tests: move a multi-statement debuglocks hook into a shell script for Windows Before this patch, a part of "test-push-hook-lock.t" fails unexpectedly on Windows environment, because semicolon (";") isn't recognized as the command separator by "cmd.exe". This is fixed the same way as a similar issue in 7c253c23de3b.

tests: fix globs for Windows test-largefiles-update.t, test-subrepo.t, test-tag.t, and test-rename-dir-merge.t still warn about no result returned because of unnecessary globs that test-check-code-hg.t wants, relating to output for pushing to, pulling from and moving X to Y.

run-tests: include quotes in the HGEDITOR value when storing sys.executable This fixes test-install.t on Windows that broke in 2122b82b6987 when shlex.split() was added to the debuginstall command: @@ -7,8 +7,11 @@ checking installed modules (*mercurial)... (glob) checking templates (*mercurial?templates)... (glob) checking commit editor... + Can't find editor 'c:\Python27\python.exe -c "(omitted)"' in PATH + (specify a commit editor in your configuration file) checking username... - no problems detected + 1 problems detected, please check your install! + [1] What happens is that shlex.split() on Windows turns this: c:\Python27\python.exe -c "import sys; sys.exit(0)" into this: ['c:Python27python.exe', '-c', 'import sys; sys.exit(0)'] While technically a regression, most programs on Windows live in some flavor of 'Program Files', and therefore the environment variable needs to contain quotes anyway to handle the space. This wasn't handled prior to the shlex() change, because it tested the whole environment variable to see if it was an executable, or split on the first space and tested again.

mq: when adding headers in plain mode, separate them from message (issue4453) c87f2a5a6e49 did a clean-up in one direction ... but we want it in the other direction.

mq: introduce insertplainheader - same naive implementation as before

mq: when setting message in plain mode, separate it from header (issue4453) Fix inconsistent handling of plain header separation in mq patcheader - and contrary to c87f2a5a6e49, do it in the direction of having an empty line between header and description. Plain patches are like mails and should thus have an empty line between headers and body in compliance with RFC 822 3.1.

setdiscovery: avoid a full changelog graph traversal We were definitely being suboptimal here: we were constructing two full sets, one with the full set of common nodes (i.e. a graph traversal) and one with all nodes. Then we subtract one set from the other. This whole process is O(commits) and causes discovery to be significantly slower than it should be. Instead, keep track of common incrementally and keep undecided as small as possible. This makes discovery massively faster on large repos: on one such repo, 'hg debugdiscovery' over SSH with one commit missing on the client and five on the server went from 4.5 seconds to 1.5. (An 'hg debugdiscovery' with no commits missing on the client, i.e. connection startup time, was 1.2 seconds.)

ancestor: add a way to remove ancestors of bases from a given set This and missingancestors can share state, which will turn out to be perfect for set discovery.

ancestor: add a way to add to bases of a missing ancestor object This will be useful for setdiscovery, since with that we incrementally add to our knowledge of common nodes.

ancestor: add a way to test whether a missing ancestor object has bases This is pretty trivial so there's no unit test coverage for it. This will be used by setdiscovery.

ancestor: remove now-unused missingancestors function Callers should use revlog.incrementalmissingrevs instead.

revlog: switch findmissing* methods to incrementalmissingrevs This will allow us to remove ancestor.missingancestors in an upcoming patch.

revlog: add a method to get missing revs incrementally This will turn out to be useful for discovery.

test-ancestor: add support for multiple tests against one incremental object In upcoming patches we'll add more operations to the object, and this prepares for testing those operations.

test-ancestor: move naive missing ancestor algorithm into a class This mirrors the change to the real missing ancestor algorithm in a previous patch.

ancestor.missingancestors: turn into a state-keeping class This allows multiple efficient missing ancestor queries against the same set of bases. In upcoming patches we'll also define ways to grow the set of bases. The fact that the test output hasn't changed establishes this patch's correctness.

ancestor.missingancestors: calculate start point after filtering revsvisit Any revs that are filtered out are also in basesvisit, which means they wouldn't be returned in the missing list anyway. There's no need to explore such revs or their ancestors. The 'if not revsvisit' check moves down because we can't call max() on an empty set.

ancestor.missingancestors: don't discard from basesvisit We only actually care about whether revsvisit is empty, so we can let basesvisit grow to arbitrary size. It turns out that this actually helps performance. For a large repo with hundreds of thousands of commits, hg perfrevset 'only(0, tip)' (basically the worst case, involving a full DAG traversal) goes from 1.63 seconds to 1.50. hg perfrevset 'only(tip, 0)' remains unchanged at 1.98 seconds.

test-ancestor: use random testing for missing ancestors We're going to make changes to the missing ancestor algorithm, and random testing will give us much more confidence than a fixed set of tests.

test-ancestor: define a main function We're going to add to it in upcoming patches.

test-ancestor: test iteration for lazyancestors This has some test coverage in test-revlog-ancestry.py, but not very much.

ancestor.lazyancestors: take parentrevs function rather than changelog Principle of least privilege, and it also brings this in line with missingancestors.

remove: avoid a bogus warning about no tracked files when removing '.' Previously, any files relative to the root of the repo that match the -I patterns would be deleted, but the command exited with 1 after printing a warning: $ hg remove -S -I 're:.*.txt' . removing sub1/sub2/folder/test.txt removing sub1/sub2/test.txt not removing .: no tracked files

remove: support remove with explicit paths in subrepos

remove: recurse into subrepositories with --subrepos/-S flag Like 'forget', git and svn subrepos are currently not supported. Unfortunately the name 'remove' is already used in the subrepo classes, so we break the convention of naming the subrepo function after the command.

cmdserver: protect pipe server streams against corruption caused by direct io Because pipe-mode server uses stdio as IPC channel, other modules should not touch stdio directly and use ui instead. However, this strategy is brittle because several Python functions read and write stdio implicitly. print 'hello' # should use ui.write() # => ch = 'h', size = 1701604463 'ello', data = '\n' This patch adds protection for such mistakes. Both stdio files and low-level file descriptors are redirected to /dev/null while command server uses them.

cmdserver: postpone creation of pipe server until run() This makes it easy to swap file descriptors while running command server.

cmdserver: use given streams as pipe channels like other commands Because commandserver itself is an hg subcommand, it shouldn't use stdio directly in principle.

revset.only: use cl.findmissingrevs ancestor.missingancestors is really an implementation detail.

manifestmerge: use already existing fl2 synonym for m2.flags(f) Probably not a noticeable performance gain, but shortens the code slightly.

merge: drop underscore prefix from _checkunknown() The method has been called from commands.py since 3eab42088be4 (update: just merge unknown file collisions, 2012-02-09), so drop the underscore prefix that suggests that it's private.

transaction: drop special handling for phases and bookmarks generation We are still doing double backups, but now that we have proper location handling this is less of an issue. Dropping this simplifies the code before we add some pending-related logic. This also ensures we actually test the new 'location' mechanism.

transaction: use 'location' instead of 'vfs' objects for file generation The argument is now a location name. The location must be present in the 'vfsmap' provided to the transaction at creation time.

transaction: use 'location' instead of 'vfs' in the addbackup method This unlock the backup of file outside of store (eg: bookmarks).

addbackup: handle file in subdirectory The current naming scheme ('journal.backups.<file>') resulted is bad directory name when 'file' was in a subdirectory. We now extract the directory name and create the backupfile within it. We plan to use file in a subdirectory for cachefile.

addbackup: use the vfs for the backup destination too The backup file location was always computed using the opener, bypassing the 'location' setting. (And making the feature broken.)

transaction: set backupentries version to proper value Now that all mechanisms are in place, we can advertise it with a proper new version.

transaction: support cache file in backupentries We do not want to abort if anything wrong happen while handling a cache file. Cache file have way to be invalidated and if old/bad version stay no misbehavior will happen. Proper value will eventually be computed and the wrong will be righten. This changeset use the transaction reporter (usually writing on stderr) to write details about failed cache handling. This will only apply to write operation using a transaction. The usual update during read only operation will stay a debug message. I was on the way to bring these message back to debug level when I realised it could be a feature. People with write access to the repository are likely to have the power to fix error related to cache (and it is valuable to fix them). So let the things as is for now.

transaction: use the location value when doing backup We finally use the 'location' value coupled with the 'vfsmap' to restore backup for the right file.

transaction: pass a vfs map to the transaction The goal is to allow access to file outside ofthe store directory from the transaction. The obvious target are the `bookmarks` file. But we can envision usage for cache too. We keep passing a main opener explicitly because a lot of code rely on this default opener. The main opener (operating on store) is using an empty key ''.

transaction: change the on disk format for backupentries We need to store new data to improve the current transaction logic: - location: We want to generate and backup file outside of the 'store' (eg: bookmarks, or various cache files). This requires knowing and preserving where each file is located. The value of this new field is a string. It will be used as a key for a vfs mapping. - cache: We would like to handle cache file in the transaction code. This Will help to have cache consistent with the repository state and avoid performance issue on big repository like Mozilla. However, failure to handle cache file should not result in a transaction failure. We add a new field that carry this information. The value is boolean, A True value mean any error while handling this file can be ignored. Those two mechanisms are not implemented yet, but they are now persisted in the on disk file. Support for new mechanisms is coming in later changeset. We update the file format now and will introduce the new features in later changeset. The format version is set to 0 until we actually support the new feature. This will prevent misunderstanding between incomplete and final client. Support for reading both version 1 and (future) version 2 could be achieved (using default value when reading version 1) but has not been seen as necessary for now.

silenttestrunner: add environment variable to make tests noisy again As I've been working on complicated extension code it's been handy to be able to get standard unittest verbose output so I can find crashers more efficiently.

largefiles: update comments to refer to the right overridden method This cleans up leftovers from b228ad1f79d7 (largefiles: override calculateupdates instead of manifestmerge, 2014-03-02).

revlog: cache chain info after calculating it for a rev (issue4452) This dumb cache works surprisingly well: on a repository with typical delta chains ~50k in length, unbundling a linear series of 5000 revisions (changelogs and manifests only) went from 60 seconds to 3.

manifest: add matches() method Move the code in context._manifestmatches() into a new manifest.matches(). It's a natural place for the code to live and it allows other callers to easily use it. It should also make it easier to optimize the new method in alternative implementations of the manifest (same reasoning as with manifest.diff()).

context.status: pass status tuple into _buildstatus By passing a status tuple (instead of the current list), we can access the status fields by name and make it a little more readable.

context.status: avoid de- and reconstructing status tuple We can just modify the status tuple we got from dirstate.status() instead of deconstructing it and constructing a new instance, thereby simplifying the code a little.

context.status: make _dirstatestatus() return an status tuple Letting _dirstatestatus() return an scmutil.status instance also means that _buildstatus() will always return such an instance, so we can remove the conversion from the call sites.