inotify: server: use wprefix everywhere, introduce prefixlen self.wprefix = self.repo.root + '/' = self.repo.wjoin('') + '/' Since we'll eventually get rid of self.repo, keep only the first form. use self.prefixlen to compute only once the root prefix size.

inotify: server._updatestatus: simplify control flow

url: add support for custom handlers in extensions

patchbomb: add --flag to put flags in subject prefixes --flag foo uses: [PATCH foo] or [PATCH M of N foo] depending on the number of patches. Multiple flags are supported: --flag foo --flag bar gives [PATCH foo bar]

Merge with stable

Merge with i18n-stable

Merge with i18n

merge with i18n-stable

i18n-pt_BR: miscellaneous fixes

i18n-pt_BR: synchronized with 2de7d96593db

config: improve code readability

mq: re-phrase the qrefresh help about the user and date options

mq: add the date with qrefresh, even if missing (issue1768) When setting the date with qrefresh (-d|-D), and the patch does have neither a 'Date:' nor a '# Date' header, add the date field as such: - if the patch has a 'From:' header, add a 'Date:' header - otherwise, upgrade to Mercurial extended patch format and add '# Date'

mq: upgrade the patch to HG format when adding the author name with qrefresh When setting the author field with qrefresh (-u|-U), and the patch does have neither a 'From:' nor a '# User' header, add the author name as such: - if the patch has a 'Date:' header, add a 'From:' header - otherwise, upgrade to Mercurial extended patch format and add # From

mq: add function to check if a header exists in a patch

mq: fix coding style in qrefresh

Merge with crew-stable.

hook: fix full path imports on Windows (issue1779) Bottom portion fixes full path imports on source installs on Windows. The top portion further fixes full path imports on binary installs. Initial patch by Roman V. Kiseliov

patch: marked string for translation

diffstat: scale adds/removes proportionally to graph width The previous method of scaling had a tendency to include graph lines that went past the output width when the file with the most changes had a very large number of changes.

test-clone-failure: reenable perm to allow deletion

patch: create file even if source is not /dev/null as long as the first hunk is -0,0 and the source and destination have the same name. Matt found a patch like this, and /usr/bin/patch accepts it.

Merge with crew

Fix failing darcs test

Merge with mpm

Merge backout of 26fa0e31011d

Backed out changeset 26fa0e31011d. On OS X, it broke test-issue322, test-issue660, test-permissions, test-symlink-addremove, and test-walk.

mercurial.ini: provide more useful info * Add warning about not editing the site-wide file on Vista * More detail about where to create user Mercurial.ini * Provide examples for ssh tools * List extensions, all disabled * add [patch] section, for fixing patch behavior when using win32text * example [extdiff] section, all disabled * example [diff] section, all disabled

churn: correct output when run in subdirectories

i18n: move unrelated line out of try-except block

i18n: updated outdated comment

merge backout of 5724cd7b3688

Backed out changeset 5724cd7b3688 (depends on other changes, no tests)

merge with crew-stable

dispatch: also pass level argument to __import__ for ignored modules I wanted to check if mercurial.demandimport could speed up the loading of PyObjC, and ran into this: the level argument for __import__, available in Python 2.5 and later, is silently dropped when doing an 'import *'. I have no idea what these arguments mean, but this minor change made it work. (Oh, and because of that 'from ... import *', PyObjC still took about 2s...)

lsprof: make profile not die when imported modules changes (issue1774)

notify: do not mime encode multipart templates Mulitpart templates should take care of this themselves. See http://www.selenic.com/pipermail/mercurial/2009-July/027017.html Also catch potential parsing errors gracefully.

kill trailing whitespace

color: simplify & improve colorization of qseries Use 'repo.mq.series' to obtain patch names, instead of deriving them from the qseries output. This is both simpler and more robust, and fixes colorization of patches without a message with -s/--summary.

gendoc: documentation section describing available extensions

localrepo: document the locking scheme a little better - localrepo: document lock(), wlock() methods - lock: add class docstring

keyword: use util.datestr for utcdate filter, expose through variable

keyword: reference templating help, add utcdate filter example

keyword: remove "help keyword" from test Testing help output does not make sense as we have no way to check for errors in content which would be the only reasonable test here.

keyword: update copyright

hgwebdir: add links to the main repository page in notfound.tmpl

Merge with crew-stable

help: improve "hg help revisions" Add info about branch names, and other clarifications.

rebase: link to RebaseExtension This new page has the usual user-oriented information about the extension.

patchbomb: use a list instead of indented paragraphs

help: un-indent list in "hg help diffs"

help: add "templates" as an alias for "templating" help topic

gendoc: make anchors for help topic names

gendoc: rename underlined/bold to section/subsection

help: use field lists for lists of extensions

help: format templating help topic with a field list

minirst: parse field lists

minirst: simplify bullet list indentation computation

minirst: indent literal blocks with two spaces The vast majority* of them are formatted like this in the source, so this basically reverts the output to how it looked before we got the minirst parser. *: the help on templating use four spaces for some examples and will now shown with an indentation of just two spaces.

help: mark literal blocks

patchbomb: reST syntax for literal blocks in help text

Merge with crew-stable

mq: Parse commit message after we find start of changeset patch patch.extract extracts just the message below HG changeset patch. So to not parse comments above the patch, we reset the message if we find a changeset patch as we know the commit message comes after the changeset patch data. This is similar to the behavior of patch.extract. The drawback of this approach, after a qref -e the comment above HG changeset patch is gone. Without this patch the message would become part of the commit message in the HG changset patch part.

doc: fix quotes mismatches affecting rst

Merge with crew-stable

test-repair-strip: Don't rely on cat error message

test-gpg: suppress secure memory warning Some platforms (Solaris for one) does not support secure memory and would give a warning.

Merge with main

keyword: refactor kwdemo and make output translatable Better reflect the actual behaviour of the extension: - Make map arguments and -f/--rcfile not mutually exclusive but extend the current configuration - Map arguments and -f/--rcfile both override the defaults even when -d/--default is specified - -d/--default only overrides the current configuration Inform the user about extending/overriding behaviour, but only at the beginning; the following messages become terser, making the output translatable without too much code clutter. Rephrase help (use "short/long" option notation etc.).

commands: Check if helptext contains a newline before we split

Merge with crew-stable

Makefile: remove non-default wrapping of hg.pot strings

help: wrap help strings at 70 characters

zeroconf: wrap docstrings at 70 characters

win32mbcs: wrap docstrings at 70 characters

transplant: wrap docstrings at 70 characters

share: wrap docstrings at 70 characters

record: wrap docstrings at 70 characters

rebase: wrap docstrings at 70 characters

purge: wrap docstrings at 70 characters

patchbomb: wrap docstrings at 70 characters

parentrevspec: wrap docstrings at 70 characters

pager: wrap docstrings at 70 characters

notify: wrap docstrings at 70 characters

mq: wrap docstrings at 70 characters

keyword: wrap docstrings at 70 characters

interhg: wrap docstrings at 70 characters

highlight: wrap docstrings at 70 characters

hgk: wrap docstrings at 70 characters

hgcia: wrap docstrings at 70 characters

graphlog: wrap docstrings at 70 characters

fetch: wrap docstrings at 70 characters

extdiff: wrap docstrings at 70 characters

convert: wrap docstrings at 70 characters

color: wrap docstrings at 70 characters

churn: wrap docstrings at 70 characters

children: wrap docstrings at 70 characters

bugzilla: wrap docstrings at 70 characters

bookmarks: wrap docstrings at 70 characters

acl: wrap docstrings at 70 characters

commands: wrap docstrings at 70 characters It is no longer necessary to wrap the docstrings at 70 characters in the source -- with the reST parser, they are re-formatted to fit the terminal when shown. However, wrapping the docstrings at 78 characters makes life harder for translators because it marks a lot of strings as fuzzy for no good reason. When un-marking them, the translators would have to examine each string again and determine if it is merely re-wrapped or if the content was also changed. The long lines also introduce very ugly linebreaks in the .po files if they are processed using the standard Gettext tools without using something like '--width 85' all the time.

Merge with crew

Merge with stable

email: Catch exceptions during send. Catch SMTPlib exceptions to avoid mercurial showing a backtrace during an e.g. authentication error.

update: make a naked except more specific (issue1766)

tests/hghave: bzr114 checks for bzr >= 1.14 Everything but the implementation indicated that bzr114 also should be true for "or higher".

fix issue 1763: strip chars from end of line when parsing gitpatch lines

issue1251: bail if darcs version is too old

Merge with crew-stable

gendoc: don't translate topic strings twice (issue1760)

dirstate: fold paths using the just added util.realpath() function. Using the one true canonical path of a directory entry allows equivalent paths to be treated gracefully. Equivalent means, in this case, differing directory entries resulting in the same, unique file system link to a file.

util: add realpath() for getting the 'true' path. The function is implemented for Mac OS X using the F_GETPATH fcntl, and a basic implementation for Windows is provided as well. On other POSIX systems, vanilla os.path.realpath() is used.

bookmarks: optimize update routine in case track.current is set

bookmarks: Teach addchangset to respect track.current addchangeset didn't respect track.current, resulting in hg pull to forward all bookmarks although track.current is set to True. Reported by Christos Trochalakis.

bookmarks: Wrap commictx instead of commit (issue 1515)

Merge with stable

Merge with i18n

i18n-de: fix indention

i18n-da: switch back to "rettelse" for "patch"

i18n-da: updates in gpg, highlight, inotify, and mq