hgext: replace uses of hasattr with util.safehasattr

globally: use safehasattr(x, '__iter__') instead of hasattr(x, '__iter__')

globally: use safehasattr(x, '__call__') instead of hasattr(x, '__call__')

safehasattr: new function to work around hasattr being broken

windows: check util.mainfrozen() instead of ad-hoc checks everywhere

notify: rewrite user documentation The main intent is to turn the reference help into a configuration walkthrough. It also fix several things: - Do not suggest to use it for commit notifications, it cannot work - Fix notify.strip default value - Mention that subscriptions can be setup in Mercurial configuration files - Improve notify.strip and notify.domain documentation

cmdserver: repo.invalidate() on every runcommand This will trigger the filecache and recreate every cached property that was changed by something other than this cmdserver instance (e.g. by running 'hg commit' at the cmdline).

help: add "web/logoimg" setting description

filecache: fix check-code complaint

localrepo: unify tag related info into a tagscache class

localrepo: make invalidate() walk _filecache

localrepo: decorate manifest() with filecache

localrepo: decorate _bookmarks/current with filecache nodebookmarks() can be cached as well, leaving for later

localrepo: decorate changelog() with filecache

localrepo: refresh filecache entries after releasing a repo.lock()

localrepo: decorate dirstate() with filecache We refresh the stat info when releasing repo.wlock(), right after writing it. Also, invalidate the dirstate by deleting its attribute. This will force a stat by the decorator that actually checks if anything changed, rather than reading it again every time. Note that prior to this, there was a single dirstate instance created for a localrepo. It was invalidated by calling dirstate.invalidated(), clearing its internal attributes. As a consequence, the following construct is no longer safe: ds = repo.dirstate # keep a reference to the repo's dirstate wlock = repo.wlock() try: ds.setparents(...) finally: wlock.release() # dirstate should be written here Since it's possible that the dirstate was modified between lines #1 and #2, therefore changes to the old dirstate won't get written when the lock releases, because a new instance was created by the decorator.

localrepo: add a cache with stat info for files under .hg/

scmutil: introduce filecache The idea is being able to associate a file with a property, and watch that file stat info for modifications when we decide it's important for it to be up-to-date. Once it changes, we recreate the object. On filesystems that can't uniquely identify a file, we always recreate. As a consequence, localrepo.invalidate() will become much less expensive in the case where nothing changed on-disk.

posix, windows: introduce cachestat This class contains a stat result, and possibly other file info to reliably determine between two points in time whether a file has changed. Uniquely identifying a file gives us that reliability because we either atomic rename or append. So one of two will happen: the file 'id' will change, or the size of the file will change. posix implements it simply by calling os.stat() and checking if the result has st_ino. For now on Windows we always assume the path is uncacheable. This can be improved on NTFS due to file IDs: http://msdn.microsoft.com/en-us/library/aa363788(v=vs.85).aspx So we need to find out if a file path is on an NTFS drive, for that we have: - GetVolumeInformation, which unfortunately only works with a root path (but is available on XP) - GetVolumeInformationByHandleW, works on a full file path but requires Vista or higher

util: eliminate wildcard imports

templater: use a global funcs table

ui: config path relative to repo root

ui: providing no default value to configpath should not raise an Error

ui: fix error, base can not be a list

merge with stable

test-i18n, i18n-pt_BR: updated test message changed in 41c3a71c318d

i18n-pt_BR: synchronized with a934b9249574

merge with stable

rebase: remove trailing whitespace found by check-code

hooks: use python 2.4 compatible exception handling

url: handle urls of the form file:///c:/foo/bar/ correctly

dispatch: avoid double backslashes in error message The use of %r in the format string caused Python to display Windows paths with double backslashes.

hgweb: add a "web/logoimg" setting to customize the web logo image This change complements the existing web/logourl setting, and lets the user customize the logo image that is shown on many of the hg server pages. If this setting is not set, hglogo.png is used.

util: move windows and posix wildcard imports to begin of file

util: move "default" hidewindow to posix.py There is a hidewindow in win32.py, which we get via windows.py

util: move "default" lookupreg to posix.py There is a lookupreg in win32.py, which we get via windows.py

util: move "default" unlinkpath to posix.py we have a unlinkpath in windows.py

util: move "default" makedir to posix.py makedir is already defined in win32.py, which gets imported into util.py via windows.py if we are running on Windows

summary: allow color to highlight active bookmark

summary: show bookmarks separate from tags and note active mark (issue2892)

localrepo: make requirements attribute of newly-created repos contain a set This is for internal consistency, as this attribute typically contains a set

localrepo: fix comment on set

revert: restore check for uncommitted merge (issue2915) (BC) This will restore the pre-1.9 behavior.

localrepo: add set method to iterate over a given revset This should allow replacing a number of hand-rolled graph algorithms.

revset: add formatspec convenience query builder

revset: allow bypassing alias expansion For internal usage of revset queries, we don't want aliases breaking things.

merge with stable

subrepo: don't commit in subrepo if it's clean If a subrepo has changed relative to the outer repo's substate, but the subrepo itself is clean, we don't need to commit in the subrepo.

rebase: block collapse with keepbranches on multiple named branches (issue2112) Collapse and keepbranches should be blocked when there is more than one named branch on the branch that's going to be rebased.

merge with stable

hbisect: do not assume that min(good) is an ancestor of min(bad) The included test used to report "inconsistent state", which is incorrect. While this situation cannot occur when the user sticks to the suggested bisect sequence. However, adding more consistent good/bad information to the bisect state should be tolerated as well.

hbisect: more consistent variable name

hbisect: confine loop to the relevant interval In this context we know that ancestors[rev] == None for all rev <= goodrev, so looping further back is unnecessary (and confusing).

win32: assign winstdout to sys.__stdout__ as well (issue2888) On Windows sys.stdout was being replaced with winstdout, which caused util.system() to redirect its output (due to 406b6d7bdcb9). That causes interactive tools (such as vim) to stop working.

alias: note interaction of shell aliases with early opts in help

alias: fix up test results

hooks: redirect stdout/err/in to the ui descriptors when calling python hooks We need to make sure that python hooks I/O goes through the ui descriptors so it doesn't mess the command server protocol.

alias: pass local ui to shell alias This avoids an additional config read

aliases: drop cwd manipulation Already done in _dispatch

dispatch: move shell alias handling after early arg handling --cwd, --repo, and --config obviously should be handled first

merge with stable

rebase: reset bookmarks (issue2265 and issue2873)

merge with stable

cmdserver: take repo.baseui as our ui The ui passed to server() is really repo.ui, that is it contains its local configuration as well. When running commands that use a different repo than the servers cached repo, we don't want to use that ui as the baseui for the new repo.

web: Output a correct date in short format (issue2902)

test-commandserver: explicitly close opened file

verify: fix mq misfire

merge with stable

merge with i18n

i18n-pt_BR: synchronized with 35f5cfdd0427

merge with stable

commit: suppress spurious new head message for duplicate commit (issue2893)

osutil: emulate os.listdir's OSError for long names (issue2898)

verify: filter messages about missing null manifests (issue2900)

merge with crew

subrepo: use working copy of .hgsub to filter status (issue2901)

notify: remove unused url() method

test-notify-changegroup: test notifications on unbundle

run-tests: fix summary when accepting changes interactively Accepted changes were not counted as success.

merge with stable

verify: filter messages about missing null manifests (issue2900)

cmdserver: restore old working dir after dispatch when we have --cwd

dispatch: don't use request repo if we have --cwd

eol: ignore IOError from deleted files in commitctx A Mercurial repo signals a file is deleted by raising IOError when the file's data is requested. This IOError is normally caught by localrepository.commitctx. With the eol extension enabled and EOL mappings in place, the eolrepo subclass should ignore IOError because a deleted file has no line endings to process. This issue exhibited itself when performing an incremental hg convert of a revision with deleted files to a repo with an existing .hgeol file.

scmutil: add missing import of re see https://bitbucket.org/tortoisehg/thg/issue/929

dispatch: fix checking of rpath in _getlocal This avoids an extra config read

ui: rename _is_trusted to _trusted is and do are pointless, underbar separators are style violations

merge with stable

eol: fix silly test-gendoc breakage by escaping control characters

eol: document new eol.fix-trailing-newline setting

eol: fix missing trailing newlines in comitted files Some text editors (Eclipse, for example) do not add trailing newlines, so diffs often contain annoying "\ No newline at the end of file". This patch to eol extension simply adds trailing newline on commit.

eol: eol.only-consistent can now be specified in .hgeol

merge with stable

commands: use mergetoolopts when a command supports --tool

revsets: actually catch type error on tip^p1(tip) (issue2884) The previous commit was empty.

hgcia: Set default value of strip to -1 (issue2891) For backward compatibility the root path of the repository is not displayed unless the user has specified an explicit strip value.

commands: improve help for -y/--noninteractive Before, the help text said that Mercurial would assume 'yes' for all prompts, but this is confusing since many prompts don't have any 'yes' choice. It now more accurately describes what will happen.

bookmarks: simplify warning code

bookmarks: drop superfluous strip

bookmarks: add a warning for non empty malformed line

bookmarks: more robust parsing of bookmarks file

merge with stable

revsets: catch type error on tip^p1(tip) (issue2884)

revsets: do the right thing with x^:y (issue2884) Given an operator ^ that's either postfix or infix and an operator : that's either prefix or infix, the parser can't figure out the right thing to do. So we rewrite the expression to be sensible in the optimizer.

win32mbcs: wrap two more functions to be wrapped. util.checkwinfilename() and util.checkosfilename() are wrapped. The former, used by scmutil, is not friendly for Shift_JIS bytes. The latter is an dynamic alias of checkwinfilename() used by other modules.

commands, merge: call setconfig on the right ui In practice this doesn't change anything since ui == repo.ui here, but setting repo.ui explicitly here is clearer since hg.merge will use repo.ui later on.

peer: change arg name to convey it can be a repo as well

progress: add a changedelay to prevent parallel topics from flapping (issue2698) When combined with the earlier change to make the progress object truly a singleton, this prevents the progress bar swapping on 'hg clone --pull' on a local filesystem. Thanks to timeless for lots of debugging help at the Copenhagen sprint to isolate the root cause of this and a first draft an idea that would fix it.

progress: make progress bar a singleton to avoid double-progress ui bugs This helps issue2698 a little, but isn't a complete fix, since generators still can produce some weird progress bar switching.

progress: remove superfluous parens

keyword: reuse already present working contexts for match Shortens overlong line as side-effect.

merge with stable

discovery: quiet note about heads This was changing output on in/out -v for no good reason.

patch: fix parsing patch files containing CRs not followed by LFs Since 1e64e1e12195 , patch lines containing a CR not followed by a LF would be incorrectly splitten, causing a failure to apply the patch.

treediscovery: rename stop() in tests to fix failures on AIX. It seems ksh, the default shell on AIX, does not permit the creation of a function called stop(). test-treediscovery.t and test-treediscovery-legacy.t both fail on AIX with error 'syntax error at line 25 : `(' unexpected'. Fix by renaming stop() in the scripts to tstop(). For completeness rename start() to tstart() to match. Both tests then pass on AIX. Add check for the use of stop() in a shell script to check-code.

fileset: fix typo in binary() doc

help: fileset foo.lst was named files.lst

merge with stable

Added signature for changeset de9eb6b1da4f

Added tag 1.9 for changeset de9eb6b1da4f

util: rename the util.localpath that uses url to urllocalpath (issue2875) util is never imported by any other name than util, so this is mostly just a simple search and replace from util.localpath to util.urllocalpath (assuming other uses of util.localpath already has been renamed).

tests: restore previous fix for date ambiguity in cvs test 017ab404e588 changed some sleeps after cvs update to sleeps before commit. I don't know why that should work, but I have a machine where it doesn't work. Commits regularly fails becuase cvs doesn't notice that a file has been changed. Restoring the previous sleep between the update and the edit makes cvs notice the change.

merge with stable

run-test: revert most of 439ed4721a6d, timeout uses fallback for terminate()

run-tests: fallback to SIGTERM if subprocess.Popen does not have terminate()

subrepo: correct revision in svn checkout A Subversion subrepo checkout uses a url and --revision which does not do the correct thing when specifying a revision of a branch that has since been deleted and recreated. The checkout needs to specify the revision as URL@REV instead.

wix: catch up with more added files revsets: 1.7::tip and adds("mercurial/help/**") 1.7::tip and adds("mercurial/templates/**")

i18n-pt_BR: translate template filter descriptions

i18n-ja: recover template translations from 15200b46165b These translations were removed by 68a697421a57 because the original strings were transfered to the docstrings (264f292a0c6f) without the corresponding sources being added to the Makefile for extraction.

i18n-ru: fixed mixed up msgid and msgstr in annotate

i18n-ru: translated missing strings for commands, some typos corrected

i18n-pt_BR: synchronized with 9adce4b38ed1

merge with stable

merge with i18n

i18n-ja: synchronized with 5c046e4b9652

i18n-ru: translated help for urls. Thus, all stock commands and help translated

i18n-ru: translated help for templating

i18n-sv: synchronized with 85857aa6e666

i18n-pt_BR: synchronized with 6ba51c81ff75

i18n-ru: translated help for multirevs-subrepos

i18n-da: translate hg.1.html header

i18n: merge with stable

i18n-ru: translated help for merge-tools

i18n-ru: translated help for hgignore

i18n-pt_BR: add lost translation used in test-i18n.t Changeset e6a6e7ebd585 replaced the Portuguese translation with the English original and broke the test.

i18n: merge with main

i18n: merge with main

merge with stable

i18n-ru: translated help for merge-tools

i18n-ru: translated help for hgignore

i18n-pt_BR: synchronized with 72e4fcb43227

i18n-ru: drop executable bit from .po file

merge with stable

i18n: merge with main This includes four i18n changesets transplanted from the default branch to the stable branch since they were made after the freeze.

i18n-ru: glossary translated

i18n-ru: tranlated docstrings from filesets added in 1575dc5d399a

i18n-ru: translated help for extensions and filesets

i18n-ru: translated new config.txt

i18n: merge with main

i18n-pt_BR: synchronized with 94eea58da2a3

fileset: add i18n hints for keywords

merge with i18n

merge with i18n

18n-sv: translated all new fuzzy strings

i18n-ru: glossary translated

i18n-ru: tranlated docstrings from filesets added in 1575dc5d399a

i18n-ru: translated help for extensions and filesets

merge with i18n

i18n-pt_BR: synchronized with 6c7283faa967

merge with i18n

i18n-ru: translated new config.txt

i18n-pt_BR: messages from crew f9c056f48018

i18n-pt_BR: synchronized with 15200b46165b

wix: add missing templates/paper/diffstat.tmpl

hgweb: treat branch attribute `closed' as more important than `inactive' In the branches page, branches that are closed and are merged into another branch are displayed as `inactive'. This patch changes that behaviour to show these branches as `closed'. For me, the `closed' attribute is more important than the `inactive' attribute. Branches that are not closed, and are merged into other branches will still be shown as `inactive'. Branches that are closed, and are not merged into other branches will still be shown as `closed'.

tests: add basic commandserver test

color: for the sake of "less -R", default to ansi in auto mode (issue2792) less with the -R option only supports ANSI color codes, and terminfo mode issue non-ANSI color codes (depending on the database). The -r option can work around this, but there are users currently using the -R option, and defaulting to terminfo would break colorization in less for them without warning.

color: fix TypeError with auto mode on win32 when colors aren't available (issue2871) This can happen when stdout isn't a terminal (e.g., it's being piped).

test-revert: add revert --all case when nothing changed supposed to succeed without notice

subrepos: be smarter about what's an absolute path (issue2808)

test-wireprotocol.py: rename to test-wireproto.py for consistency 1. The module named 'wireproto' is tested. 2. There already is a test-wireproto.t for related tests on shell level.

classes: fix class style problems found by b071cd58af50 This makes test-wireprotocol.py work on Python 2.4

check-code: fix class style checking (with tests) - old-style classes were only checked for one-letter class names - add check for new-style classes with empty parent class, because this is not available in Python 2.4

tests: ignore inotify extension in test-duplicateoptions.py The inotify extension is only available on linux and setup.py will not install it on other platforms - but it will of course always be there in the source. test-duplicateoptions.py tried to load most extensions (including inotify if available). When the local uninstalled Mercurial was used it would thus always load the inotify extension and cause a warning on unsupported platforms. The inotify extension is not relevant for this test, so now we explicitly ignore it.

error: Add a hint argument to RepoError This use the same mechanism than Abort. Except clause stay distinct because RepoError add "!" at the end of the message.

test-revert: add revert --all case when nothing changed supposed to succeed without notice

merge with stable

color: be silent when falling back to ECMA-48 mode if "auto" mode was desired Don't warn about missing setab/setaf terminfo entries unless we explicitly asked for terminfo mode; fallback to ECMA-48 mode silently.

tests: hardcode terminfo entry for the terminfo color test For the terminfo color test, make sure that the terminfo entry used is one of our own choosing, by delivering a special "hgterm" entry (a copy of ncurses' xterm-color), compiling it, and specifically pointing curses to it using the TERMINFO and TERM environment variables. This means we can ignore the variability in different terminal definitions on different platforms.

localrepo: mark progress step for translation

revert: improve hints on abort when reverting to parent without --all BEFORE: $ hg revert abort: no files or directories specified (use --all to discard all changes) AFTER: Uncommitted changes (using --all *will* nuke edits): $ hg revert abort: no files or directories specified (uncommitted changes, use --all to discard all changes) Clean working directory (using --all won't discard anything): $ hg revert abort: no files or directories specified (use --all to revert all files)

dispatch: set config options on the request repo.ui

dispatch: return read config options

dispatch: set global options on the request repo.ui If the request has a repo, and global options such as --verbose are passed, we need to set those explicitly on the repo.ui.

cmdserver: assign repo.baseui before running commands There are places in the code that use localrepository.baseui (see hg.remoteui), we need the ui descriptors (and possibly other things) to be set correctly on it, so output written to the remoteui descriptors ends up at the right place. Before this change, tests such as 'test-bookmarks-pushpull.t' didn't work.

cmdserver: copy repo.ui before running commands

filemerge: use ui out descriptor when calling util.system

dispatch: make sure unspecified global ui options don't override old values

merge with stable

requirements: show all missing features in the error message. Displaying all missing featureis help people to solve the issue (choosing the right version, creation the right repo)

test-revert: add case for nothing changed

dispatch: set descriptors on the request repo.ui

dispatch: pass the correct ui to runcommand For commands that take a repo, repo.ui is passed as the ui to runcommand. This restores that behaviour when the request has a repo.

archive: wrap the ui descriptor so it doesn't get closed archival.tarit closes the fileobj if one is passed to it, which is undesired when reading from '-'.

transplant: use ui out descriptor when calling util.system

bisect: use ui out descriptor when calling util.system

extdiff: use ui out descriptor when calling util.system

ui: use ui out descriptor when calling util.system

call reposetup for repositories returned from peer (issue2864) Extension setup functions were not being called when repositories were obtained using the peer function instead of the repository function.

tests: solaris [ doesn't know -e

convert: fix git convert on solaris - it cannot remove environment variables

tests: HGRCPATH must be exported after assignment - at least on Solaris

merge with stable

backout of d04ba50e104d: allow to qpop/push with a dirty working copy The new behavior was breaking existing tools that relied on a sequence such as this: 1) start with a dirty working copy 2) qimport some patch 3) try to qpush it 4) old behavior would fail at this point due to outstanding changes. (new behavior would only fail if the outstanding changes and the patches changes intersect) 5) innocent user qrefreshes, gets his local changes in the imported patch It's worth considering if we can move this behavior to -f in the future.

merge with stable

merge with stable

update: do not use the term 'update' when mentioning reverting one file and give a more precise hint for how to revert such a file I'm using the term 'revision' instead of 'changeset' in this change to be consistent with the REV we use in the synopsis.

cmdserver: fix read-loop string concatenation

dispatch: check for None before closing repo We were trying to call close() if repo == None and req.repo != None. This can happen when running commands that don't take a repo.

revert: mention update in hint of abort when reverting to non-parent and explicitly warn about uncommitted changes Examples: BEFORE: $ hg par -q 7:e81a2efd53d4 $ hg revert -r 2 abort: no files or directories specified (use --all to discard all changes) AFTER: Clean working directory (revert can be easily undone, no edits to be lost): $ hg revert -r 2 abort: no files or directories specified (use --all to revert all files, or 'hg update 2' to update) Uncommitted changes (revert --all *does* discard edits and is pretty hard to undo or even impossible if --no-backup is specified): $ hg revert -r 2 abort: no files or directories specified (uncommitted changes, use --all to discard all changes, or 'hg update 2' to update)

update: clarify that update changes the parent revison

merge with stable

revset: fix aliases with 0 or more than 2 parameters The existing code seemed to have incorrect assumptions about how parameter lists are represented by the parser. Now the match and replace functions have been merged and simplified by using getlist().

match: fix bug caused by refactoring in cfc89398f710

revert: be more helpful on uncommitted merges BEFORE: $ hg revert abort: no files or directories specified (use --all to discard all changes) AFTER: $ hg revert abort: no files or directories specified (uncommitted merge, use --all to discard all changes, or 'hg update -C .' to abort the merge)

opener: add self._audit (issue2862)

cmdserver: write the hello message as one chunk on the 'o' channel This is a guaranteed by the protocol: clients know they need to read one chunk off of the 'o' channel and treat that as the hello message. They should ignore fields they don't recognize so they stay compatible with new versions of the server in case we decide to add something.

fileset: copied takes no arguments

revset and fileset: fix typos in parser error messages

fileset: use ParseError pos field correctly The pos field is intended to describe the position of the error - it should not be used for other potentially interesting information.

revset: fix parameter name in implementation of follow()

merge with stable

revert: be more explicit that changes are lost BEFORE: $ hg revert abort: no files or directories specified (use --all to revert all files) AFTER: $ hg revert abort: no files or directories specified (use --all to discard all changes)

dispatch: do not close the request repo Closing here means we've closed the repo passed to us in the request, which is not our responsibility. This is essential for bundlerepo, and possibly other localrepository subclasses who do something in their close().

hook: write hook output to ui fout descriptor output on stderr will also be written to ui.fout, unless sys.__stdout__ is passed in (see util.system), thus not changing previous behavior. This fixes a bug where hooks run through the command server would mess up with the command protocol, sending non-channeled data to the client.

merge with stable

check-code: don't mark debug messages for translation

i18n: remove translation of debug messages

cmdserver, runcommand: properly handle the client sending no arguments No real reason for a client to do this, but still possible. Previously if the client sent no arguments, a list with an empty string [''] would be used as the arguments to dispatch, which would cause hg to complain about an ambiguous command. Instead, we simply check for no arguments and use an empty list instead (which is equivalent to invoking hg with no args on the command line).

cmdserver: don't raise EOFError when trying to read 0 bytes from the client