record: use absolute path instead of os.chdir Record was changing the current directory to `repo.root` in order to be able to feed `command.commit` file name relative to this `repo.root`. This is a bit overkill and prevent an incoming fix to rebase. This would also break multi-threaded usage. Instead we just feed `command.commit` with absolute path name. works as well as before but without chdir.

help: improve description of phases.checksubrepos setting The existing description was a bit hard to understand.

clone: do not turn hidden changeset public on publishing clone (issue3935) Before this changeset local clone of a repo with hidden changeset would include then in the clone (why not) and turn them public (plain wrong). This happened because the copy clone publish by dropping the phaseroot file entirely making everything in the repo public (and therefore immune to obsolescence marker). This changeset takes the simplest fix, we deny the copy clone in the case of hidden changeset falling back to pull clone that will exclude them from the clone and therefore not turning them public. A smarter version of copy clone could be done, but I prefer to go for the simplest solution first.

convert: use branchmap to change default branch in destination (issue3469) The fix for issue2653 broke the ability to map the default branch of a source repository to a non-default named branch in the destination repository. Leave the default behaviour as is, but allow the branch name "None" to be used to map to a non-default named branch in the destination repository.

dispatch: take --hidden from individual commands into account The command server would otherwise ignore that option, since the repo object is only created once.

doc: fix mistake about matching against directories in "pattern.txt" This fixes mistake of documentation about matching against directories in "pattern.txt" introduced by 50db996bccaf. ".hgignore" treats specified "glob:" pattern as same as one specified for "-X" option: it can match against directories, too. For reference, extra regexp string appended to specified pattern for each types are listed below: see also "match.match()" and "match._regex()" for detail. ============= ========== =============== type cmdline -I/-X ============= ========== =============== glob/relglob '$' '(?:/|$)' path/relpath '(?:/|$)' '(?:/|$)' re/relre (none) (none) ============= ========== =============== Appending '$' means that the specified pattern should match against only files.

dispatch: make "_checkshellalias()" invoke "findcmd()" with "strict=True" Before this patch, shell alias may be executed by abbreviated command name unexpectedly, even if abbreviated command name matches also against the command provided by extension. For example, "rebate" shell alias is executed by "hg reba", even if rebase extension (= "rebase" command) is enabled. In this case, "hg reba" should be aborted because of command name ambiguity. This patch makes "_checkshellalias()" invoke "cmdutil.findcmd()" always with "strict=True" (default value). If abbreviated command name matches against only one shell alias even after loading extensions, such shell alias will be executed via "_parse()". This patch doesn't remove "_checkshellalias()" invocation itself, because it may prevent shell alias from loading extensions uselessly.

rebase: abort cleanly when we encounter a damaged rebasestate (issue4155)

i18n-ja: change phrasing for "or"

i18n-ja: fix some quoting problems for space character usage

i18n-ja: synchronized with f493c2f67430

i18n: fix non-matching 1st line indentations When generating documentation, indentation must match for getting the same view for translated messages. Often an output is generated anyway, but it can look different. When a syntactically wrong indentation change is done, runrst will fail (this is detected by test-gendoc.t). Fix the simple places. When translation knowledge is necessary, the entry is marked as fuzzy (and therefore skipped when generating translations). A translator can fix it later.

i18n: fix some non matching quotation marks in translations A test for this is in preparation.

i18n: do not translate rst syntax .. note:: .. note:: is rst syntax which must not be translated. Fix this in the translations. This is not the first time this happens, so there should be a note for the translator. A later patch will change the generation of the po files to write this automatically. A test in i18n/check-translation.py could help as well.

i18n-pt_BR: synchronized with 268a5ab5c27b

subrepo: make it possible to update to hidden subrepo revisions When a subrepo revision was hidden it was considered missing and mercurial was unable to update to the corresponding parent revision. Instead warn the user of the problem and let it choose what to do (the default is to udpate anyway).

subrepo: remove unnecessary else clause in hgsubrepo._get This revision has no behaviour change. It simply removes an unnecessary else that follows an if / return block. The change looks big because a big chunk of code has been unindented one level.

subrepo: do not try to get hidden revisions If a subrepo revision is hidden (because it was amended, for example) it does not make sense to try to "get" it from the remote subrepository. Note that in order to avoid making the change look bigger than it is, this adds an unnecessary else clause. This will be removed on a follow up patch.

parsers: fix 'unsigned expression is always true' warning (issue4142) On Mac OS gcc-llvm throws an -Wtautological-compare warning because flen is defined as an unsigned integer, therefore flen < 0 is always true.

wix: pull in new templates

phase: properly compute ancestors of --rev on push (issue3786) Now that discovery is working on unfiltered changeset, I had a good occasion to look at that bug again. This let me realise that a trivial node vs rev comparision was the cause of this two years old bugs… Happy second birthday phases!

hgext: fix one typo in rebase

templater: selecting a style with no templates does not crash (issue4140) Running `hg log --style compact` (or any other style) raised a traceback when no template directory was there. Now there is a message: Abort: style 'compact' not found (available styles: no templates found, try `hg debuginstall` for more info) There is no test because this would require to rename the template directory. But this would influence other tests running in parallel. And when the test would be aborted the wrong named directory would remain, especially a problem when running with -l.

makefile: add $ to environment variables in help message

Added signature for changeset 8862469e16f9

Added tag 2.9-rc for changeset 8862469e16f9

merge with i18n

i18n: fix trailing '::' on translated messages Reported by FUJIWARA Katsunori <foozy@lares.dti.ne.jp>.

merge default into stable for 2.9 code freeze

tests: stop tracking tests/gpg/random_seed test-gpg.t left the random_seed file as modified. That was slightly confusing ... and it was accidentally changed in ea4996754d91. The seed is created on demand and there is no reason to track it. There is also no reason to leak state between test runs so we let the test clean up after running.

tests: 'hghave tic' also requires curses support in Python test-status-color.t would fail when using a Python without curses.

test-import-merge: mangle file in binary mode This should fix a test breakage on Windows.