan Cully <brendan@kublai.com> [Thu, 16 Jul 2009 21:05:24 -0700] rev 9167
Merge with hg
Matt Mackall <mpm@selenic.com> [Sat, 18 Jul 2009 12:34:38 -0500] rev 9166
changelog: move delayopener outside of class to eliminate reference cycle
Matt Mackall <mpm@selenic.com> [Sat, 18 Jul 2009 11:26:35 -0500] rev 9165
changelog: swap opener to switch delay modes
Matt Mackall <mpm@selenic.com> [Sat, 18 Jul 2009 11:25:55 -0500] rev 9164
changelog: factor out _delayname
Matt Mackall <mpm@selenic.com> [Sat, 18 Jul 2009 11:25:54 -0500] rev 9163
changelog: _delaycount -> _divert
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:26 +0200] rev 9162
doc: link man pages to one another
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:26 +0200] rev 9161
doc: fixup font markup in man pages
This exposed a bug in rst2man where it neglects to escape a literal
backslash. A patch has been applied upstream, but not yet packaged in,
say, Debian unstable. A forward-compatible work-around has therefore
been put in place.
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:26 +0200] rev 9160
help: fix bad indentation in template filters list
The indentation showed up in the hg.1.html file.
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:26 +0200] rev 9159
doc: add newline after literal blocks in man pages
The rst2man writer leaves no space between a literal block and the
following paragraph. This patch corrects this.
It has also been applied upstream. This does not conflict with this
change since any number of newlines can be added without effecting the
rendered man page.
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:26 +0200] rev 9158
doc: use reStructuredText for man and HTML pages
The Makefile now requires the rst2html and rst2man programs. Both can
be found in Debian testing or downloaded from the Docutils homepage:
http://docutils.sf.net/
http://docutils.sf.net/sandbox/manpage-writer/
The new HTML and man pages no longer contain huge amounts of
un-wrapping literal blocks, thanks to how snippets of reStructuredText
can easily be included inside other reStructuredText documents.
The HTML pages now have anchors for all sections, including the help
topics in hgrc.1 which were missing from the old HTML pages.
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:26 +0200] rev 9157
commands: use minirst parser when displaying help
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:25 +0200] rev 9156
minimal reStructuredText parser
Martin Geisler <mg@lazybytes.net> [Thu, 16 Jul 2009 23:25:25 +0200] rev 9155
util: wrap at termwidth-2 by default
Matt Mackall <mpm@selenic.com> [Thu, 16 Jul 2009 15:40:13 -0500] rev 9154
Merge with stable
Matt Mackall <mpm@selenic.com> [Thu, 16 Jul 2009 14:49:52 -0500] rev 9153
Add ui.progress API
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:41:19 -0400] rev 9152
tags: support 'instant' tag retrieval (issue548)
- modify _readtagcache() and _writetagcache() to read/write tag->node
mapping for global tags
- if (and only if) tip unchanged, use that cached mapping to avoid
reading any revisions of .hgtags
- change so tag names are UTF-8 in memory in tags.py, and converted to
local encoding as late as possible (in localrepository._findtags())
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:42 -0400] rev 9151
tags: implement persistent tag caching (issue548).
- rename findglobaltags() to findglobaltags1() (so the "no cache"
implementation is still there if we need it)
- add findglobaltags2() and make findglobaltags() an alias for it
(disabling tag caching is a one-line patch)
- factor out tagcache class with methods readcache() and writecache();
the expensive part of tag finding (iterate over heads and find
.hgtags filenode) is now in tagcache.readcache()
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9150
localrepo: add destroyed() method for strip/rollback to use (issue548).
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9149
Factor tags module out of localrepo (issue548).
Currently only handles reading tags, and will soon grow support for
tag caching. Could eventually deal with updating tags too.
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9148
localrepo: factor updatetags() out of readtags() (issue548).
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9147
localrepo: improve readability of _findtags(), readtags() (issue548).
- rename many local variables
- add some comments
- refactor call to line.split() (catch ValueError rather
than checking length of return value: one less local variable)
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9146
localrepo: rename in-memory tag cache instance attributes (issue548).
- self.tagscache to self._tags
- self._tagstypecache to self._tagtypes
- this is for consistency, readability, privacy, and to subtly hint
that "caching" is something else
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9145
localrepo: factor _findtags() out of tags() (issue548).
This makes in-memory caching the sole responsibility of localrepo,
eliminating some localrepo code that was duplicated in mq and
bookmarks.
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9144
test-tags: enhance the test to probe tag caching better (issue548).
- give detailed dumps of .hgtags and localtags content
- repeat some query operations to expose cache bugs
- ensure that rollback/strip undo tagging operations
Greg Ward <greg-hg@gerg.ca> [Thu, 16 Jul 2009 10:39:41 -0400] rev 9143
test-tags: clarify test output; simplify test script a bit (issue548).
- use simpler way ('hg id') to get current changeset id
- add 'echo' statements to guide the eye when reading output
- add some more output to clarify the state of .hgtags
- drop '-d' option from every commit/tag, since run-tests.py
does this (although with a different timestamp, so changeset
ids differ)
Martin Geisler <mg@lazybytes.net> [Wed, 15 Jul 2009 17:44:47 +0200] rev 9142
merge with crew-stable
Brendan Cully <brendan@kublai.com> [Tue, 14 Jul 2009 18:00:37 -0700] rev 9141
Merge with crew-stable
Bryan O'Sullivan <bos@serpentine.com> [Tue, 14 Jul 2009 17:21:10 -0700] rev 9140
Merge with crew-stable
Brendan Cully <brendan@kublai.com> [Tue, 14 Jul 2009 16:51:52 -0700] rev 9139
Merge with crew-stable
Martin Geisler <mg@lazybytes.net> [Wed, 15 Jul 2009 00:24:20 +0200] rev 9138
merge with mpm
Martin Geisler <mg@lazybytes.net> [Wed, 15 Jul 2009 00:19:15 +0200] rev 9137
merge with crew-stable
Nicolas Dumazet <nicdumz.commits@gmail.com> [Mon, 13 Jul 2009 09:50:26 +0900] rev 9136
for calls expecting bool args, pass bool instead of int
str.splitlines and email.message.as_string both expect a bool argument
defaulting at False: replace f(1) by f(True) and f(0) by f()
Benoit Boissinot <benoit.boissinot@ens-lyon.org> [Tue, 14 Jul 2009 20:24:16 +0200] rev 9135
merge with -stable
Martin Geisler <mg@lazybytes.net> [Tue, 14 Jul 2009 20:10:23 +0200] rev 9134
merge with crew-stable
Simon Heimberg <simohe@besonet.ch> [Tue, 14 Jul 2009 14:05:07 +0200] rev 9133
store: eliminate reference cycle in fncachestore
no reference to self in fncacheopener
Inspired by Adrian Buehlmann
Shun-ichi GOTO <shunichi.goto@gmail.com> [Fri, 10 Jul 2009 19:34:08 +0900] rev 9132
win32mbcs: add special wrapper for osutil.listdir().
osutil.listdir() may misinterpret 0x5c of MBCS 2nd. byte as path
separator. New wrapper prevents this by adding it beforehand.
Shun-ichi GOTO <shunichi.goto@gmail.com> [Fri, 10 Jul 2009 15:52:01 +0900] rev 9131
win32mbcs: wrapper supports keyword arguments and dict result.
The keyword arguments are also decoded before calling original.
And dict of return value is also encoded back.
Cédric Duval <cedricduval@free.fr> [Fri, 10 Jul 2009 19:45:31 +0200] rev 9130
gendoc: fall back to pure modules if C extensions are not available (issue1711)
James Abbatiello <abbeyj at gmail.com> [Tue, 14 Jul 2009 22:42:05 -0400] rev 9129
Don't copy hidden files/directories during `setup.py install`
This is useful if a copy of Mercurial is stored in a Subversion repository
so that the .svn directories don't get copied.
Dan Villiom Podlaski Christiansen <danchr@gmail.com> [Fri, 10 Jul 2009 13:40:25 +0200] rev 9128
commands: hide deprecated commands.
A command is considered deprecated if the word "DEPRECATED" is found
in the doc string. Such commands are hidden from non-verbose help.
Henri Wiechers <hwiechers@gmail.com> [Sun, 12 Jul 2009 07:56:43 +0200] rev 9127
hgignore.5.txt: improved description of matching
Improved the description of the matching behavior used with .hgignore.
Made some minor language improvements.
Brendan Cully <brendan@kublai.com> [Tue, 14 Jul 2009 17:59:36 -0700] rev 9126
Merge with bos
Brendan Cully <brendan@kublai.com> [Tue, 14 Jul 2009 17:58:36 -0700] rev 9125
Fix issue1738 for strip too.
I see no reason to open every touched file at once.
Bryan O'Sullivan <bos@serpentine.com> [Tue, 14 Jul 2009 17:18:45 -0700] rev 9124
setup.py: don't pollute the current directory with temporary files
Brendan Cully <brendan@kublai.com> [Tue, 14 Jul 2009 16:50:37 -0700] rev 9123
Make patch.diff filelog cache LRU of 20 files. Fixes issue1738.
20 files is as fast as 200 for hg diff -r 28015:30103 of mozilla-central.
Ideally we'd use util.lrucachefunc, but the interface doesn't quite work.
Wagner Bruna <wbruna@softwareexpress.com.br> [Tue, 14 Jul 2009 17:12:12 -0300] rev 9122
url: fix use of non-int port in https connections via proxy
Complements eef406165507 (issue1725).
Greg Ward <greg-hg@gerg.ca> [Sun, 12 Jul 2009 22:33:00 -0400] rev 9121
test-fetch: fix non-portable sed regex.
(s/...\+/.../ appears to be a GNU-ism: this test broke on OS X and
NetBSD. Changing \+ to * fixes it, although that is a slightly less
strict regex.)
Henrik Stuart <hg@hstuart.dk> [Mon, 13 Jul 2009 20:19:17 +0200] rev 9120
branch heads: optimise computation of branch head cache (issue1734)
The previous branch heads cache implementation iterated all ancestors
for each new revision in the repository, causing a massive slowdown on
cloning larger repositories.
Matt Mackall <mpm@selenic.com> [Tue, 14 Jul 2009 12:18:22 -0500] rev 9119
Merge with stable
Steve Borho <steve@borho.org> [Sun, 12 Jul 2009 00:46:43 -0500] rev 9118
cmdutil: fall back to filename if glob expand has errors
On Windows, Mercurial tries to glob expand provided filenames as a
convenience to the user. Unfortunately, there are valid filenames
which are not valid glob patterns. In those cases, we should fallback
to the original provided filename.
Nicolas Dumazet <nicdumz.commits@gmail.com> [Mon, 13 Jul 2009 21:55:17 +0900] rev 9117
inotify: server: explicitely ignore events in subdirs of .hg/ (issue1735)
Nicolas Dumazet <nicdumz.commits@gmail.com> [Mon, 13 Jul 2009 16:49:05 +0200] rev 9116
inotify: fix issue1375, add a test.
The biggest problem was the data structure, which did not allow changing
a file into a directory or vice versa. This problem is fixed by b55d44719b47.
The walk() method also had an issue in this case:
- we know 'path' as a directory. inotify server sleeps.
- 'path' is deleted
- 'path' is recreated as a file
- the server catches up here, and see the deletion. it instantiates a scan(),
which in its turn calls for walk(repo, path).
- walk() then assumes that 'path' is a directory and calls os.listdir on it,
which raises an OSError(errno.ENOTDIR)
Catch the error, and yield the file instead of the directory contents.
Nicolas Dumazet <nicdumz.commits@gmail.com> [Tue, 26 May 2009 23:00:35 +0900] rev 9115
inotify: server: new data structure to keep track of changes.
== Rationale for the new structure ==
Current structure was a dictionary tree. One directory was tracked
as a dictionary:
- keys: file/subdir name
- values:
- for a file, the status (a/r/m/...)
- for a subdir, the directory representing the subdir
It allowed efficient lookups, no matter of the type of the terminal leaf:
for part in path.split('/'): tree = tree[part]
However, there is no way to represent a directory and a file with the same name
because keys are conflicting in the dictionary. Concrete example:
Initial state:
root dir
|- foo (file)
|- bar (file)
# data state is: {'foo': 'n', 'bar': 'n'}
Remove foo:
root dir
|- bar (file)
# Data becomes {'foo': 'r'} until next commit.
Add foo, as a directory, and foo/barbar file:
root dir
|- bar (file)
|-> foo (dir)
|- barbar (file)
# New state should be represented as:
{'foo': {'barbar': 'a'}, 'bar': 'n'}
however, the key "foo" is already used and represents the old file.
The dirstate:
D foo
A foo/barbar
cannot be represented, hence the need for a new structure.
== The new structure ==
'directory' class. Represents one directory level.
* Notable attributes:
Two dictionaries:
- 'files' Maps filename -> status for the current dir.
- 'dirs' Maps subdir's name -> directory object representing the subdir
* methods
- walk(), formerly server.walk
- lookup(), old server.lookup
- dir(), old server.dir
This new class allows embedding all the tree walks/lookups in its own class,
instead of having everything mixed together in server.
Incidently, since files and directories are not stored in the same
dictionaries, we are solving the previous key conflict problem.
The small drawback is that lookup operation is a bit more complex:
for a path a/b/c/d/e we have to check twice the leaf, if e is a directory or a
file.
Martin Geisler <mg@lazybytes.net> [Sun, 12 Jul 2009 21:41:12 +0200] rev 9114
merge with crew-stable
Martin Geisler <mg@lazybytes.net> [Sun, 12 Jul 2009 21:37:24 +0200] rev 9113
gitweb, monoblue: fix double-spacing in file view (issue1733)
The error was introduced in 338412820a57.
Simon Heimberg <simohe@besonet.ch> [Fri, 10 Jul 2009 17:54:04 +0200] rev 9112
util: use propertycache in opener instead of __getattr__