perf: add option to perfstatus to get the status of unknown files When status needs to look at unknown files (e.g. when running hg status), it needs to use a completely different algorithm than when it doesn't (e.g. when running hg diff).

dirstate: test normalize is truthy instead of using a no-op lambda hg perfstatus -u on a working directory with 170,000 files, without this change: ! wall 1.869404 comb 1.850000 user 1.170000 sys 0.680000 (best of 6) With this change: ! wall 1.839561 comb 1.830000 user 1.120000 sys 0.710000 (best of 6)

merge with stable

hgweb: avoid generator exhaustion with branches

hgweb: fix iterator reuse in atom feed generation

tests: don't hardcode errno==2 for ENOENT Hurd seems to set ENOENT to 2 + 2**30, unlike everyone else.

osutil: tab damage, how i hate thee

osutil: write a C implementation of statfiles for unix This makes a big difference to performance. In a clean working directory containing 170,000 files, performance of "hg --time diff" improves from 2.38 seconds to 1.69.

merge with i18n

i18n-ja: synchronized with f94ead934067

Added signature for changeset 0c10cf819146

Added tag 2.4.1 for changeset 0c10cf819146

osutil: fix tab damage

Merge with crew

osutil: factor out creation and init of listdir_stat

dirstate: avoid use of zip on big lists In a clean working directory containing 170,000 tracked files, this improves performance of "hg --time diff" from 1.69 seconds to 1.43. This idea is due to Siddharth Agarwal.

dirstate: move file type filtering to its source This prepares us to move to a much faster statfiles implementation on Unix.

clfilter: rename `unfilteredmeth` to `unfilteredmethod` As originally intended.

clfilter: fix a false positive in the test-obsolete.t We push between two repo which once filtered looks unrelated. Weakness in the current implementation allows this push to be done without -f. But later improvement with filtering will make this push fails for unrelatedness. However we want this push to fail for including bumped changeset. So we had a smaller push --force to make them related.

clfilter: ensure that filecache on localrepo is unfiltered All filecache usage on repo is for logic that should be unfiltered. The caches should be common to all filtered instances, and computation must be done unfiltered. A dedicated storecache subclass is created for this purpose.

clfilter: add a propertycache that must be unfiltered Some of the localrepo property caches must be computed unfiltered and stored globally. Some others must see the filtered version and store data relative to the current filtering. This changeset introduces two classes `unfilteredpropertycache` and `filteredpropertycache` for this purpose. A new function `hasunfilteredcache` is introduced for unambiguous checking for cached values on unfiltered repos. A few tweaks are made to the property cache class to allow overriding the way the computed value is stored on the object. Some logic relative to _tagcaches is cleaned up in the process.

largefile: status is buggy on repoproxy, so run unfiltered For some yet-unkown reason, largefile status does not work on repoproxy. As status is not affected by filtering, we run it unfiltered. Na'Tosha Bard's view on this issue: "but, well, largefiles status is kind of an unholy piece of code"

clfilter: mq should not warn about filtered mq patches MQ warns when qstatus contains unknown nodes. With changelog filtering, a node may be unknown because it is filtered. Thus, an unfiltered repo is used for this check.

clfilter: ensure that mq performs commits on unfiltered repos

clfilter: use unfiltered repo for bookmark push logic The remote location of the bookmark may be filtered locally. This changeset ensures that bookmark movement logic has access to all the repo's content.

clfilter: `bookmark.validdest` should run on unfiltered repo The logic recently added to `bookmark.validdest` uses data about obsolete changesets to see if a bookmark destination is valid. Obsolete changesets are likely to be filtered, so we need to work on an unfiltered repository.

clfilter: unfilter some parts of the push logic Computation of common changesets during push needs to be done on the widest set possible. An unfiltered version of the repo is kept for discovery and various revset calls. The discovery code itself enforces the filtering of unserved outgoing changeset.

subrepo: add argument to "diff()" to pass "ui" of caller side (issue3712) (API) Color extension achieves colorization by overriding the class of "ui" object just before command execution. Before this patch, "diff()" of abstractsubrepo and classes derived from it has no "ui" argument, so "diff()" of hgsubrepo uses "self._repo.ui" to invoke "cmdutil.diffordiffstat()". For separation of configuration between repositories, revision 573bec4ab7ba changed the initialization source of "self._repo.ui" from "ui"(overridden) to "baseui"(plain) of parent repository. And this caused break of colorization. This patch adds "ui" argument to "diff()" of abstractsubrepo and classes derived from it to pass "ui" object of caller side.

clfilter: prevent unwanted warning about filtered parents as unknown During changectx __init__ the dirstate's parents MAY be checked. If the repo is filtered, this check will complain "working directory has unknown parents" even if the parents are perfectly known. This may happen when the repo is used for serving and the dirstate has parents that are secret, as those secret changesets will be filtered.

clfilter: strip logic should be unfiltered Strip is a "write" operation that needs to be aware of the whole repo's content before destroying changesets. Only the low level function is altered. The top level command will still process its argument filtered (if any filtering is in place).

clfilter: verify logic should be unfiltered To verify a changelog obviously needs all of it. The verify logic now ensures it works on an unfiltered repository.

clfilter: phases logic should be unfiltered Phase computations and boundary movements need to be aware of all revisions that exist in the repository to return correct results.

clfilter: unfilter computation of obsolescence related computation All obsolescence related sets need to be computed on the full unfiltered version of the repository, in particular because several of them (obsolete, extinct) are used to compute the hidden revisions. On a filtered repo, revset predicates related to these sets will be properly filtered because of revset's own pre-filtering.

clfilter: ensure changeset creation in the repo is run unfiltered This applies to both creation of new commits and application of changegroups.

clfilter: ensure changegroup generation is run unfiltered Changegroup generation logic needs to be aware of the whole repository to work properly.

clfilter: ensure `rollback` is run unfiltered Rollback logic needs to be aware of the whole repository to work properly.

clfilter: ensure cache invalidation is done on the main unfiltered repo The proxy version will not hold any cache for now. But we have to ensure all cache operations are done on the unfiltered version.

clfilter: ensure that tag logic runs unfiltered The current tag logic is not aware of filtering. We keep the status quo, ensuring that the tag cache is computed as before: without any filtering.

clfilter: ensure `branchcache` logic runs unfiltered The current branchcache construction is not aware of filtering. We keep the status quo, ensuring that the branch cache logic is computed as before: without any filtering.

clfilter: introduce an `unfilteredmethod` decorator This decorator ensure the method in run on an unfiltered version of the repository. See follow-up commit for details. This decorator is not named `unfiltered` because it would clash with the `unfilteredmethod` on `localrepo` itself.

clfilter: introduce an "unfiltered" method on localrepo This commit is part of the changelog level filtering effort. It returns the main "unfiltered" version of a repo-like object. For localrepo this means the same localrepo object. But this method will be overwritten by the filtered versions of a repository to return the core unfiltered version of the repo. Introducing this simple method first allows later commits to prepare for the use of a filtered version of a repository. A new repo method is added because a lot of users may call it. At the end of this series of commits, about 40 calls exist in core and hgext.

clfilter: remove usage of `range` and `xrange` in scmutil.revrange For changelog level filtering to take effect it need to be used for any iteration. This changeset removes usage of `range` and `xrange` that survived the first pass.

hgweb: display diff for a changeset against any parents (issue2810) During merge of branches, it is useful to compare merge results against the two parents. This change adds this support to hgweb. To specify which parent to compare to, use rev/12300:12345 where 12300 is a parent changeset number. Two links are added to changeset web page so that one can choose which parent to compare to.

branch: add missing repo argument to checknewlabel scmutil.checknewlabel takes a repo object as its first argument. When the call to this function was added in e689b0d91546, the first argument was mistakenly set to 'None'.

Merge with stable.

rebase: fix pull --rev options clashing with --rebase (issue3619) Rebase also have a plain `--rev` option used to select the rebase set (as `--base` or `--source` would). But the content of the --rev option was intended for the remote repo and is irrelevant for the local rebase operation. We expect `hg pull --rebase` to stick with the default behavior here: hg rebase --base . --dest tip(branch(.)) The `rev` option is dropped from the option passed to rebase.

dirstate: remove obsolete comment from setbranch This comment should have been removed in e689b0d91546, when the call to scmutil.checknewlabel was removed.

merge with stable

tests: fix broken fix of test-phases output

update: allow update to existing branches with invalid names (issue3710) Starting with 361ab1e2086f, users are no longer able to update a working copy to a branch named with a "bad" character (such as ':'). Prior to v2.4, it was possible to create branch names using "bad" characters, so this breaks backwards compatibility. Mercurial must allow users to update to existing branches with bad names. However, it should continue to prevent the creation of new branches with bad names. A test was added to confirm that 'hg update' works as expected. The test uses a bundled repo that was created with an earlier version of Mercurial.

commit: increase perf by building a new addlist instead of editing the old one When commiting to a repo with lots of files (>170000), manifest.py:addlistdelta takes some time because it's editing a large array many times. Changing it to build a new array instead of editing the old one saves around 0.04 seconds on a 1.64 second commit. A 2.5% gain. The gain here is pretty minor, but it was blatantly at the top of the profiler report and the fix is straight forward. I tested it by comparing the arrays produced by the new and old logic while running all of the tests.

template engine: convert generator-based iterator to list-based iterator If a template iterator is implemented with generator, the iterator is exhau= sted after we use it. This leads to undesired behavior in template. This chang= e converts a generator-based iterator to list-based iterator when template en= gine first detects a generator-based iterator. All future usages of iterator wi= ll use list instead.

command: remove phase from the list of basic command This is not a basic command. There is no reason new user should needs to know about it. Thanks to Matt Mackall for pointing this.

merge with stable

phases: fix missing "error" module import (issue3707)

test-mq-qrefresh: test that the patch at qtip is indeed empty ...after qrefreshing a non-existent file

tests: only call check-code once The accepted warnings list is now empty, let's try to keep it that way.

ancestor: fix a comment (followup to 0b03454abae7)

revlog: allow reverse iteration with revlog.revs We often need to perform rev iteration in reverse order. This changeset makes it possible to do so, in order to avoid costly reverse or reversed() calls later.

convert: add config option to use the local time zone The default for the time zone offset in a converted changeset has always been 0 (UTC). With this patch, the converted changeset is modified so that the local offset from UTC is specified as the time zone offset. The option is specified as the boolean convert.localtimezone (default False). Example usage: hg convert -s cvs --config convert.localtimezone=True example-cvs example-hg IMPORTANT: the patch only applies to conversions from cvs or svn. The documentation for the option only appears in those two sections in the convert help text.

rebase: use revlog.findmissingrevs to compute detach set

revlog: add rev-specific variant of findmissing This will be used by rebase in an upcoming commit.

revlog: switch findmissing to use ancestor.missingancestors This also speeds up other commands that use findmissing, like incoming and merge --preview. With a large linear repository (>400000 commits) and with one incoming changeset, incoming is sped up from around 4-4.5 seconds to under 3.

ancestor: faster algorithm for difference of ancestor sets One of the major reasons rebase is slow in large repositories is the computation of the detach set: the set of ancestors of the changesets to rebase not in the destination parent. This is currently done via a revset that does two walks all the way to the root of the DAG. Instead of doing that, to find ancestors of a set <revs> not in another set <common> we walk up the tree in reverse revision number order, maintaining sets of nodes visited from <revs>, <common> or both. For the common case where the sets are close both topologically and in revision number (relative to repository size), this has been found to speed up rebase by around 15-20%. When the nodes are farther apart and the DAG is highly branching, it is harder to say which would win. Here's how long computing the detach set takes in a linear repository with over 400000 changesets, rebasing near tip: Rebasing across 4 changesets Revset method: 2.2s New algorithm: 0.00015s Rebasing across 250 changesets Revset method: 2.2s New algorithm: 0.00069s Rebasing across 10000 changesets Revset method: 2.4s New algorithm: 0.019s

bisect: add example for limiting bisection to specified directories The bisect command does not have an option to limit itself only to subdirectories, but it's possible to use revsets for the --skip option for the same effect. Given the relative obscurity of revsets, it helps to have this as another example for bisect.

subrepo: use posixpath when diffing, for consistent paths This fixes a Windows failure in test-subrepo-recursion.t introduced by c84ef0047a94.

run-tests: fix an unnoticed check-code violation

run-tests: add a --compiler option Without this option, it is not possible to run the test suite on Windows using mingw's gcc as the compiler.

run-tests: make build command line less intimidating Use a dict for parameters to the format string, instead of a ridiculous number of positional parameters.

hooks: be even more forgiven of non-fd descriptors (issue3711) Looks like there are instances where sys.stdout/stderr contain file handles that are invalid. We should be tolerant of this for hook I/O redirection, as our primary concern is not garbling our own output stream.

hooks: delay I/O redirection until we actually run a hook (issue3711) We were attempting to redirect I/O even if no hook was actually getting called. This defers redirection until we've found something to do.

util: make chunkbuffer non-quadratic on Windows The old str-based += collector performed very nicely on Linux, but turns out to be quadratically expensive on Windows, causing chunkbuffer to dominate in profiles. This list-based version has been measured to significantly improve performance with large chunks on Windows, with negligible overall overhead on Linux (though microbenchmarks show it to be about 50% slower). This may increase memory overhead where += didn't behave quadratically. If we want to gather up 1G of data to join, we temporarily have 1G in our list and 1G in our string.

revset: backed out changeset 54cedee86e51 This was causing clones of the hg repo to go from 12.4s to 14.7s.

hgk: specify some colours explicitly in hex Tk 8.6b3 uses web colours, where green is defined as #008000, not #00ff00, and grey is #c0c0c0, not #808080, so specify those colours explicitly.

hgk: no committer please Generate committer only if we really have it. Update test-hgk.t accordingly.

hgk: use Ttk instead of plain Tk Use Ttk (themed Tk) for most of the widgets. Default to xpnative theme on Windows, clam otherwise. Provide a shim for Tk 8.4 without Tile/Ttk.

check-code: move i18n check from warning to error

i18n: wrap false positives for translation detection

eol: don't refer to a random name-captured ui object Previously, we used a ui object captured at the time of monkeypatching.

mq: don't refer to a random name-captured repo object Previously, we used a repo object captured at the time of monkeypatching.

Merge with crew-stable

commit: increase perf by avoiding checks against entire repo subsets When commiting to a repo with lots of history (>400000 changesets) checking the results of revset.py:descendants against the subset takes some time. Since the subset equals the entire changelog, the check isn't necessary. Avoiding it in that case saves 0.1 seconds off of a 1.78 second commit. A 6% gain. We use the length of the subset to determine if it is the entire repo. There is precedence for this in revset.py:stringset.

commit: increase perf by avoiding unnecessary filteredrevs check When commiting to a repo with lots of history (>400000 changesets) the filteredrevs check (added with 5c89e7fa5bc2) in changelog.py takes a bit of time even if the filteredrevs set is empty. Skipping the check in that case shaves 0.36 seconds off a 2.14 second commit. A 17% gain.

merge with crew-stable

grep: remove useless while condition A revised version of 35ba170c0f82 was sent to the list that fixed this <http://markmail.org/message/jmfuiise5igcyh2m>, but the older version of the patch was applied.

merge with mpm

test-pathencode: more aggressively check for python < 2.6

diff: move index header generation to patch In an upcoming patch, we will add index information to all git diffs, not only binary diffs, so this code needs to be moved to a more appropriate place. Also, since this information is used for patch headers, it makes more sense to be in the patch module, along with other patch-related metadata.

patch: make _addmodehdr a function under trydiff addmodehdr is a header helper, same as diffline, so it doesn't need to be a top-level function and can be nested under trydiff. In upcoming patches we will generalize this approach for all headers.

diff: rewrite diffline Make diffline more readable, using strings with placeholders rather than appending to a list from many ifs that makes difficult to understand the actual output format.

diff: swap and simplify diffline args Args order swapped, since a, b are more important than revs (which is only used on non-git format), and change to read opts from context.

diff: change how quiet mode supresses diffline Before, quiet mode produced no diffline header for mercurial as a side effect of not populating "revs". This was a weird side effect, and we will always need revs for git index header that will be added in upcoming patches, so now we just check ui.quiet from diffline directly.

diff: move diffline to patch module diffline is not part of diff computation, so it makes more sense to place it with other header generation in patch module. In upcoming patches we will generalize this approach for all headers added in the patch, including the git index header.

diff: unify calls to diffline diffline was called from trydiff for binary diffs and from unidiff for text diffs. In this patch we unify those calls into one. diffline is also a header, not part of diff mechanisms, so it makes sense to remove that responsibility from the mdiff module. In upcoming patches we will move diffline to patch module and keep grouping responsibilities.

diff: move b85diff to mdiff module b85diff generates a binary diff, so we move this code to mdiff module along with unidiff for text diffs. All diffing mechanisms will be in the same place. In an upcoming patch we will remove the responsibility to print the index header from b85diff and move it back to patch, since it's a patch metadata header, not part of the diff generation.

share: always set default path to work with subrepos (issue3518) set the default path in any case because creating subrepo looks this up. Subrepos are cloned, not shared.

vfs: optimize __call__ by not calling util.split for reads dirname and basename are not used on reads

run-tests: backout 4a4173519b63 This deleted work in progress to eliminate child processes for -j.

test-pathencode: make a 2.4-safe import of collections

tests: add a randomized test for pathencode This is a probabilistic test - it generates different test cases on every run, unless invoked from the command line with a specific seed. The default number of tests run should make the tests take about a second to complete on a semi-modern laptop.

webcommands: allow hgweb's archive to recurse into subrepos Currently when obtaining an archive snapshot of a repository via the web interface, subrepositories are not taken in the snapshot. I introduce an option, archivesubrepos, which allows this.

Merge with crew-stable.

grep: don't search past the end of the searched string '*' causes the resulting RE to match 0 or more repetitions of the preceding RE: >>> bool(re.search('.*', '')) >>> True This causes an infinite loop because currently we're only checking if there was a match without looking at where we are in the searched string.

convert: fix a too-long line nag

convert: add support for converting git submodule (issue3528) Previously, convert aborted upon encountering a git submodule. This patch changes it so that it now succeeds. It modifies convert_git to manually generate '.hgsub' and '.hgsubstate' files for each git revision, so as to convert git sub modules to non-mercurial subrepositories.

run-tests.py: remove runqueue's results parameter since it is now a global

makefile: don't use system hgrc when running hg in-place This suppresses the printing of spurious error messages if a global hgrc refers to extensions that the in-place hg can't access.

setup: print subprocess stderr if there is any I just spent 1.5 days trying to debug a failing buildbot because setup.py was silently dropping the errors that were being printed by in-place hg.

run-tests: open child-parent pipes in binary mode Python's pickle is a binary format.

amend: force editor only if old message is reused (issue3698) This regression was added by 9732473aa24b. It triggered the editor even if --message or --logfile were provided.

grep: don't search past the end of the searched string '*' causes the resulting RE to match 0 or more repetitions of the preceding RE: >>> bool(re.search('.*', '')) >>> True This causes an infinite loop because currently we're only checking if there was a match without looking at where we are in the searched string.

bookmarks: introduce a bmstore to manage bookmark persistence Bookmarks persistence still showed a fair amount of its legacy as a monkeypatching extension. This encapsulates all bookmarks serialization and parsing in a single class, and offers a single location where other bookmarks storage engines can be substituted in. As a result, many files no longer import the bookmarks module, which strikes me as an encapsulation win. This doesn't do anything to the current bookmark state yet, but I'm hoping put that in the bmstore class as well.

run-tests: add --time option to log times for each test --time also prints out the wall-clock time each test takes in descending order.

run-tests: remove resultslock since it serves no useful purpose Each child process has its own copy of the results dict, so all access to the results dict is serial.

run-tests: use pickle to send results from children to parent JSON would probably be preferable, but it isn't available for Python 2.4 or 2.5.

bookmark: remove useless line in `validdest` This case is already covered by the `if old == new:` conditional at the start of the function.

bookmark: issue a single call to `allsuccessors` per loop Update to this code was minimalist when `allsuccessors` argument were changed from a list to a set. As this code is getting my attention again I realised we can drastically simplify this part of the code by issue a single call to `allsuccessors`.

bookmark: simplify mutability check in `validdest` There is no need to check phase by hand. We already have a ctx method for that.

Merge with stable

Merge with mpm

bundle: add revset expression to show bundle contents (issue3487) bundle() revset expression returns all changes that are present in the bundle file (no matter whether they are in the repo or not). Bundle file should be specified via -R option.

tags: add formatter support

manifest: add formatter support

status: use condwrite to avoid zero-width format string hack

formatter: add condwrite method This makes handling conditional output tidier

run-tests: skip unnecessary operations on the args in the child proces The arguments are already sorted and never empty.

test-subrepo: adapt for Windows after 17c030014ddf Fixes @@ -606,7 +606,7 @@ issue3276_ok $ rm repo/s/b $ hg -R repo revert --all - reverting repo/.hgsubstate + reverting repo\.hgsubstate reverting subrepo s $ hg -R repo update 1 files updated, 0 files merged, 0 files removed, 0 files unresolved on Windows

test-mq-qrefresh: adapt for Windows after 39b7052b217b Fixes @@ -208,7 +208,7 @@ $ echo 'orphan' > orphanchild $ hg add orphanchild $ hg qrefresh nonexistentfilename # clear patch - nonexistentfilename: No such file or directory + nonexistentfilename: The system cannot find the file specified $ hg qrefresh --short 1/base $ hg qrefresh --short 2/base on Windows

merge with stable

Added signature for changeset 195ad823b5d5

Added tag 2.4 for changeset 195ad823b5d5

tests: fix test for issue3581 for vfat on Linux

filemerge: only run test for issue3581 on non-windows environments

openpath: specify binary mode and use util.posixfile follow-up to 0e2846b2482c

i18n-pt_BR: synchronized with afa7e6fa820b

merge with i18n

i18n-pt_BR: synchronized with 36ed69d4593d

i18n-ja: synchronized with 0188ddfa844e

subrepo: only do clean update when overwrite is set (issue3276) Files in a subrepo were overwritten on update. But this should only happen on a clean update (example: -C is specified). Use the overwrite parameter introduced for svn subrepos in c19b9282d3a7 to decide whether to merge changes (as update) or remove them (as clean). The new function hg.updaterepo is intruduced to keep all update calls in hg. test-subrepo.t is extended to test if an untracked file is overwritten (issue3276). (Update -C is already tested in many places.) The first two chunks are debugging output which has changed. (Because overwrite is not always true anymore for subrepos) All other tests still pass without any change.

i18n: change output of largefiles for summary to distinguish from one for outgoing Before this patch, largefiles extension uses "largefiles: No remote repo" message not only for "outgoing" as status report, but also for "summary" as summarized information. This sharing prevents message translators from inserting white spaces between "largefiles:" and "No remote repo" in translated message to align column position of summarized information. This patch changes output of largefiles for summary to distinguish from one for outgoing. This patch puts "no remote repo" into parentheses, because this is not summarized information.

i18n: make column positioning message of MQ summary output translatable Before this patch, one of column positioning messages of MQ summary output is not translatable, and patches are always listed up at fixed column position like below, when patch queue is not empty: mq: 1 applied, 1 unapplied But column positioning messages for other summarized information are translatable, so other information may be displayed at different column position on non-English locale environments. This patch makes column positioning message of MQ summary output translatable.

i18n: add "i18n" comment to column positioning messages of "hg summary" This comment makes it easier to distinguish such messages from others for message translators.

i18n: add "i18n" comment to column positioning messages of "hg log" This comment makes it easier to distinguish such messages from others for message translators.

i18n: add "i18n" comment to error messages of template functions

icasefs: make case-folding collision detection as deletion aware (issue3648) Before this patch, case-folding collision is checked simply between manifests of each merged revisions. So, files may be considered as colliding each other, even though one of them is already deleted on one of merged branches: in such case, merge causes deleting it, so case-folding collision doesn't occur. This patch checks whether both of files colliding each other still remain after merge or not, and ignores collision if at least one of them is deleted by merge. In the case that one of colliding files is deleted on one of merged branches and changed on another, file is considered to still remain after merge, even though it may be deleted by merge, if "deleting" of it is chosen in "manifestmerge()". This avoids fail to merge by case-folding collisions after choices from "changing" and "deleting" of files. This patch adds only tests for "removed remotely" code paths in "_remains()", because other ones are tested by existing tests in "test-casecollision-merge.t".

mq: fix qrefresh case sensitivity (issue3271) When calling qrefresh with filenames, the filenames were being treated as case-sensistive on case-insensitive file systems. So 'qrefresh foo' would not match file 'Foo', and it failed silently. This fix makes it work correctly on case-insensitive file systems. Previously the matching function was applied directly to the filenames. Now we apply the matching function through repo.status, which handles the case logic for us. A side effect of using repo.status is that if the qrefresh file doesn't exist, there is output stating it doesn't exist. Adds a test to an existing mq refresh case test.

url: use open and not url.open for local files (issue3624)

revset: accept @ in unquoted symbols (issue3686)

filemerge: use util.shellquote when calling merge (issue3581)

merge with i18n

i18n-pt_BR: synchronized with 71c1513fd560

clone: show status "updating to bookmark @" if bookmark @ is not on the default branch, we show updating to bookmark @ on branch <name>

test-clone: fix directory level 92980a8dfdfe inserted a "cd a" but didn't restore the previous directory.

patchbomb: -c is not an alias for the --confirm option The help text shows "-c/--confirm", which implies that -c is an alias for --confirm. Actually, it is an alias for --cc. Correct the text by removing -c.

dirstate: handle dangling junctions on windows (issue2579)

largefiles: don't copy largefiles from working dir to the store while converting Previously, if one or more largefiles for a repo being converted were not in the usercache, the convert would abort with a reference to the largefile being missing (as opposed to the previous patch, where the standin was referenced as missing). This is because commitctx() tries to copy all largefiles to the local store, first from the user cache, and if the file isn't found there, from the working directory. No files will exist in the working directory during a convert, however. It is not sufficient to force the source repo to be local before proceeding, because clone and pull do not download largefiles by default. This is slightly less than ideal because while the conversion will now complete, it won't be possible to update to revs with missing largefiles unless the user intervenes manually, because there is no default path pointing back to the source repo. Ideally these files would be cached during the conversion. This check could have been done in reposetup.commitctx() instead, but this ensures the local store directory is created, which is necessary to enable the standin matcher. The rm -> 'rm -f' change in the test is to temporarily suppress an error clearing the cache- as noted, the cache is is not repopulated during convert. When that is fixed, this can be changed back and the verification errors will disappear too.

largefiles: respect the rev when reading standins in copytostore() (issue3630) When the rev isn't specified, the standin for the working copy gets read. But convert doesn't update the working copy for each cset it processes, so there is no standin and the 'hg convert' would abort complaining about the standin being missing. Note that if the largefile is not in the user cache, 'hg convert' complains about the largefile itself missing from the destination repo.

subrepo: subrepo isolation, pass baseui when cloning a new subrepo (issue2904) Create the repo with baseui because it should only get the global configuration. After this patch issue2904 is finally fixed.

subrepo: more isolation, only use ui for hg.peer when there is no repo ui contains repo specific configuration, so do not use it when there is a repo. But pass it to hg.peer when there is no repo. Then it only contains global configuration.

peer: subrepo isolation, pass repo instead of repo.ui to hg.peer Do not pass ui because it contains the configuration of the repo. It is the same object as repo.ui. When a repo is passed to hg.peer, the global configuration is read from repo.baseui.

subrepo: isolate configuration between each repositories in subrepo tree Before this patch, repository local configurations are not isolated between repositories in subrepo tree, because "localrepository" objects for each subrepositories are created with "ui" instance of the parent of each ones. So, local configuration of the parent or higher repositories are visible also in children or lower ones. This patch uses "baseui" instead of "ui" to create repository object: the former contains only global configuration. This patch also copies 'ui.commitsubrepos' configuration to commit recursively in subrepo tree, because it may be set in not "repo.baseui" but "repo.ui".

bookmarks: backed out new message from changeset 52c7e171e355 Not strictly a bugfix, hasn't achieved consensus yet.

clone: print bookmark name when clone activates a bookmark Similar to the message that prints the checked out branch name. Without this, the user might accidentally move the @ bookmark.

clone: activate @ bookmark if updating to it The message "updating bookmark @ failed!" in test-bookmarks-pushpull.t is correct, because the changeset that the @ bookmark points to is not pushed to the target repository.

test-clone.t: check that branch "@" is not automatically checked out This verifies that c9339efed653 fixes the second problem mentioned in issue3677.

test-bookmarks.t: check that bookmark "default" is not automatically checked out This verifies that c9339efed653 fixes the problem originally reported in issue3677.

clone: make sure to use "@" as bookmark and "default" as branch (issue3677) Before this change a bookmark named "default" or a branch named "@" would cause the wrong changeset to be checked out. The change in output of test-hardlinks.t is due to the fact that no unneeded tag lookups for the tags "@" or "default" happen, therefore the cache file is not created.

bookmark: simplify nodemap check introduced in the previous changeset

bookmark: prevent crashing when a successor is unknown locally (issue3680) The `%ln` revset substitution does not accept unknown node. We prune unknown node from potential successors before computing descendants. This have no impact on the result of this function. - Descendants of unknown changeset as unknown, - all successors of unknown changesets are already return by the call who returned those same unknown changesets, - unknown changesets are never a valid destination for a bookmark.

bookmark: complexity pull-push test to have deeper tree This changeset only touch test. The previous test was correct, it tested that the successors of an old bookmark position was seen as a valid destination for bookmark. However, a newer version is made for two reason: (1) The new test check further. It check that the descendant of the successors is a valid destination (2) An ever more complexe test is needed to validate a future fix to issue 3680 Splitting complexification of the test and actual bugfix help to reduce the noise in the bugfix changeset. Issue 3680 is NOT fixed by this changeset.

amend: fix incompatibity between logfile and message option (issue3675) Bug introduced by 9732473aa24b

test-remove: adapt to differing error message on Windows On Windows, this part of the test failed with $ hg rm --after nosuch - nosuch: No such file or directory + nosuch: The system cannot find the file specified [1] Fixed by glob-ing away the error message if the test is run on Windows (see for example test-bad-pull.t line 3 for precedent). test-remove.t now passes on Windows.

test-remove: fix \ vs. / issues on Windows This part of the test failed with $ hg rm --after d1 - removing d1/a + removing d1\a on Windows.

verify: fix all doubled-slash sites (issue3665)

patchbomb: respect --in-reply-to for all mails if no intro message is sent Before this change, the thread hierarchy looked like this: PARENT PATCH1/x PATCH2/x PATCH3/x ... Now it is: PARENT PATCH1/x PATCH2/x PATCH3/x ... With an introductory message the behaviour is unchanged: PARENT INTRO PATCH1/x PATCH2/x PATCH3/x ...

test: add test for the issue introduced by e410be860393 (issue3669) e410be860393 (released as Mercurial 2.3) introduced the issue that the revset program started with 40 hexadecimal letters caused unexpected result at "hg log" execution. This issue was already fixed by bde1185f406c (released as 2.3.1), but there is no test to examine whether this issue is certainly fixed or not: no test fails even if bde1185f406c is backed out. This patch adds test for this issue. Added test is also confirmed to fail, when it is tested against: - Mercurial 2.3, or - Mercurial 2.3.1 or later with backing bde1185f406c out

obsolete: add missing line feed in debug error message

merge with stable

localrepo: translate "push includes X changeset" only once This was introduced in 12fdaa30063a

merge stable heads

Merge with stable

Merge repeated-slash fix

verify: tolerate repeated slashes in a converted repo (issue3665) These slashes are a hangover from issue3612, fixed in e4da793998bf. Although the bugfix in that commit is correct, the test it adds does not replicate the conditions for the bug correctly.

scmutil: generalize message to make it more i18n-friendly

update: check for missing files with --check (issue3595)

remove: don't return error on directories with tracked files Spotted by Sergey <sergemp@mail.ru>

largefiles: use 'default' instead of 'default-push' when pulling (issue3584) This only applies to downloading largefiles, and only when no source for the pull is explicitly provided. The repository itself was properly being pulled via 'default' previously. Using --all-largefiles is not necessary on a bare pull to test this (this existing test is merely a convenience), but it is required to test pulling on the rebase path. Note that the errors generated in the --rebase case are because the repo specified doesn't have the largefiles in its cache (though they are in the user cache), so the errors are misleading. Specifying --all-largefiles when cloning to 'b' fixes this, but instead of errors, it reports caching only 5 largefiles instead of the 9 that come up missing. Likely this is because the largefile download procedure tries to download missing files for each rev, and some of the files have standins in more than one rev that gets pulled.

scmutil: add mustaudit delegation to filtervfs (issue3673)

scmutil: abstract out mustaudit delegation

clone: don't %-escape the default destination (issue3145)

test-largefiles: fix failing test on Windows The test failed due to / vs. \ in paths issues. Fixes 08d11b82d9fc.

test-push-http: fix failing test on Windows The test failed due to problems with escaping. Fixes 8474be4412ca.

Added signature for changeset d118a4f4fd16

Added tag 2.4-rc for changeset d118a4f4fd16

merge default into stable for 2.4 code freeze

hgwebdir: make collapsed folders easier to distinguish from repositories Add a "/" character after the collapsed folder names, to make them easier to distinguish from regular repository and subrepository entries.

help: indicate help omitting if help document is not fully displayed Before this patch, there is no information about whether help document is fully displayed or not. So, some users seem to misunderstand "-v" for "hg help" just as "the option to show list of global options": experience on "hg help -v" for some commands not containing verbose containers may strengthen this misunderstanding. Such users have less opportunity for noticing omitted help document, and this may cause insufficient understanding about Mercurial. This patch indicates help omitting, if help document is not fully displayed. For command help, the message below is displayed at the end of help output, if help document is not fully displayed: use "hg -v help xxxx" to show more complete help and the global options and otherwise: use "hg -v help xxxx" to show the global options For topics and extensions help, the message below is displayed, only if help document is not fully displayed: use "hg help -v xxxx" to show more complete help This allows users to know whether there is any omitted information or not exactly, and can trigger "hg help -v" invocation. This patch causes formatting help document twice, to switch messages one for omitted help, and another for not omitted. This decreases performance of help document formatting, but it is not mainly focused at help command invocation, so this wouldn't become problem.

httpclient: fix calling convention violation

largefiles: distinguish "no remote repo" from "no files to upload" (issue3651) Before this patch, when no files to upload, "hg outgoing --large" and "hg summary --large" show "no remote repo", even though valid remote repository is specified. It is because that "getoutgoinglfiles()" returns None, not only if no valid remote repository is specified, but also if no files to upload. This patch makes "getoutgoinglfiles()" return empty list when no files to upload, and makes largefiles show "no files to upload" message at that time. This patch doesn't test "if toupload is None" route in "overrideoutgoing()", because this route is not executed unless remote repository becomes inaccessible just before largefiles specific processing: successful execution of "orig()" means that at least one of "default", "default-push" or dest is valid one, and that "getoutgoinglfiles()" never returns None in such cases. At "hg summary --large" invocation, this patch shows message below: largefiles: (no files to upload) This follows the message shown by "hg summary" with MQ: mq: (empty queue)

push: refuse to push bumped changeset This applies the same logic as used for `obsolete` and `unstable` changesets. Refuse to push them without force. We'll probably want to factor this logic with two new functions `pctx.troubled()` and `ctx.troubles()`. But I'm waiting for the third "trouble" to make it into core.

obsolete: simplify push abort message to: push includes xxx changeset: yyyyyyyyyy

context: add a `bumped` method to `changectx` Same as `unstable()`, returns true if the changeset is bumped.

obsolete: add a flag that allows fixing "bumped" changeset The first obsolescence flag is introduced to allows for fixing the "bumped" changeset situation. bumpedfix == 1. Creator of new changesets intended to fix "bumped" situation should not forget to add this flag to the marker. Otherwise the newly created changeset will be bumped too. See inlined documentation for details.

debugobsolete: add --flags option This options allows to specify the `flag` part of obsolete markers. For details about marker flags, check the `mercurial/obsolete.py` documentation. Some random flag are added to a marker to test this feature.

revset: add a bumped revset Select bumped changesets.

obsolete: add the detection of bumped changeset. Bumped changesets are non-public changesets that tries to succeed a public() changeset.

obsolete: have `allsuccessors` takes a list of nodes Additional logic, used to detect mutable history troubles, will need to quickly compute successors of a whole set of changeset.

obsolete: rename `anysuccessors` into `allsuccessors` The "any" prefix looks like it returned a boolean. `allsuccessors` is more accurate.

obsolete: rename `getobscache` into `getrevs` The old name was not very good for two reasons: - caller does not care about "cache", - set of revision returned may not be obsolete at all. The new name was suggested by Kevin Bullock.

largefiles: always create the cache and standin directories when cloning The standin matcher only works if the .hglf directory exists (and it won't exist if 'clone -U' is used, unless --all-largefiles is also specified). Since not even 'update -r null' will get rid of the standin directory, this ensures that the standin directory always exists if the repo has the 'largefiles' requirement. This requirement is only set after a largefile is committed, so these directories will not be created for repos that have the extension enabled but have not committed a largefile. With the standin directory in place, 'lfconvert --to-normal' will now be able to download the required largefiles when converting a repo that was created with 'clone -U', and whose files are not in the usercache. The downloadlfiles command could probably be put inside the 'largefiles' requirement conditional too, but given that the user specified --all-largefiles, there is likely an expectation to print out the number of largefiles downloaded, even if it is 0.

largefiles: fix a traceback in lfconvert if a largefile is missing (issue3519) The largefile may be missing for various reasons, including that a remote repository was cloned without the --all-largefiles option. Therefore, it seems reasonable to attempt to download the missing files and failing that, abort and indicate the affected file and revision so the user can manually fix the problem.

bookmarks: further flatten code This hopefully clarifies the behavior when no NAME is passed, by separating the branches for listing bookmarks vs. deactivating the current bookmark.

scmutil: add bad character checking to checknewlabel This factors out the checks from tags and bookmarks, and newly applies the same prohibitions to branches. checknewlabel takes a new parameter, kind, indicating the kind of label being checked. Test coverage is added for all three types of labels.

tag: disallow '\0' in tag names This brings the list of disallowed characters in line with bookmarks.

dirstate: use scmutil.checknewlabel to check new branch name

bookmarks: use scmutil.checknewlabel Validation is pulled up into the commands module to avoid an import cycle.

scmutil: add function to validate new branch, tag, and bookmark names For now the new function only checks to make sure the new label name isn't a reserved name ('tip', '.', or 'null'). Eventually more of the checks will be unified between the different types of labels. The `tag` command is trivially updated to use it. Updating branches and bookmarks to use it is slightly more invasive and thus reserved for later patches.

bookmarks: disallow bookmarks named 'tip', '.', or 'null' This makes bookmarks reject the same reserved names as tags and branches.

bookmarks: abort directly on invalid name This consolidates the abort message for a bookmark name containing invalid characters into one place, and renames the valid() method to checkvalid() to reflect the fact that it's no longer a predicate.

bookmarks: remove redundant check for newline New bookmarks are already checked for illegal characters (':', '\0', '\n', and '\r') in bookmarks.valid().

test-tag: test that all reserved names are rejected

amend: do a bare kill of temporary changeset Before this changeset the temporary changeset created by amend was made a precursors on the amend result. This add unnecessary complexity to the obsolescence graph. This temporary commit will probably disappear in the future. It is an unwanted byproduct of amend that nobody cares about.

amend: add noise in extra to avoid creating obsolescence cycle (issue3664) Obsolescence cycle are bad and should be avoided as much as possible. The current amend implemented touch changeset meta data as few as possible. This make is easy for amend to result in the same node than a precursors. We add some deterministic noise in extra to avoid this. In practice, the hex of the amended changeset is stored in 'amend_source' extra key.

changelog: extract description cleaning logic in a dedicated function The amend logic have use for it.

bookmarks: replace code-redundant comment with something more useful

bookmarks: remove another uneeded return

update: remove unnecessary argument check

grep: colorize all fields Colors were picked in accordance to GNU grep.

test-grep: add a test for -l

clfilter: remove use of xrange in revset For changelog level filtering to take effect it need to be used for any iteration. Some remaining use of `xrange` in revset code is replace by proper use of `changelog.revs` or direct iteration over changelog.

largefile: use `self` in repo method instead of `repo` Most method added (or overwritten) to repo by largefile works on `repo` instead of `self`. This currently works without trouble because `self` and `repo` are likely the same. However this is semantically dubious and this may cause issue for filtering. `self` may be proxy object different from the `repo` one. This changeset fix that and use `self` when applicable.

test: use proper subclassing in `test-issue2137.t`. To use changelog filtering on the repository, we plan to use "proxy" object that perfectly mock a repository but with a filtered changelog. Altering the `repo.commit` function using `extensions.wrapfunction` will prevent the logic to propagate to the proxy class by the mean of inheritance. We changes the extension to use subclassing as expectable.

run-tests: remove dead code for supporting old test scripts

run-tests: handle windows crlf in .py tests again Before af7c6bc48d8d all crlf occurrences in test output on Windows were simply changed to lf. In af7c6bc48d8d it was replaced by more clever handling in the .t test runner ... but the .py runner was forgotten and many .py tests were failing on Windows. The crlf/lf replacement is now reintroduced in the py test runner.

test-run-tests.t: fix wrong test case for cr handling on Windows An incorrect and failing test case was introduced in af7c6bc48d8d.

win32mbcs: add reversing wrapper for some unicode-incompatible functions. This changeset fix the problem to use win32mbcs with mercurial 2.3 or later. The problem is brought by side effect of modification of encoding.upper() (changeset 17236:9fb8312dbdbd) because upper() does not accept unicode string argument. So wrapped util.normcase() which uses upper() will fail. In other words, upper() and lower() are unicode incompatible. To fix this issue, this changeset adds new wrapper for reversed conversion (unicode to str) for lower() and upper() to use them safely.

convert: normalize paths in filemaps (issue3612) convert doesn't normalise double slashes in paths. Path normalization is applied when a path is loaded into filemap and when a file lookup request is issued to filemap.

repair: use node to track post-strip bookmark target Revision numbers are unstable when non-consecutive revs are stripped.

keyword: move commands.inferrepo for code maintenance

largefiles: update lfutil.findoutgoing() discovery method It looks like this method missed the updates in 30273f0c776b (which changed the preferred discovery method from findcommonincoming() to findcommonoutgoing()), and cd956049fc14 (which rolls up the outgoing lists into a single object).