largefiles: deprecate --all-largefiles for pull The same can be achieved with --lfrev pulled() and we shouldn't advertise unnecessary command line options.

largefiles: implement pull --all-largefiles as a special case of --lfrev

largefiles: drop --cache-largefiles again This goes a step further than d69585a5c5c0 and backs out the unreleased --cache-largefiles option. The same can be achieved with --lfrev heads(pulled()) and we shouldn't introduce unnecessary command line options.

largefiles: introduce pulled() revset expression for use in --lfrev This provides a general way to do what already can be done with --all-largefiles and --cache-largefiles.

largefiles: introduce pull --lfrev option The revset will be evaluated after the changesets has been pulled, and missing largefiles from matching revisions will be pulled to the local caches. This in combination with revsets will make it possible to specify different strategies for pulling largefiles. The revset expressions used for this option might be quite complex and will probably be most useful from scripts or an alias ... but less complicated than configuring hooks.

largefiles: refactor overridepull internals

largefiles: introduce lfpull command for pulling missing largefiles

largefiles: update help Some clarifications, and some clean-up after --cache-largefiles was introduced.

largefiles: fix cat of non-largefiles from subdirectory We were calling back to the original commands.cat from inside the walk loop that handled and filtered out largefiles. That did however happen with file paths relative to repo root and the original cat would fail when it applied its own walk and match on top of that. Instead we now duplicate and modify the code from commands.cat and patch it to handle both normal and largefiles. A change in test output shows that this also makes the exit code with largefiles consistent with the normal one in the case where one of several specified files are missing. This also fixes the combination of --output and largefiles.

largefiles: don't store whole file in memory for 'cat'

mergetools: rename 'base' to 'merged' in meld This makes it clear which panel is the target of the merge operation.

mergetools: avoid losing the merged version with meld Add -o $output. When using Meld as intended (merge from left and right into the center panel), the merged version is written to the wrong file without this option ($base, a temporary file, which is ignored by Mercurial). Add meld.check=changed as a secondary safety net.

templatekw: add default styles for hybrid types (issue3887) This allows elements like file_copies to be printed as 'name (source)' when used with join.

largefiles: improve repo wrapping detection Before this patch, repo wrapping detection in "reposetup()" of largefiles can detect only limited repo wrapping: replacing target functions by another one named as "wrap". So, it can't detect repo wrapping even in recommended style: replacing "__class__" of repo by derived class. This patch can detect repo wrapping in both styles below: - replacing "__class__" of repo by derived class (recommended style): class derived(repo.__class__): def push(self, *args, **kwargs): return super(derived, self).push(*args, **kwargs) repo.__class__ = derived - replacing function of repo by another one (not recommended style): orgpush = repo.push def push(*args, **kwargs): return orgpush(*args, **kwargs) repo.push = push

hgweb: respond HTTP_NOT_FOUND when an archive request does not match any files

archive: raise error.Abort if the file pattern matches no files Note that we could raise this exception even if no pattern were specified, but the revision contained no files. However this should not happen in practice since in that case commands.py/archive would exit earlier with an "no working directory: please specify a revision" error message instead.

ui: add 'force' parameter to traceback() to override the current print setting This will allow a current traceback.print_exc() call in dispatch to be replaced with ui.traceback() even if --traceback was not given on the command line.

ui: add support for fully printing chained exception stacks in ui.traceback() Currently, only SubrepoAbort has a cause chained to it.

subrepo: chain the original exception to SubrepoAbort The tracebacks in subrepos are truncated at the point where the original exception is caught and SubrepoAbort is raised in its place since 9e3910db4e78. That hides the most relevant subrepo methods when an error occurs. Python 2.x doesn't support chaining exceptions, so it is manually done here for manual printing later.

debugrebuildstate: rename to debugrebuilddirstate There is a lot of state, but this command is for rebuilding the dirstate.

debugstate: rename to debugdirstate There is a lot of state, but this command is for debugging the dirstate.

debugrebuildstate: clarify that rev can't be specified without -r -r has a default value of '' in the command line. The function default value of 'tip' is thus never used and any attempt at specifying revisions without -r will fail. It seems like then intended behavior was that 'hg debugrebuildstate' without any parameters should set the parents to tip. That would be very confusing now when the command primarily is used to recover from incorrect stat info. It is apparently undocumented that '' is the same as '.' ... unless it is passed in a place where revsets are used.

check-code: check txt files for trailing whitespace

check-code: catch trailing space in comments

spelling: fix typos and spelling errors

wireproto: clarify cryptic 'remote: unsynced changes' error message on push The message was not very much to the point and did not in any way help an ordinary user. 'repository changed while preparing/uploading bundle - please try again' is more correct, gives the user some understanding of what is going on, and tells how to 'recover' from the situation. The 'bundle' aspect could be seen as an implementation detail that shouldn't be mentioned, but I think it helps giving an exact error message. The message could still leave the user wondering why Mercurial doesn't lock the repo and how unsafe it thus is. Explaining that is however too much detail.

export: export working directory parent by default A common usecase for export is to preview the patch that will be patchbombed or to see what changed in a revision found by bisect. Showing the working directory parent is thus a useful and obvious default.

commit: allow closing "non-head" changesets Backout acd61dc44a39. The changeset prevented closing non-head changesets but did not provide any rationale or test case and I don't see what value it adds. Users might have their reasons to commit something anywhere - and close it immediately. And contrary to the comment that is removed: The topo heads set is _not_ included in the branch heads set of the current branch. It do not include closed topological heads. The change thus prevented closing commits on top of closing commits. A valid usecase for that is to merge closed heads to reduce the number of topological heads. The only existing test coverage for this is the failing double close in test-revset.t. It was added in 3cc2e34d7a7d and seems to not be intentional.

record: ignore trailing content when parsing patches - introduce 'other' lines This makes record work more like import which ignores for instance mail footers in a patch file. This also makes it possible for TortoiseHg to preview unapplied patches containing such footers.

record: abort on malformed patches instead of crashing

localrepo: use "vfs.rename()" instead of "util.rename()" This patch makes "_journalfiles()" return a list of pairs of journal file and corresponded vfs instance instead of a list of journal files in full path, to use "vfs.rename()" instead of "util.rename()" in "aftertrans()". "undofiles()" still returns a list of undo files in full path, because "repair.strip()" expects such list. It'll be also made to return a list of pairs of undo file and corresponded vfs at vfs migration for "repair.strip()" in the near future.

localrepo: use "vfs.setflags()" instead of "util.setflags()"

localrepo: use "vfs.readlink()" instead of "os.readlink()"

localrepo: use "vfs.islink()" instead of "os.path.islink()"

localrepo: use "vfs.rename()" instead of "util.rename()"

localrepo: use "vfs.exists()" instead of "os.path.exists()"

localrepo: use vfs instead of "os.path.*" in sharedpath checking In the point of view of efficiency, "vfs" instance created in this patch should be passed to and reuse in "store.store()" invocation just after patched code block, because "store" object is initialized by vfs created with "self.sharedpath". eBut to focus just on migration from direct file I/O API accessing to vfs, this patch uses created vfs as temporary one. Refactoring around "store.store()" invocation will be done in the future.

vfs: split "expand" into "realpath"/"expandpath" to apply each separately Before this patch, vfs constructor applies both "util.expandpath()" and "os.path.realpath()" on "base" path, if "expand" is True. This patch splits it into "realpath" and "expandpath", to apply each functions separately: this splitting can allow to use vfs also where one of each is not needed.

unionrepo: read-only operations on a union of two localrepos unionrepo is just like bundlerepo without bundles. The implementation is very similar to bundlerepo, but I don't see any obvious way to generalize it. Some most obvious use cases for this would be log and diff across local repos, as a kind of preview of pulls, for instance: $ hg -R union:repo1+repo2 heads $ hg -R union:repo1+repo2 log -r REPO1REV -r REPO2REV $ hg -R union:repo1+repo2 log -r '::REPO1REV-::REPO2REV' $ hg -R union:repo1+repo2 log -r 'ancestor(REPO1REV,REPO2REV)' $ hg -R union:repo1+repo2 diff -r REPO1REV -r REPO2REV This is going to be used in RhodeCode, and Bitbucket already uses something similar. Having a core implementation would be beneficial.

subrepo: fix exception on revert when "all" option is omitted Since fafdff7e9c43, backout does not set opts['all'], which causes KeyError at hgsubrepo.revert.

histedit: update tests for new intervention framework

test-subrepo: add tests for subrepo "storeclean" checks These tests verify that subrepos are not pushed when their store is clean versus a given target repository.

subrepo: do not push mercurial subrepos whose store is clean This patch stops mercurial from pushing unmodified subrepos. An unmodified subrepo is one whose store is "clean" versus a given target subrepo. Note that subrepos may have a clean store versus a target repo but not versus another. This patch handles this scenario by individually keeping track of the state of the store versus all push targets. Tests will be added on the following revision.

subrepo: implement "storeclean" method for mercurial subrepos The mercurial subrepo "storeclean" method works by calculating a "store hash" of the repository state and comparing it to a cached store hash. The store hash is always cached when the repository is cloned from or pushed to a remote repository, but also on pull as long as the repository already had a clean store. If the hashes match the store is "clean" versus the selected repository. Note that this method is currenty unused, but it will be used by a later patch. The store hash is calculated by hashing several key repository files, such as the bookmarks file the phaseroots file and the changelog. Note that the hash comparison is done file by file so that we can exit early if a pair of hashes do not match. Also the hashes are calculated starting with the file that is most likely to be smaller upto the file that is more likely to be larger.

util: add notindexed optional parameter to makedirs function

subrepo: introduce storeclean method Currently this method is unused and it is not implemented for any specific subrepo type (it always returns False). It receives a remote repository path because a repository may have a clean store versus a given repository but not versus another.

subrepo: introduce storeclean helper functions These helper functions are currently unused but will be used to implement the cleanstore method that will be introduced later.

dispatch: exit with status 1 for an InterventionRequired exception (bc)

histedit: switch from util.Abort to util.InterventionRequired where appropriate (bc)

rebase: switch from util.Abort to util.InterventionRequired where appropriate (bc)

dispatch: catch InterventionRequired and print the message with no prefix

error: introduce new InterventionRequired exception Future changes will use this type instead of util.Abort for signalling the user that intervention is required, as in some rebase and histedit aborts.

util: remove unreachable code Found by Cython.

util: remove no-op assignment Found by Cython.

merge more evolve stuff away (what?)

synthrepo: add missing import of sys Found using Cython.

transplant: pass source through to recover

repoview: remove more unreachable code Found using Cython.

merge away some failed evolve fat-fingering

pager: add missing imports of errno and shlex Found using Cython.

largefiles: quiet (and document) undefined name errors (issue3886) These names were found using Cython; I was completely puzzled until I searched the rest of the tree. It's icky to mess with another module's namespace, but ickier yet to do so without a comment :-)

color: add missing import of error Found using Cython.

pvec: use the correct name for an identifier Found using Cython.

repoview: remove unreachable code Found using Cython.

pvec: use the correct name for an identifier Found using Cython.

repoview: remove unreachable code Found using Cython.

mail: add missing import of sys Found using Cython.

statichttprepo: add missing import of os Found using Cython.

worker: add missing import of errno Found using Cython.

histedit: support editing of the first commit (issue3767)

test-commit-amend.t: fix check-code violation from 3a72c89a83ec

util: improve doc for checkcase

scheme: don't crash on invalid URLs

amend: support amending merge changesets (issue3778)

worker: catch all exceptions, try to exit usefully/safely

lock: if we fork, ensure that only the parent releases This prevents us from having a bunch of errant worker processes all try to release a lock if a problem occurs. (Releasing the lock more than once is harmless; it's invoking the associated callbacks we want to avoid.)

tag: clarify cryptic error message when tagging null revision

setup: make error message for missing Python headers more helpful

obsolete: ensure all markers have a date Obsolescence creates a sparse DAG mostly composed of a lot of small independent chain of markers. Date is the only simple and "reliable" way to sort them. The existence of a date is now enforced at creation time as I'm more and more convinced that date will have a key role in obsolescence markers exchange.

zsh_completion: make use of `debuglabelcomplete` command Use `debuglabelcomplete` command when populating labels list, instead of calling `hg` three times: for branches, bookmarks and tags. Do not pass string being completed to `hg debuglabelcomplete` (as `$words[$CURRENT]`), since it breaks `_hg_revrange` completion (`--rev 2.5:2.5.<Tab>`) for no apparent benefit. Also complete `hg view` with labels, not just tags, and drop unused `_hg_tags`.

dirs: use mutable strings internally perfdirs results for a working dir with 170,000 files: Python 638 msec C 244 C+int 192 C+int+str 168 In the large repo above, the nearly 0.5 second time improvement is visible in commands like "hg add" and "hg update". hg add Python 1100 msec C+int+str 600 hg update (with nothing to do) Python 2800 msec C+int+str 2240

dirs: use mutable integers internally These integers are not visible to Python code, so this is safe. perfdirs results for a working dir with 170,000 files: Python 638 msec C 244 C+int 192

scmutil: rewrite dirs in C, use if available This is over twice as fast as the Python dirs code. Upcoming changes will nearly double its speed again. perfdirs results for a working dir with 170,000 files: Python 638 msec C 244

scmutil: use new dirs class in dirstate and context The multiset-of-directories code was open coded in each of these modules; this change gets rid of the duplication.

scmutil: add a dirs class This encapsulates the "multiset of directories" structures that are currently open-coded (and duplicated) in both the dirstate and context modules. This will be used, and optionally replaced by a C implementation, in upcoming changes.

scmutil: migrate finddirs from dirstate

merge

manifestmerge: handle workdir removed, remote removed with flags This can happen when a file with flags is removed or deleted in the working directory and also not present in m2. The obvious solution is to add a __delitem__ override to manifestdict that removes the file from flags if necessary, but that has a significant performance cost in some cases, e.g. hg status --rev rev1 --rev rev2 <file>.

dicthelpers.diff: compare against default for missing values This is not only a bit faster, but also aligns with callers' expectations better since we can legitimately have manifestdict's _flags set to '' instead of unset. hg perfmergecalculate -r . before: ! wall 0.139582 comb 0.140000 user 0.140000 sys 0.000000 (best of 59) after: ! wall 0.126154 comb 0.120000 user 0.120000 sys 0.000000 (best of 74) hg perfmergecalculate -r .^ before: ! wall 0.236333 comb 0.240000 user 0.240000 sys 0.000000 (best of 36) after: ! wall 0.212265 comb 0.210000 user 0.210000 sys 0.000000 (best of 45)

templater: back out 0615b22da148, it breaks schemes ({1})

debuglabelcomplete: compute active branch heads correctly The previous computation was simply wrong.

merge with i18n Note: i18n work should normally be done on stable

hg-i18n-it: minor fixes

templater: fix check-code error

smtp: verify the certificate of the SMTP server for STARTTLS/SMTPS Before this patch, the certificate of the SMTP server for STARTTLS or SMTPS isn't verified. This may cause man-in-the-middle security problem (stealing authentication information), even though SMTP channel itself is encrypted by SSL. When "[smtp] tls" is configured as "smtps" or "starttls", this patch: - uses classes introduced by preceding patches instead of "SMTP" or "SMTP_SSL" of smtplib, and - verifies the certificate of the SMTP server, if "[smtp] verifycert" is configured as other than False "[smtp] verifycert" can be configured in 3 levels: - "strict": This verifies peer certificate, and aborts if: - peer certification is not valid, or - no configuration in "[hostfingerprints]" and "[web] cacerts" This is default value of "[smtp] verifycert" for security. - "loose": This verifies peer certificate, and aborts if peer certification is not valid. This just shows warning message ("certificate not verified"), if there is no configuration in "[hostfingerprints]" and "[web] cacerts". This is as same as verification for HTTPS connection. - False(no verification): Peer certificate is not verified. This is as same as the behavior before this patch series. "hg email --insecure" uses "loose" level, and ignores "[web] cacerts" as same as push/pull/etc... with --insecure. Ignoring "[web] cacerts" configuration for "hg email --insecure" is already done in "dispatch._dispatch()" by looking "insecure" up in the table of command options.

sslutil: abort if peer certificate is not verified for secure use Before this patch, "sslutil.validator" may returns successfully, even if peer certificate is not verified because there is no information in "[hostfingerprints]" and "[web] cacerts". To prevent from sending authentication credential to untrustable SMTP server, validation should be aborted if peer certificate is not verified. This patch introduces "strict" optional argument, and "sslutil.validator" will abort if it is True and peer certificate is not verified.