extensions: fix lookup of hgext.foo modules

localrepo: demand-load changeset, manifest, and dirstate

Merge with crew

extensions: fix load of module/__init__.py on OS X

localrepo.status: only append stuff to the clean list if list_clean is true This is consistent with dirstate.status.

mq: reduce matcher abuse The match function should match only the files we're interested in.

commands.remove: don't use the unknown list

commands.commit: don't use the unknown list

localrepo.add: avoid some stats There's still an extra lstat done by dirstate.update.

revert: use util.lexists

addremove: use util.lexists

commands.commit: symlinks are supported

mq: grab locks before starting a transaction

Merge with crew

extensions: load modules in module/__init__.py form. For example, convert=/path/to/convert now works.

test-patchbomb: sed -e may not work with extended regexps Non-GNU versions of the sed(1) utility (i.e. on Solaris or BSD) may not support extended regexps (i.e. including \| branches), or may need a special option to enable it (like -E instead of -e). Expand the sed RE in tests/test-patchbomb to fix running this test on FreeBSD and Solaris.

Correct tests/test-patchbomb* files: mv test-patchbomb test-patchbomb.out hg revert -r 1cf908c00479 test-patchbomb

patchbomb: add --date option

Don't validate email config if we're not sending email.

patchbomb: 0c61124ad877 moved setremoteconfig into cmdutil

dispatch: restore a dropped shlex import Add test-dispatch as a place for command parsing and dispatch tests.

merge with main

extensions: add missing sys import

Revive commands.dispatch as a simple way to call hg from other pythons scripts. This was already used by hg-ssh and others and is really convenient if you just want a native equivalent of calling the command line version.

dispatch: accept an abbreviation of --repository

dispatch: move findrepo to cmdutil

localrepo: don't search from path All current callers already pass in the repo root. This normalizes things a bit. Now all repo types take a direct path to repo root and only the command line interface (or the dispatcher) searches from the current directory.

dispatch: use the repo path found by early parsing This lets us simplify things a bit.

dispatch: parse and apply -R early This allows us to use extensions specified in .hg/hgrc. Unfortunately, this requires us to parse that file twice, but performance impact on the version command appears to be less than 1%.

dispatch: parse and apply cwd opt early

dispatch: rename variables after code motion

dispatch: hoist debugging hook to runcatch

dispatch: move dispatching code to cmdutil

dispatch: move runcommand to cmdutil

dispatch: pull final command execution into its own function

dispatch: reduce nesting by moving parsing exceptions

dispatch: move signal and exception handling to its own function

Create a separate module for managing extensions

extensions: separate function for loading single extensions

dispatch: reorder functions

localrepo: break out the repo-finding walk into its own function

Respect locale environment variables on darwin. In python 2.4+ on darwin, locale.getpreferredencoding() returns mac-roman regardless of what LC_CTYPE, LANG etc are set to. This can produce hard-to-notice conversion errors if input text is not in mac-roman. So this patch overrides it with setlocale/getlocale if the environment has been customized, on the assumption that the user has done so deliberately.

Small cleanups for the new tag code

gitweb: Display branch and tag labels Pages which display labels: - summary - shortlog - changelog - changeset - search

Merge with mpm

commit: fix bug where dirstate for removed file is confused This fixes issue586

hg serve: don't lookup user-supplied names/addresses

Split convert extension into common and repository type modules

Removed trailing \n in calls to util.Abort()

Some small cleanups for convert extension: - repo doesn't need to be imported - util.Abort doesn't need \n - space after comma - long lines - spacing

Avoid extra filelogs entries. Right now, there are some situations in which localrepo.filecommit can create filelog entries even though they're not needed. For example: - permissions for a file have changed; - qrefresh can create a filelog entry identical to its parent (see the added test); - convert-repo creates extra filelog entries in every merge where the first parent has added files (for example, changeset ebebe9577a1a of the kernel repo added extra filelog entries to files in the arch/blackfin directory, even though the merge should only touch the drivers/ata directory). This makes "hg log file" in a converted repo less useful than it could be, since it may mention many merges that don't actually touch that specific file. They all come from the same basic problem: localrepo.commit (through filecommit) creates new filelog entries for all files passed to it (except for some cases during a merge). Patch and test case provided by Benoit. This should fix issue351.

Make sure the changelog mentions files whose flags changed These changes don't actually need a new filelog entry (see next patch), but if we don't mention the files in the changelog, it becomes much harder to find all changesets that touch a file (not even a "hg log --removed file" will work).

run-tests.py: set ui.slash = True This should make things easier for pysh.

test-config-case: print only the Section we're interested in. This allows us to add additional settings to the global hgrc file created by run-tests.py without breaking this test.

Add ui.slash hgrc setting This will make most commands print paths using "/", regardless of the value of os.sep.

docopy: use repo.pathto to format paths for printing

Add dirstate.pathto and localrepo.pathto. Every time util.pathto is called, we have to pass the repo root and the repo cwd. dirstate.pathto is a simple convenience function that knows about the root and the cwd arguments. It's still possible to pass the cwd as an optimization. localrepo.pathto is a convenience function that just calls dirstate.pathto, just like localrepo.getcwd. dirstate.pathto becomes a single point that converts most (all?) paths from the internal representation to some OS-specific relative path for display purposes.

Use absolute paths in revert.

Use absolute paths in docopy.

Use absolute paths in addremove. This is more consistent with other places in the code, which only use the "relpath" returned by cmdutil.walk for display purposes.

Add some more smart when initializing destination repository

Fix hgext/convert/__init__.py rights

Include convert extension in packages setup (reported by Christian Ebert and Eric Hopper)

convert: map CVS HEAD to default branch

ui: make readsections() abort when configuration cannot be read.

Removed trailing whitespace and tabs from python files

Fixed python2.3 incompatibility (rsplit) in cvs code of convert extension

Add a wrapper script for convert extension

Turns convert.py into a real extension

Move convert-repo to hgext/convert/__init__.py

Fixed test output for changed glog (. instead of @)

Issue a warning if "-r ." is used with two working directory parents. Rationale for not aborting instead: The first parent is usually more important as it is the local branch during a merge and commands like 'hg diff' and 'hg diff -r.' behave still identically (except for the warning of course). Added a test for log -r. with one and two parents.

graphlog: Print . instead of @ for working directory parents "." is accepted as an alias for the first working directory parent by -r/--rev

Remove atomicfile class. The interface provided by opener(atomic=True) is inherently unsafe: if an exception is raised in the code using the atomic file, the possibly incomplete file will be renamed to its final destination, defeating the whole purpose of atomic files. To get around this, we would either need some bad hacks involving sys.exc_info (to make sure things work in except: blocks), or an interface to say "file is complete; rename it". This is the exact interface provided by atomictempfile. Since there are no remaining users of the atomicfile class, just remove it.

Use atomictemp files to write the dirstate. atomic files are not safe in all cases.

test-serve: let run-tests.py kill the daemons If the daemons die while the tests are still running, we'll get additional "killed!" messages, making the test fail.

Fix bug in "hg serve -v"

Add test reproducing a bug in "hg serve -v"

Merge with crew-stable

hgk: display tags that contain spaces

hgk: added -selectbackground grey to canvases

merge with stable

Obfuscate author in gitweb shortlog (issue578)

Fixed indentation in hgext/notify.py

merge with crew-stable

wsgicgi: change stdin to binary mode

Also check the index file size when deciding whether to reload a revlog. This should fix some spurious failures from test-clone-pull-corruption (issue552).

use .extend instead of .append in readtags

hbisect: fix a typo in error message

patch.diff: avoid calling workingctx().manifest() Right now, to generate the manifest of the working dir, we have to perform a full walk of the working dir, which will be very slow, especially if we're interested in only a small part of it. Since we use the manifest only to find out the mode of files for git patches, manually build an execf function to do it. This should fix issue567.

Merge with crew-stable

hgk - fix CR issues on windows

Add sending date to notify message. When using SMTP, no date field was set into the message, which causes it to be displayed as being send on 1st january 1970 on most MUA.

Merge with crew.

patchbomb: Fail early if no revs given to email

patchbomb: Fix typo.

patchbomb: Hoist sender config higher.

Automated merge with http://hg.intevation.org/mercurial/crew

patchbomb: Validate email config before we start prompting for info.

Introduce find_exe. Use instead of find_in_path for programs. The behaviour of find_in_path was broken for config options containing path names, because it always searched the given path, even when not necessary. The find_exe function is more polite: if the name passed to it contains a path component, it just returns it.

ui: Rename has_config to has_section.

patchbomb: Defer the import of readline. If imported at the top of the module, the import appears to succeed, but raw_input doesn't acquire magic editing fu. I suspect this has something to do with the newish demandimport code, because the prior code worked with demandload.

patchbomb: Don't prompt for headers until sure we have revs to export. The prior behaviour was to always prompt for headers, and only then bomb out if there were actually no revs to send.

Add support for url#id syntax This allows you to do: hg clone http://server/repo#stable which is equivalent to: hg clone -r stable http://server/repo Future incoming, outgoing, and push commands will default to using this id because it's recorded in the default path. Other commands that accept URLs (push, pull, bundle, incoming, and outgoing) also accept this syntax.

clone: checkout the "default" branch if available

mq: pass a URL to clone

Warn about large files on hg add

merge with stable

Make "hg incoming -r revision_number" work for remote repos (issue566) Uses the remote repository's lookup method as pull does and only transfers what is needed to calculate incoming changesets. One minor problem: As only the needed changesets are transfered and stored in a local bundle repository, the tip tag of this bundle is shows despite not being the tip changeset of the remote repository. (see "+tag: tip" in tests/test-incoming-outgoing.out in this patch)

Corrected similarity parameter used in test-addremove 0.5 meant 50% in the early stages of addremove -s, now 50 has to be used.

addremove: comparing two empty files caused ZeroDivisionError (found by Giorgos Keramidas)

Add archive download links to hgweb search template

merge with crew

gitweb: Fixed-up search template It seems that the search template has been forgoten over time, and needed major revamping.

gitweb: Add search box to the summary page

gitweb: file diff: Display metadata in same order as changeset does

gitweb: file annotate: Display metadata in same order as changeset does

gitweb: file rev template: Display metadata in same order as changeset does

gitweb: changeset template: remove manifest link from cset metadata

purge: add --include and --exclude options