largefiles: optimize update speed by only updating changed largefiles Historically, during 'hg update', every largefile in the working copy was hashed (which is a very expensive operation on big files) and any largefiles that did not have a hash that matched their standin were updated. This patch optimizes 'hg update' by keeping track of what standins have changed between the old and new revisions, and only updating the largefiles that have changed. This saves a lot of time by avoiding the unecessary calculation of a list of sha1 hashes for big files. With this patch, the time 'hg update' takes to complete is a function of how many largefiles need to be updated and what their size is. Performance tests on a repository with about 80 largefiles ranging from a few MB to about 97 MB are shown below. The tests show how long it takes to run 'hg update' with no changes actually being updated. Mercurial 2.1 release: $ time hg update 0 files updated, 0 files merged, 0 files removed, 0 files unresolved getting changed largefiles 0 largefiles updated, 0 removed real 0m10.045s user 0m9.367s sys 0m0.674s With this patch: $ time hg update 0 files updated, 0 files merged, 0 files removed, 0 files unresolved real 0m0.965s user 0m0.845s sys 0m0.115s The same repsoitory, without the largefiles extension enabled: $ time hg update 0 files updated, 0 files merged, 0 files removed, 0 files unresolved real 0m0.799s user 0m0.684s sys 0m0.111s So before the patch, 'hg update' with no changes was approximately 9.25s slower with largefiles enabled. With this patch, it is approximately 0.165s slower.

mq: make qimport --push push all imported patches (issue3130) Only the first imported one was pushed.

convert: tolerate spaces between splicemap parent ids (issue3203) Splicemap lines are documented in hg help convert like: key parent1, parent2 but parsed like: key, parents = line.strip().rsplit(' ', 1) parents = parents.replace(',', ' ').split() The rsplit() call was introduced to handle spaces in keys for the generic mapfile format. Spaces can appear in svn identifiers since they contain path components. This logic makes less sense with splicemap since svn identifiers can also appear on the right side, even if it is a bit less likely. Given the parsing is theorically broken, I would rather follow what is documented already and is correct in the main case where all identifiers are hg hashes. Also, using svn identifiers in a splicemap sounds difficult as they are not easily accessible.

test-commandserver: flush stdout

localrepo: clear _filecache on rollback (issue3261) Files are being replaced by rollback but the corresponding data in localrepo isn't actually updated for things like bookmarks, phases, etc. Then when rollback is done, the cache is updated thinking it has the most up-to-date data, where in fact it is still pre-rollback. We clear _filecache to force everything to be recreated.

scmutil: update cached copy when filecached attribute is assigned (issue3263) When assigning a new object to filecached properties, the cached object that was kept in the _filecache map was still holding the old object. By implementing __set__, we track these changes too and update the cached copy as well.

cmdserver: invalidate the dirstate when running commands (issue3271) The dirstate is invalidated separately outside of invalidate() which is already being called (other callers of invalidate() seems to suggest the separation is there for a reason).

localrepo: reset _phasesdirty flag after writing

import: handle git renames and --similarity (issue3187) There is no reason to discard copy sources from the set of files considered by addremove(). It was done to handle the case where a first patch would create 'a' and a second one would move 'a' to 'b'. If these patches were applied with --no-commit, 'a' would first be marked as added, then unlinked and dropped from the dirstate but still passed to addremove(). A better fix is thus to exclude removed files which ends being dropped from the dirstate instead of removed. Reported by Jason Harris <jason@jasonfharris.com>

context: make workingctx.forget() really warn about untracked files

largefiles: check whether specified patterns are related to largefiles strictly current 'lfiles_repo.status()' implementation examines whether specified patterns are related to largefiles in working directory (not to STANDIN) or not by NOT-EMPTY-NESS of below list: [f for f in match.files() if f in lfdirstate] but it can not be assumed that all in 'match.files()' are file itself exactly, because user may only specify part of path to match whole under subdirectories recursively. above examination will mis-recognize such pattern as 'not related to largefiles', and executes normal 'status()' procedure. so, 'hg status' shows '?'(unknown) status for largefiles in working directory unexpectedly. this patch examines relation of pattern to largefiles by applying 'match()' on each entries in lfdirstate and checking wheter there is no matched entry. it may increase cost of examination, because it causes of full scan of entries in lfdirstate. so this patch uses normal for-loop instead of list comprehensions, to decrease cost when matching is found.

merge with stable

log: remove caching of all visited revisions (issue3253) Not only does this eat all available memory for some users, it's slower.

pull: backout change to return code This bit a number of people.

convert: use splicemap entries when sorting revisions (issue1748) When sorting revisions before converting them, we have to edit the revision graph using splicemap entries. Otherwise, a spliced revision may be converted before its synthetic parents. Invalid splicemap revisions are now detected before starting the conversion.

convert: turn splicemap into a simple dictionary Parsing the splicemap as a mapfile was a pain because map does not let us override its parsing code and splicemap entries are not key/values. Besides we had no need for mapfiles extra features. Just parse the splicemap and return a dictionary.

debugrevspec: mention --verbose to print the parsed tree

largefiles: only cache largefiles in new heads This fixes a serious performance regression in largefiles introduced in Mercurial 2.1. Caching new largefiles on pull is necessary, because otherwise largefiles will be missing (and unable to be downloaded) when the user tries to merge or rebase a new head with an old one. But this is an expensive operation and should only be done for heads that are new from the pull, rather than on all heads in the repository.

merge with stable

mq: restore _branchtags() fast path (issue3223) Since a5917346c72e, mq saves the nodeid of the first applied patch to cache/branchheads, which breaks the optimized cache handling introduced in fbf8320f25c8. The problem is the revision being committed is appended to mqrepo.applied after the commit succeeds, which means mqrepo._branchtags() performs a regular update and write the first applied patch to the branch cache. One solution is to set a context variable _committingpatch on the mqrepo while it is committing a patch and to take it in account when deciding to fast-path mqrepo._branchtags(). Not really elegant but it works. The changes to test-mq-caches.t reverse changes introduced by a5917346c72e. The cache should not have been updated with mq records. The changes to test-keyword.t are indirectly caused by a5917346c72e. Reported and analyzed by Yuya Nishihara <yuya@tcha.org> Notes: - qpush still makes a slow path _branchtags() call when checking heads. Maybe this can be optimized. - be careful when merging this patch in default as secretcommit() was renamed newcommit() right after the end of the code freeze.

mq: ensure all mq commits are made with secretcommit() Having a common code path helps fixing things globally.

convert/bzr: ignore nested repos when listing branches (issue3254) Reported by A.S. Budden <abudden@gmail.com>

tests: tighten checks for octal escapes in shell printf. printf on AIX default shell ksh (89) says \1 is an invalid escape. It insists on at least 2 digits. This causes failures in test-keyword.t and test-status.t. check-code.py already looks out for \NNN and recommends using Python for outputting octal values. Extend the check to \NN and \N and fix up resulting failures.

phase: when phase cannot be reduced, hint at --force and return 1 (BC) Before, trying to change the phase of a "public" node to "draft" would result in: $ hg phase --draft . no phases changed [0] With this patch: $ hg phase --draft . cannot move 1 changesets to a more permissive phase, use --force no phases changed [1] On partial failures, the exit status is now 1 instead of 0 and the number of changed nodes is displayed while it was only with --verbose. It seems useful to tell the user that despite the apparent failure, something changed. $ hg phase --draft '5 or 7' cannot move 1 changesets to a more permissive phase, use --force phase changed for 1 changesets [1]

revset: fix alias substitution recursion (issue3240) The revset aliases expansion worked like: expr = "some revset" for alias in aliases: expr = alias.process(expr) where "process" was replacing the alias with its *unexpanded* substitution, recursively. So it only worked when aliases were applied in proper dependency order. This patch rewrites the expansion process so all aliases are expanded recursively at every tree level, after parent alias rewriting and variable expansion.

update: just merge unknown file collisions The unknown file collision rule was introduced as an extension of the "should be clean when merging" rule. Unfortunately, it got applied to the normal update path, which should be happy to merge local changes. This patch gives us merges for unknown file collisions on update, while preserving abort for merge and update -c.

merge: don't use unknown() This removes use of unknown files for building the synthetic working directory manifest used by manifestmerge. Instead, we adopt the strategy used by _checkunknown. Side-effect: unknown files are no longer moved by remote directory renames, and now are left alone like ignored files.

merge: refactor unknown file conflict checking Previously, we would do a full working directory walk including unknown files to perform a merge. In many cases, this was painful because unknown files greatly outnumbered tracked files and generally had no useful effect on the merge. Here we instead wait until we find a file in the destination that's not tracked locally and detect if it exists and is not ignored. This is usually cheaper but can be -more- expensive in the case where we're adding a huge number of files. On the other hand, the cost of statting the new files should be dwarfed by the cost of eventually writing them. In this version, case collisions are detected implicitly by os.path.exists and wctx[f] lookup.

update: use normal update path with --check (issue2450) This avoids clobbering unknown files on update by not using overwrite mode.

fetch: use update rather than clean when updating (issue3246) We already verify the working directory is "clean" before starting so there's no advantage to clobbering the working directory.

merge with stable

mdiff: adjust hunk offsets with --ignore-blank-lines (issue3234) When diffing the following documents with --ignore-blank-lines (-B): $ cat > a <<EOF > > > > b > x > d > EOF and: $ cat > b <<EOF > b > y > d > EOF the context lines are taken from the first document, even if the lines differ (with -w or -b) or if the number of lines differ (with -B). In the second case, we have to adjust the hunk new lines offsets or we end with inconsistent diffs like (see the @@ offsets): diff -r 0e66aa54f318 a --- a/a +++ b/a @@ -1,4 +1,3 @@ b -x +y d Note that having different context lines in a and b means the diff can be applied but is not invertible. Reported by Nicholas Riley <com-selenic@sabi.net>

merge with stable

merge with i18n

merge with i18n

i18n-de: fix rst markup in command output (issue3247)

merge with i18n stable

i18n-ru: fix hg command rst markup

i18n-ja: fix rst markup in hg command output

i18n-de: fix rst markup in command output (issue3247)

merge with i18n stable

i18n-de: fix hg command rst markup and quoting (issue3247)

merge with i18n stable

i18n-de: fix rst markup in command output (issue3247)

windows: use 'str.replace()' instead of combination of split() and join() 'str.replace()' is twice (or more) as fast as combination of split() and join(). combination of split() and join() in current 'util.pconvert()' implementation comes only from historical reason when win32mbcs extension was introduced.

largefiles: reduce redundant splitstandin/standin combination assumptions: - 'lfutil.splitstandin(f)' strips 'f' of '.hglf/', if it can - 'lfile(f)' examines 'lfutil.standin(f) in manifest' - 'lfutil.standin(f)' adds '.hglf/' to 'f' and - 'f' should start with '.hglf/' there, because it is already examined by 'lfutil.isstandin(f)' then: lfile(lfutil.splitstandin(f)) => lfile(<f without '.hglf/'>) => lfutil.standin(<f without '.hglf/'>) in manifest => <<f without '.hglf/'> with '.hglf/'> in manifest => f in manifest

largefiles: reduce OR-ing of same conditions

subrepo: rewrite handling of subrepo state at commit (issue2403) When the contents of .hgsubstate are stale (either because they've manually been tweaked or partial updates have confused it), we get confused about whether it actually needs committing. So instead, we actively consult the parent's substate and compare it the actual current state when deciding whether it needs committing. Side effect: lots of "committing subrepo" messages that didn't correspond with real commits disappear. This change is fairly invasive for a fairly obscure condition, so it's kept on the default branch.

subrepo: add basestate method This will allow us to sort out what's going on with state .hgsubstate files more precisely.

bdiff: fix malloc(0) issue in fixws() If fixws() is called on a zero-length string, malloc(0) is called and expected to return a pointer. Which it does on e.g. Linux. AIX returns NULL, which it is also legal, but the malloc() is then assumed to have failed. So ensure a valid pointer is always returned.

forget: show warning messages for forgetting in subrepo correctly in 'cmdutil.forget()': for f in match.files(): if match.exact(f) or not explicitonly: .... is equal to: for f in match.files(): if True: .... because 'f' from 'match.files()' should 'match.exact(f)': - 'match.files()' returns 'self._files' - 'match.exact(f)' examines 'f in self._fmap', - 'self._fmap' of match is 'set(self._files)' then, 'explicitonly' wants to suppress warning messges, if it is true (= 'cmdutil.forget()' is invoked from 'subrepo.forget()'). so, current code should be fixed as: if not explicitonly: for f in match.files(): ....

tests: fix command line overflow on ksh "<command> `hg manifest`" overflows the puny 1024 byte command line length limit on AIX default shell ksh. Replace with "hg manifest | xargs <command>".

use util.localpath() instead of 'str.replace()' to unify path conversion

i18n: use util.pconvert() instead of 'str.replace()' for problematic encoding some problematic encodings use backslash as part of multi-byte characters. util.pconvert() can treat strings in such encodings correctly, if win32mbcs is enabled, but str.replace() can not.

i18n: use util.pconvert() instead of 'str.replace()' for problematic encoding some problematic encodings use backslash as part of multi-byte characters. util.pconvert() can treat strings in such encodings correctly, if win32mbcs is enabled, but str.replace() can not.

merge with stable

mq: make qprev return the previous applied patch (issue3245) The current behaviour is to return the previous one in the series but at the same time the implementation is buggy because it does not take guarded patches in account.

mq: fix qnext when all remaining patches are guarded When all remaining patches are guarded, qnext used to return the last of the queue anyway.

convert/bzr: test tags conversion

convert/bzr: handle empty bzr repositories (issue3233)

convert/bzr: convert all branches (issue3229) (BC) Instead of opening the target bzr checkout as a single branch, we try to open it as a repository. This has the following effects: - All branches are now converted - bzr branch names are preserved. Previously, the selected branch was always converted as 'default'. Branches without a name or 'trunk' are mapped to 'default branch. - Lightweight checkouts are no longer supported. Maybe they can be, I did not try to fix that at all. Implementation notes: - This was a quick fix, I have no knowledge of bzr API besides browsing 2.0.3 sources. - The fix was only tested on OSX against bzr 2.4.2. - Tags discovery does not handle collisions. I have no idea how tags work in bzr so maybe such collisions are not possible.

convert/bzr: expect unicode metadata, encode in UTF-8 (issue3232) Before this patch, metadata and file names were interpreted like: - unicode objects were converted to UTF-8 - non unicode objects were left unchanged Looking at the code and bzr being known for transcoding filenames, we expect everything to be returned as unicode objects, and we want to encode them in UTF-8, like the subversion source does. To do that, we just remove the custom implementation of .recode().

test-mq: fix sed call on OSX - The regular sed on OSX wants: -i "" not -i"" - GNU sed wants: -i"" not -i "" Backups are fine.

mq: rename secretcommit to newcommit This function create secret changeset only when mq.secret is true

merge with stable

i18n-de: update po and remove fuzzies

i18n-de: Rework and fixes continued

Added signature for changeset 2aa5b51f310f

Added tag 2.1 for changeset 2aa5b51f310f

phases: don't complain if cset is already public on pushkey (issue3230)

i18n-pt_BR: synchronized with 140b6282ac79

merge with i18n

mq: prevent rewriting operation on public changeset The following operation are denied if the mq changeset is public * qrefresh * qpop * qfold note: qimport was already denied.

i18n-pt_BR: miscellaneous spelling and wording fixes

i18n-ja: synchronized with 0f99e709ce08

i18n-ru: synchronized with c0b2986b37b8; public phase name changed

i18n-pt_BR: synchronized with c0b2986b37b8

qpush: avoid trying to manage existing history (issue2218)

bundlerepo: try to find containing repo on creation (issue1812)

phases: add resync example to help topic

mq: fix secret description in help

pull: return 1 when no changes found (BC) Currently we have the following return codes if nothing is found: commit incoming outgoing pull push intended 1 1 1 1 1 documented 1 1 1 0 1 actual 1 1 1 0 1 This makes pull agree with the rest of the table and makes it easy to detect "nothing was pulled" in scripts.

push: don't treat bookmark as a found change Treating bookmarks as a found change results in confusing documentation or output and is less useful for scripting.

glossary: add phase terms

merge with i18n

i18n-sv: snchronized with ed9f40bc346f

i18n-ru: phases help translated

i18n-ru: minor corrections on unclear strings

i18n-ru: synchronized with 2605fc990725

i18n-ru: verifying uncertain translations.

phases: allow phase name in phases.new-commit settings Before this commit only phase index where accepted

qfinish: do not set secret changeset to draft if mq.secret=false If mq.secret=false, mq should not touch phase

mq: take mq.secret configuration into account when picking the default phase

qimport: when mq.secret=True set qimported revision as secret

qrefresh: keep changeset phase during refresh

phases: write default phase data as soon as possible. Marks phase data as dirty when computing default phase. Actual writing is done when the lock is released. So, read only operation don't write phase data because they don't lock the repo.

phase: accept old style revision specification

push: return 1 if no changes found (issue3228) Currently we have the following return codes if nothing is found: commit incoming outgoing pull push intended 1 1 1 1 1 documented 1 1 1 0 1 actual 1 1 1 0 0 This fixes the lower-right entry.

push: more precise failure check on subrepo push This will let us distinguish between nothing to push and push failed

push: return 0 when updating bookmarks

push: fix coding errors in unused branch This branch will get re-enabled in a subsequent fix

bookmarks: move current bookmark on update -u and bare pull -u (issue3222) Currently, this won't update when a #branch spec is in the URL

phases: fix verify with secret csets Verify uses repo.cancopy() to detect whether a repo is a plain old local repo, so it was giving a confusing error message when secret changesets were present.

mq: add secret setting

merge with i18n

i18n-pt_BR: synchronized with f2c49d9dee57

tests: don't rely on 'false' exit code in test-run-tests.t On Solaris: $ sh -c 'false ; echo $?' 255

tests: don't use alias alias doesn't work in Solaris sh.

tests: don't include plain 'diff -U0' content in test-diff-unified.t The only purpose of 'diff -U0' was to show that 'hg diff -U0' did almost the same. Unfortunately 'diff -U0' on Solaris gives incorrect line numbers. Masking it with globs wouldn't it make sense, so removing it is the best solution.

help: add examples to phases topic

update: note updated bookmark This brings update in line with pull: we notify the user when an action indirectly updates a bookmark.

resolve: mention merge-tools topic in help

revset: add tests for 'remote()' predicate

revset: fix documentation for 'remote()' predicate current documentation for 'remote()' predicate is wrong about specification of parameters. there are 3 patterns: # of param: id: remote: - 0 current branch "defult" remote - 1 specified "defult" remote - 2 specified specified

revset: fix 'remote()' failure when remote repo has more revs than local current 'remote()' implementation uses 'r' variable, even if it is not initialized when 'if n in repo' is not true. this causes unexpected exception.

help/config: fix help command markup

merge with i18n

i18n-ru: synchronized with bf502ccc46d7; fuzzies removed

i18n-ru: fix typos in `resolve` help * help entry is called 'merge-tools', not 'merge-tool'; * envvar is called HGMERGE, not HGEMERGT.

merge: defer symlink flag merging to filemerge (issue3200) Previously, we could change a normal file into a corrupt symlink when trying to merge a symlink flag. Now, we leave the flag alone and let filemerge deal with it (usually by a prompt). We also drop a redundant flag setting after filemerge (now dealt with by ms.resolve) that would cause similar corruption.

help: fix publish option spelling in phases topic

help: add missing phase config items to config topic

help/phases: remove trailing whitespace

sslutil: more helpful fingerprint mismatch message This will aid debugging for users of sites that renew certs.

help: add phases topic

i18n-pt_BR: synchronized with 0b05e0bfdc1c

copies: remove stray print

scmutil: unify some 'no changes found' messages This keeps the wording more consistent for secret csets

outgoing: display info about secret changets while no sharable changeset found

bundle: display info about secret changets while no sharable changeset found

test-rebase-collapse: Add test for rebase regression introduced in 12309c09d19a

test-commandserver: test that phase data is being refreshed

localrepo: delete _phaserev when invalidating caches

localrepo: fix phaseroots filecache definition it's in .hg/store

discovery: fix regression when checking heads for pre 1.4 client (issue3218) This code is not run by the test suite and silly name error were lurking here.

help: mark strings for translation

bookmarks: respect --debug during incoming/outgoing

largefiles: fix revert on missing largefile (issue3217)

largefiles: fix transplant for all cases (issue3192)

i18n-ru: synchronized with 9f2ed48f8cda

i18n-sv: translated phase

i18n-sv: translating strings for add, annotate, debuginstall

i18n-sv: synchronized with c2e6c5ef4555

merge with i18n

copies: eliminate criss-crosses when chaining Before the copies refactoring, we declared that if a and b were present in source and destination, we ignored copies between them. The refactored code could however report b was a copy of a and vice versa in a situation where we looked for differences between two identical changesets that copy a to b. y / x \ y'

i18n-pt_BR: synchronized with f7c8d6ee6056

merge with i18n

i18n-sv: synchronized with 7ed056f1e97d

mq: pass qbase node instead of mq statusentry in phasedefaults This was breaking my remotebranches extension in a completely mystifying way, because repo.lookup was failing to resolve the statusentry. I'm not sure how this works absent my remotebranches extension, but doing it this way looks more correct and doesn't break anything.

patch: a little bit more robust line counting on diff --stat (issue3183)

convert: subversion convert abort on revision not found (issue 3205)

bash_completion: added -o filename option to _hg_status (issue3101)

phase: alway return a value With the previous code, the phase command was returning None when displaying phase and 0 or 1 when moving them. This seemed an bit odd and displaying phase return 0 too now.

largefiles: fix addremove when no largefiles are specified

revset: include the correct first ancestor change for follow(file) Previously we always included '.', which may not touch a file. Instead, find the file revision present in '.' and add its linkrev. This matches the results of 'hg log --follow file'.

revset: remove unreferenced followfile function

revsets: provide contexts for filesets Before this change, revsets containing fileset patterns failed. This allows queries like: hg log -r "contains('set: added() and symlink()')"

fileset: don't attempt to check data predicates against removed files

revset: simplify help not about quoting

Added signature for changeset db33555eafea

Added tag 2.1-rc for changeset db33555eafea

bookmarks: correct correction of -i

commands: drop exclamation mark in abort message

update: note ways to avoid moving bookmark

push: prevent moving phases outside of the pushed subset

discovery: ensure that missingheads are always heads of everything we tried This only alter the third block executed when: * repo have secret changeset * onlyheads is not None In other case findcommonoutgoing already had this behavior

phase: extracts heads computation logics from analyzeremotephases

phase: fix warning text from invalid remote phase add missing end of line and substitution.

mq-safety: don't apply safety on non-outgoing changeset When mq changeset are secret, they don't appear in outgoing and won't be pushed. So it's not necessary to abort the push. The checkpush call is protected by lock to prevent race on phase.

phases: properly register excluded changeset when revision are specified There was no other drawback than a bad UI message for now.

bookmarks: make -i summary simpler and more accurate

revset: allow slashes in symbols suggested by Ryan Kelly

phases: do not mark debug message for translation

templatekw: fix phase keywords

phases: fix typo in warning message

rebase: drop uppercase in abort message

largefiles: add --normal option to hg add (issue3061)

largefiles: fix caching largefiles from an aliased repo (issue3212)

run-tests: expand user in --with-hg This makes it possible to run: $ ./run-tests.py --with-hg=~/hg-bin/hg

tests: let run-tests.py default to use 'sh' in $PATH instead of '/bin/sh' This makes it easier to run tests on systems that have a usable sh in the search path but not in the standard location.

tests: add 'set -x' to the .t sh scripts in run-tests.py debug mode This makes -d output much more readable when debugging the test framework or very strange test failures.

bookmarks: clarify help for -i/--inactive

revset: fix typo in message

merge default into stable for 2.1 code freeze

revset: add remote() predicate to lookup remote revisions

bookmarks: automatically advance bookmark on naked update (BC) (issue2894)

qimport: back out 2eec74d7ce95 This patch presumed mq csets must always be secret, which isn't yet established.

phases: only synchronize on common changeset when push fails If push failed we should not expect the pushed changeset to exist on remote. The common set before the push is used for phase related operation instead of common + missing. Note: * We still pull phase data even if push fails * We still try to push data even if push fails (same than bookmark)

discovery: diet discovery.prepush from non-discovery code The ``discovery.prepush`` function was doing multiple things not related to discovery. This changeset move some code into the ``localrepo.push`` method. The old ``discovery.prepush`` function jobs is now restricted to checking for multple head creation. It was then renamed ``discovery.checkheads``. This new ``discovery.checkheads`` function may receive several other changes in the future but we are a bit too much near the freeze for a wider refactoring.

qfinish: set all qfinished patch as draft, not only qbase Fix a silly bug.

largefiles: test and simplify empty directory removal in remove

tag: invalidate tag cache immediately after adding new tag (issue3210) New tags were written to .hgtags / .hglocaltags without updating or invalidating the localrepo cache. Before afd459933d5f a lock was acquired soon after the new tags had been written, and that invalidated the cache so the new tags for example could be seen in pretxncommit hooks. With afd459933d5f the lock had already been acquired at this point and the missing cache invalidation was exposed. The tag caches will now explicitly and immediately be invalidated when new tags are added.

mq: ensure mq changesets are set to secret when no phase data are found

mq: qimporting revision set them to secret

mq: have mq create secret changeset only

wireproto: refuse to lookup secret csets

qclone: ignore local qbase if secret

rebase: only advance phase on successful commit

phases: mechanism to allow extension to alter initial computation of phase This commit add a whennodata list where extension can register a callback to be called if no phase related data are found in the repository. The goal is to ensure the existing extension that move phase data in 2.1 can compute consistent phase boundary for existing repo.

qclone: add a few comment and blank line This help readability.

mq: turn changeset draft on qfinish (except if qparent is secret) As mq automatically sets changesets as secret, it should make them draft when he is done with it. We do not move them automatically to draft when we detect that something else have also set them as secret through their parents.

config: have a way to backup and restore value in config This is introduce to allow temporary overwriting of a config value while being able to reinstall the old value once done. The main advantage over using ``config`` and ``setconfig`` is that backup and restore will properly restore the lack of any config. Restoring the fact that there was no value is important to allow config user to keep using meaniful default value. A more naive approach will result in the following scenario:: Before: config(section, item, my_default) --> my_default temporal overwrite old = config(section, item) … setconfig(section, item, old) After config(section, item, my_default) --> None The first user of this feature should be mq to overwriting minimal phase of future commit.

graft: add test to check the phase of new changesets Graft actually do not keep the phase of the source changesets: no need to change the graft code and add test to check this behaviour.

rebase: fix phases movement Rebase now try to keep the phases of source changesets.

largefiles: cache new largefiles for new heads when pulling When the user pulls from a remote repository that is not his default repo, it is quite likely that he will pull a new head. This means that if he tries to merge or rebase with the other head, he will run into a problem becuase largefiles has no way of tracking where the remote repository for this other head is, so it cannot download the largefiles from this other remote repository. It will attempt to download them from its default remote repository, which will not yet contain the largefiles. This patch solves this problem by caching any new largefiles for all heads directly into the system cache at the time of the pull, so they are available later. This behavior is actually more in line with Mercurial's distributed nature, because pulling already implies we have a connection to the remote server, but merging or rebasing does not.

largefiles: remove unused lfutil.readhash()

largefiles: check if largefile could be found when archiving (issue3193)

largefiles: refactor lfutil.findfiles to be more logical

forget: fix subrepo recursion for explicit path handling When support for handling explicit paths in subrepos was added to the forget command (95174c381525), subrepo recursion wasn't taken into account. This change fixes that by pulling the majority of the logic of commands.forget into cmdutil.forget, which can then be called from both there and subrepo.forget.

add: fix subrepo recursion for explicit path handling When support for handling explicit paths in subrepos was added to the add command (9e99d2bbb1b1), subrepo recursion wasn't taken into account. This change adds an explicitonly argument to cmdutil.add to allow controlling which levels of recursion should include only explicit paths versus all matched paths.

tests: add subrepo recursion tests for add/forget with explicit paths When support for handling add/forget of explicit paths within subrepos was added (9e99d2bbb1b1/95174c381525), nested subrepos weren't handled properly. This change adds test coverage to expose the broken behavior, which will be fixed in later patches.

merge with stable

bookmarks: backout locking change in 12dea4d998ec Repo lock only applies to data in .hg/store/

changeset_printer: display changeset phase on debug level Backward compatibility make it hard to display it on higher level

phase: report phase movement When used in "set" mode, the phase command now display the number of changeset who changed phase.

largefiles: correctly handle newly added largefile on other side of merge

rebase: write series file without removed mq patches Rebase will remove empty changesets and will also completely remove the mq patch file for rebased empty patches. Starting with b28004513977 (1.9) it would preserve guards by writing the old series file back. That would however also reintroduce removed patch files in the series file and the inconsistency would make qpop + qpush fail. This patch backs out most of b28004513977 and makes sure guards are preserved without reintroducing removed patches.

revset: roots needs to be computed on full set

phases: use nodemap to check for missing nodes

repair: move phases code after invalidation code

largefiles: remove empty directories upon update (issue3202)

revset: optimize roots and children

revset: optimize building large lists in formatrevspec The large or-expressions we used to build required a substantial amount of subset filtering in orset() which was inefficient. Instead we build a single string which we process in one go with a special internal predicate.

hg-ssh: use shlex for shell-like parsing of SSH_ORIGINAL_COMMAND The Mercurial ssh protocol is defined as if it was ssh-ing to a shell account on an ordinary ssh server, and where hg was available in $PATH and it executed the command "hg -R REPOPATH serve --stdio". The Mercurial ssh client can in most cases just pass REPOPATH to the shell, but if it contains unsafe characters the client will have to quote it so the shell will pass the right -R value to hg. Correct quoting of repopaths was introduced in d8fa35c28335 and tweaked in 86fc364ca5f8. hg-ssh doesn't create the command via a shell and used a simple parser instead. It worked fine for simple paths without any quoting, but if any kind of quoting was used it failed to parse the command like the shell would do it. This makes hg-ssh behave more like a normal shell with hg in the path would do.

hooks: prioritize run order of hooks As of Mercurial 1.3, hooks are sorted in the order they are read into Mercurial. There are many instances when someone may want the hooks sorted in a specific order; this patch allows prioritizing hooks, while maintaining the existing enumeration for hooks without a priority.

merge with stable

merge: report all files in _checkunknown When doing hg up, if there is a file conflict with untracked files, currently only the first such conflict is reported. With this patch, all of them are listed. With this patch error message is now reported as a: untracked file differs b: untracked file differs abort: untracked files in working directory conflict with files in requested revision instead of abort: untracked file in working directory differs from file in requested revision: 'a' This is a follow up to an old attempt to do this here: http://selenic.com/pipermail/mercurial-devel/2011-August/033625.html

discovery: fix prepush documentation The prepush documentation claim that when we refuse to push, the second element of the returng tuple is an "outgoing" integer. value should be 0 when no outgoing changeset and 1 otherwise. In pratice if there are no outgoing changeset, "outgoing" value is alway 1 and util.Abort is raised on other error. the stable branch also include this error so it wasn't introduced by a recent refactoring.

phases: simplify phase exchange and movement over pushkey The code now only exchange draft root and only care about movement related to public//draft boundary. There is multiple reason to simplify this code: * Secret are never discovered anymore * We decided to not support more the three existing phase Removing phase index from pushkey (if ever decided) will be made in another commit.

phases: move phase according what was pushed not only what was added added This fix the lack phase movement when a locally secret changeset without added children was pushed to the repository. In such case, this changeset would be present in the bundle source, but not in the ``added`` variable.

revlog: make addgroup returns a list of node contained in the added source This list will contains any node see in the source, not only the added one. This is intended to allow phase to be move according what was pushed by client not only what was added.

phases: make secret changeset undiscoverable in all case This apply the redefined stronger semantic of secret. Secret changeset can still leak in various way. Those leak will need to be fixed individualy

phases: exclude secret when doing a local clone This is achieved by denying copy clone when any secret changeset exist.

bookmarks: primarily use repo lock, not wlock Bookmarks are repository data, not working directory data. Only the current bookmark is working directory data. Some lock shuffling is required to avoid lockout between the initial mock lock and locking of the localrepo instance that is created after copying.

localrepo: update branchcache in a more reliable way test-mq-cache.t did apparently look at stale cache content. Testing with different locking mechanism happened to update the cache more frequently and thus caused a test failure.

tests: make qnew in test-mq-caches.t stable The unstable hashes did for other reasons not show up in the test output.

localrepo: remove spurious updatebranchcache return value - it was left over from the refactoring in d01e28657429.

mq: only save dirty files once when savedirty is called multiple times

mq: remove early wlock release in qnew It seems like something that missed the refactoring in 97b734fb9c6f.

mq: use .invalidate to cancel dirty mq state when cancelling transaction Before the code optimistically relied on savedirty not being called a cancelled transaction. If it was called it could save incorrect data. Instead we now start using the invalidate method introduced in b169ba60eebe.