py3: use list of bytes rather than bytestring while extending bytes into lists Python 2: >>> ls = [] >>> ls.extend(b'abc') >>> ls ['a', 'b', 'c'] Python 3: >>> ls = [] >>> ls.extend(b'abc') >>> ls [97, 98, 99] So to make sure things work fine in Py3, this patch does: >>> ls = [] >>> ls.extend([b'a', b'b', b'c']) >>> ls [b'a', b'b', b'c'] After this `hg log -G` works!

py3: use pycompat.byteskwargs to converts kwargs to bytes baseformatter._item must contain both keys and values in bytes. So to make sure that, we convert the opts back to bytes.

py3: make adefaults keys str to be compatible with getattr getattr passes a str value of the attribute to be looked and keys in adefaults dict are bytes which resulted in AttributeError. This patch abuses r'' to make the keys str.

py3: abuse r'' to access keys in keyword arguments

py3: use pycompat.bytechr instead of chr

py3: use %d to format integers into bytestrings

py3: use pycompat.bytestr instead of bytes

py3: slice over bytes to prevent getting ascii values

py3: use encoding.unitolocal instead of .encode(encoding.encoding)

rebase: use matcher to optimize manifestmerge The old merge code would call manifestmerge and calculate the complete diff between the source to the destination. In many cases, like rebase, the vast majority of differences between the source and destination are irrelevant because they are differences between the destination and the common ancestor only, and therefore don't affect the merge. Since most actions are 'keep', all the effort to compute them is wasted. Instead, let's compute the difference between the source and the common ancestor and only perform the diff of those files against the merge destination. When using treemanifest, this lets us avoid loading almost the entire tree when rebasing from a very old ancestor. This speeds up rebase of an old stack of 27 commits by 20x. In mozilla-central, without treemanifest, when rebasing a commit from default~100000 to default, this speeds up the manifestmerge step from 2.6s to 1.2s. However, the additional diff adds an overhead to all manifestmerge calls, especially for flat manifests. When rebasing a commit from default~1 to default it appears to add 100ms in mozilla-central. While we could put this optimization behind a flag, I think the fact that it makes merge O(number of changes being applied) instead of O(number of changes between X and Y) justifies it.

changegroup: delete unused 'bundlecaps' argument (API)

localrepo: reuse exchange.bundle2requested() It seems like localrepo.getbundle() is trying to do the same thing, so let's just call the method. That way we get the same condition as there (matching any "HG2" prefix, not only "HG20").

py3: use raw strings while accessing class.__dict__ The keys of class.__dict__ are unicodes on Python 3.

py3: handle opts correctly for `hg add` opts in add command were passed again to cmdutil.add() as kwargs so we need to convert them again to str. Intstead we convert them to bytes when passing scmutil.match(). Opts handling is also corrected for all the functions which are called from cmdutil.add().

py3: handle opts correctly for rollback dryrun and force are just check for None, the value is not used. So its better to leave opts as unicodes as that wont harm us.

py3: handle opts correctly for unbundle There is just a check whether the value is None or not. So even having optupdate as unicodes won't harm us.

py3: convert opts to bytes in cmdutil.dorecord() commands.commit() calls cmdutil.dorecord() where opts are passed as unicodes being keyword arguments. This patch converts them back to bytes as they are required.

py3: make sure opts are passed and used correctly in help command opts are converted back to bytes in help.help_() where they are used. Before that it's ensured that we have a bytes value for keep variable.

py3: handle opts uniformly in commands.py Since keyword arguments can't be bytes on Python 3, we converted then to unicodes before passing into different command functions. We need to adopt a certain pattern to convert opts back to bytes. Following are some of the functions which are called from inside these command functions and should always be feeded bytes to follow the right behaviour. ui.fomattter() scmutil.match() patch.diffallopts() hg.peer() cmdutil.{show_changeset|copy|graphrevs|checkunsupportedflag} server.{createservice|runservice} There are few commands which are left out where opts is again passed to a function as keyword arguments or converting opts back to bytes is kind of not necessary. Those are cat, revert, help, unbundle and rollback. Following patches will deal with them. This patch apart from these five commands, convert opts back to bytes for rest of the commands. This fixes a lot of things which are hidden like --git works now. Similarly more flags of commands which run on Python 3 currently get fixed.

merge with stable

help: call out specific replacement configuration settings As an aside, I'm having trouble parsing the help text meaning for HG when it is unset or empty. How can it be the frozen name or searched if it is empty?

help: spelling fixes

help: attempt to clarify that pager usage is not output length based This may be too subtle of a change to get the point across, but when I first read the original text, I thought maybe the pager would only be invoked if writing more than a screenful. The distinction between this and a pager that simply exits after printing less than a screenful is important on Windows, given the inability of `more` to color output.

help: document color/pager pitfalls on Windows Even though I figured this out a few weeks ago, I was initially puzzled where the color went when I upgraded to 4.2 on a different Windows machine. Let's point users reading the help into the right direction. I wonder if we should be even more explicit about cmd.exe/MSYS/pager/color interplay, but at least all of the breadcrumbs are here (I think).

webcommands: use fctx.isbinary

annotate: use fctx.isbinary

fileset: use fctx.isbinary instead of util.binary(fctx.data()) filectx provides "isbinary" to test if the data is binary. Let's use it. This enables other filectx implementations (like LFS) to override the isbinary test.

internals: document that "branches" is a legacy wire command Modern clients use a different discovery mechanism.

match: make subinclude construction lazy The matcher subinclude functionality allows us to have .hgignore files that include subdirectory hgignore files. Today it parses the entire repo at once, even if we only need to test a file in one subdirectory. This patch makes the subinclude tree creation lazy, which speeds up matcher creation significantly in large repos with very large trees of ignore patterns.

bisect: allow resetting with unfinished graft/rebase/etc "hg bisect --reset" just deletes the state file (it doesn't move back to the starting point like rebase does); it can not conflict with an ongoing rebase etc. checkunfinished() has this documentation: It's probably good to check this right before bailifchanged(). So that's where I moved it.

py3: use %d instead of %s for integers dispatch._runcatch() always returns an integer value.

py3: make posix.getuser return a bytes

py3: replace str with bytes in isinstance()

py3: use pycompat.bytestr() instead of str() This is because str() on python 3 return unicodes

py3: alias long to int on Python 3

setup: sys.version_info always exists It is documented as existing since Python 2.0, and empirically from other OSS projects I maintain, there is no problem relying on its existance.

filelog: fix parsemeta docstring 75bb7c702317 changed the return type of filelog.parsemeta but forgot to update its docstring.

util: remove doc of long gone 'targetsize' argument Gone since fa836e050c50 (chunkbuffer: removed unused method and arg, 2007-10-11).

test-diff-color: disable pager for expected output on Windows (issue5555) Windows uses `more.com`, which unhelpfully adds an extra trailing line consisting only of '\r'. It also converts tab characters to spaces, which throws off the last two tests. Setting the 'ui.formatted' option is what allowed the pager to be used by these tests in the first place.

cleanup: drop the deprecated 'localrepo._link' method This was deprecated in favor of 'localrepo.wvfs.islink'. We can now drop it for the future 4.3.

cleanup: drop the deprecated 'localrepo.wfile' method This was deprecated in favor of 'localrepo.wvfs.join'. We can now drop it for the future 4.3.

cleanup: drop the deprecated 'localrepo.join' method This was deprecated in favor of 'localrepo.vfs.join'. We can now drop it for the future 4.3.

cleanup: drop the deprecated 'localrepo.tag' method This was deprecated in favor of 'mercurial.tags.tag'. We can now drop it for the future 4.3.

cleanup: drop the deprecated 'localrepo.opener' method This was deprecated in favor of 'localrepo.vfs'. We can now drop it for the future 4.3.

cleanup: drop the deprecated 'localrepo.wopener' method This was deprecated in favor of 'localrepo.wvfs'. We can now drop it for the future 4.3.

cleanup: drop vfs compatibility layer in scmutil All these constructors are deprecated in 4.2. We can now drop them on the default branch (future 4.3).

test-worker: exercise more about "killworkers" situation This patch adds some sleep and increases numcpus to exercise the "killworkers" situation. So race conditions could be discovered more easily, if someone changes worker.py incorrectly. Currently worker.py should be free of such issues.

test-worker: capture tracebacks more reliably The traceback test may have traceback caused by SIGTERM. Instead of grepping "Traceback", explicitly grep the exception we care about. This makes the test less flaky.

worker: rewrite error handling so os._exit covers all cases Previously the worker error handling is like: pid = os.fork() --+ if pid == 0: | .... | problematic .... --+ try: --+ .... | worker error handling --+ If a signal arrives when Python is executing the "problematic" lines, an external error handling (dispatch.py) will take over the control flow and it's no longer guaranteed "os._exit" is called (see 86cd09bc13ba for why it is necessary). This patch rewrites the error handling so it covers all possible code paths for a worker even during fork. Note: "os.getpid() == parentpid" is used to test if the process is parent or not intentionally, instead of checking "pid", because "pid = os.fork()" may be not atomic - it's possible that that a signal hits the worker before the assignment completes [1]. The newly added test replaces "os.fork" to exercise that extreme case. [1]: CPython compiles "pid = os.fork()" to 2 byte codes: "CALL_FUNCTION" and "STORE_FAST", so it's probably not atomic: def f(): pid = os.fork() dis.dis(f) 2 0 LOAD_GLOBAL 0 (os) 3 LOAD_ATTR 1 (fork) 6 CALL_FUNCTION 0 9 STORE_FAST 0 (pid) 12 LOAD_CONST 0 (None) 15 RETURN_VALUE

dispatch: take over SignalInterrupt handling from scmutil dispatch handles KeyboardInterrupt already. This makes the code more consistent, and makes worker not print "killed!" if it receives SIGTERM in most cases (in rare cases there is still "killed!" printed, which will be fixed by the next patch).

merge stable into default

Added signature for changeset bb96d4a49743

Added tag 4.2 for changeset bb96d4a49743

merge with i18n

i18n-ja: synchronized with 6e0368b6e0bb

pager: drop the support for 'pager.enable=<bool>' This option was never released except for a release candidate. Dropping compatibility with this option will free the 'pager.enable' config option for other usage in the future.

pager: rename 'pager.enable' to 'ui.paginate' This aligns with what we do for color (see 7fec37746417). Pager is a central enough notion that having the master config in the [ui] section makes senses. It will helps with consistency, discoverability. It will also help having a simple and clear example hgrc mentioning pager. The previous form of the option had never been released in a non-rc version but we keep it around for convenience. If both are set, 'ui.pager' take priority.

color: special case 'always' in 'ui.color' This lift the confusing case, where 'ui.color=always' would actually not always use color.

color: turn 'ui.color' into a boolean (auto or off) Previously, 'ui.color=yes' meant "always show color", While "ui.color=auto" meant "use color automatically when it appears sensible". This feels problematic to some people because if an administrator has disabled color with "ui.color=off", and a user turn it back on using "color=on", it will get surprised (because it breaks their output when redirected to a file.) This patch changes ui.color=true to only move the default value of --color from "never" to "auto". I'm not really in favor of this changes as I suspect the above case will be pretty rare and I would rather keep the logic simpler. However, I'm providing this patch to help the 4.2 release in the case were others decide to make this changes. Users that want to force colors without specifying --color on the command line can use the 'ui.formatted' config knob, which had to be enabled in a handful of tests for this patch. Nice summary table (credit: Augie Fackler) That is, before this patch: +--------------------+--------------------+--------------------+ | | not a tty | a tty | | | --color not set | --color not set | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color (not set) | no color | no color | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color = auto | no color | color | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color = yes | *color* | color | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color = no | no color | no color | | | | | +--------------------+--------------------+--------------------+ (if --color is specified, it always clobbers the setting in [ui]) and after this patch: +--------------------+--------------------+--------------------+ | | not a tty | a tty | | | --color not set | --color not set | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color (not set) | no color | no color | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color = auto | no color | color | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color = yes | *no color* | color | | | | | +--------------------+--------------------+--------------------+ | [ui] | | | | color = no | no color | no color | | | | | +--------------------+--------------------+--------------------+ (if --color is specified, it always clobbers the setting in [ui])

pager: document the 'pager.enable' option The 'config' helps was missing help about pager enabling/disabling.

pager: advertise the config option in the default hgrc Same as for 'ui.color', this is a critical part of the UI and we want user to find this config knob easily.

pager: document the 'pager' config section There as a 'hg help pager' section but the 'hg help config.pager' was missing.

pager: test the 'enable' config option While poking at this option I realised it was not tested.

config: drop pager from the recommended extension The extension is deprecated, we should having exposing it to users.

config: use "churn" as an example extension "Churn" is not the useful example we have, but this is the one used in 'hg help config.extensions'. As we need something to replace the deprecated 'pager' extension in the example config, we are adding 'churn'.

discovery: prevent crash caused by prune marker having no parent data If a marker has no parent information, parents field is set to None, which caused TypeError. I think this shouldn't normally happen, but somehow I have such marker in my repo.

color: point to the global help in the example hgrc This is probably the best way to have users learn more is they need too.

color: reflect the new default in the example hgrc Color is enabled by default so un-commenting the line would not have any effect. We'll point to the help in the next changeset.

color: point to the config help in global help topic We point out at the help of the config option for user who wants to learn more about the possible config value. The suggested command returns some other extra (color related) results, but this is bug to fix outside of the freeze.

color: reflect the new default in global help topic We point out that color is enabled by default.

docs: describe ui.color consistently with --color The --color option is described to accept "boolean, always, auto, never, or debug". Let's use a similar description for ui.color. Also fix the formatting to use double quotes as we seem to use for about half the values in config.txt (the other half uses double backticks). Also use uppercase Boolean for consistency within the file.

test: glob out variation from 'HGPORT' length On system where HGTEST_PORT is configured to a value an order a magnitude lower or higher than the default. The number of digit in the HGPORT changes and this test breaks.

lock: avoid unintentional lock acquisition at failure of readlock Acquiring lock by vfs.makelock() and getting lock holder (aka "locker") information by vfs.readlock() aren't atomic action. Therefore, failure of the former doesn't ensure success of the latter. Before this patch, lock is unintentionally acquired, if ENOENT causes failure of vfs.readlock() while 5 times retrying, because lock._trylock() returns to caller silently after retrying, and lock.lock() assumes that lock._trylock() returns successfully only if lock is acquired. In this case, lock symlink (or file) isn't created, even though lock is treated as acquired in memory. To avoid this issue, this patch makes lock._trylock() raise LockHeld(EAGAIN) at the end of it, if lock isn't acquired while retrying. An empty "locker" meaning "busy for frequent lock/unlock by many processes" might appear in an abortion message, if lock acquisition fails. Therefore, this patch also does: - use '%r' to increase visibility of "locker", even if it is empty - show hint message to explain what empty "locker" means

lock: avoid unintentional lock acquisition at failure of readlock Acquiring lock by vfs.makelock() and getting lock holder (aka "locker") information by vfs.readlock() aren't atomic action. Therefore, failure of the former doesn't ensure success of the latter. Before this patch, lock is unintentionally acquired, if self.parentlock is None (this is default value), and lock._readlock() returns None for ENOENT at vfs.readlock(), because these None are recognized as equal to each other. In this case, lock symlink (or file) isn't created, even though lock is treated as acquired in memory. To avoid this issue, this patch retries lock acquisition immediately, if lock._readlock() returns None "locker". This issue will be covered by a test added in subsequent patch, because simple emulation of ENOENT at vfs.readlock() easily causes another issue. "raising ENOENT only at the first vfs.readlock() invocation" is too complicated for unit test, IMHO.

httppeer: unify hint message for PeerTransportError Another raising PeerTransportError for "incomplete response" in httppeer.py uses this (changed) hint message. This unification reduces cost of translation.

revset: add i18n comments to error messages for followlines predicate This patch also includes un-quoting "descend" keyword for similarity to other error messages (this seems too trivial as a separated patch).

help: apply bulk fixes for indentation and literal blocking issues There are some paragraphs, which aren't rendered in online help as expected because of indentation and literal blocking issues. - hgext/rebase.py - paragraph before example code ends with ":", which treats subsequent indented paragraphs as normal block => replace ":" with "::" to treat subsequent paragraphs as literal block - help/pager.txt - paragraph before a list of --pager option values ends with "::", which treats subsequent indented paragraphs as literal block => replace "::" with ":" to treat subsequent paragraphs as normal block - the second line of explanation for no/off --pager option value is indented incorrectly (this also causes failure of "make" in doc) => indent correctly - help/revisions.txt - explanation following example code of "[revsetalias]" section isn't suitable for literal block => un-indent explanation paragraph to treat it as normal block - indentation of "For example" before example of tag() revset predicate matching is meaningless - descriptive text for tag() revset predicate matching isn't suitable for literal block => un-indent concatenated two paragraphs to treat them as normal block

rebase: fix incorrect configuration example This configuration example doesn't make rebase require a destination, even though help document wants to show such example.

help: use hg role of mini reST to make hyper link in HTML page

help: use mercurial as a subject of colorization and pagination Now, colorization and pagination are in Mercurial core.

packaging: add make target for linux wheels Having linux wheels is going to helps system without compiler or python-dev plus speed up the installation for everyone. I followed the manylinux example repository https://github.com/pypa/python-manylinux-demo to add a make target (build-linux-wheels) using official docker image to build python 2 linux wheels for mercurial. It generates Python 2.6 and Python 2.7 for both 32 and 64 bits architectures. I had to blacklist several test cases for various reasons: * test-convert-git.t and test-subrepo-git.t because of the git version * test-patchbomb-tls.t because of warning using tls 1.0 It's likely because the docker image is based on centos 5.0 and openssl is outdated.

tests: fix two http tests to also pass inside manylinux1 docker This will let us build Linux wheels.

pager: use less as a fallback on Unix This seems reasonable choice per discussion, and the default-default of Git. See also the inline-comment for why. https://www.mercurial-scm.org/pipermail/mercurial-devel/2017-April/097042.html

test-bookmarks: factor hooks out to a shell script for Windows The previous incarnation of the hooks weren't being run on Windows, which altered the DAG and hashes[1]. [1] https://www.mercurial-scm.org/pipermail/mercurial-devel/2017-April/096987.html

help: explain how to access subtopics in internals

log: document the characters ---graph uses to draw The meaning of : vs | was undocumented and non-obvious.

tests: demonstrate that pager.attend-<abbreviated> doesn't work A pager.attend-* value that isn't a full command or alias name doesn't work. We add explicit test coverage of this to demonstrate it.

tests: test that abbreviated command alias is also paged Explicit test coverage is good.

tests: drop unnecessary pager attend in test `log` is attended by default. We don't need to specify pager.attend-log in this test.

hgweb: change text of followlines links to "older / newer" DAG directions "descending" / "ascending" arguably do not make much sense in the web interface where changes are usually listed by "dates".

hgweb: do not show "descending" link in followlines UI for filelog heads When on a filelog head, we are certain that there will be no descendant so the target of the "descending" link will lead to an empty log result. Do not display the link in this case.

context: optimize linkrev adjustment in blockancestors() (issue5538) We set parent._descendantrev = child.rev() when walking parents in blockancestors() so that, when linkrev adjustment is perform for these, it starts from a close descendant instead of possibly topmost introrev. (See `self._adjustlinkrev(self._descendantrev)` in filectx._changeid().) This is similar to changeset c82d88dfaf59, which added a "f._changeid" instruction in annotate() for the same purpose. However, here, we set _descendantrev explicitly instead of relying on the '_changeid' cached property being accessed (with effect to set _changeid attribute) so that, in _parentfilectx() (called from parents()), we go through `if '_changeid' in vars(self) [...]` branch in which instruction `fctx._descendantrev = self.rev()` finally appears and does what we want. With this, we can roughly get a 3x speedup (including in example of issue5538 from mozilla-central repository) on usage of followlines revset (and equivalent hgweb request).

record: update help message to use operation instead of "record" (issue5432) Update the hunk selector help message to use the operation name instead of using "record" for all operations. Extract the help message in the same way as other single and multiple message line. Update tests to make sure that both "revert" and "discard" variants are tested.

tests: add lots of regex matching to test-http-bad-server.t This makes the test pass in Python 2.6. It's all one of two problems: 1) abort: error: '' (2.7) vs abort: error: (2.6) 2) 65537 (2.7) vs -1 (2.6) With some...unfortunate escaping required due to how regular expressions work. Sigh.

tests: pass only one file at a time to tail(1) It looks like the verison on Solaris only tails one file at a time.

debian: configure editor and pager to sensible-\1 by default Now that environment variables override system-wide hgrc settings, we can default Mercurial to sensible-editor and sensible-pager by default for debian users.

tests: work around flaky test failure on FreeBSD and relatives test-http-bad-server has a flaky output case that reproduces fairly often (20% or more of runs on my Mac) on BSD-family OSes. So that tests can be green for the 4.2 release, let's more or less give up on this output, and plan to pursue it in the future.

context: start walking from "introrev" in blockancestors() Previously, calling blockancestors() with a fctx not touching file would sometimes yield this filectx first, instead of the first "block ancestor", because when compared to its parent it may have changes in specified line range despite not touching the file at all. Fixing this by starting the algorithm from the "base" filectx obtained using fctx.introrev() (as done in annotate()). In tests, add a changeset not touching file we want to follow lines of to cover this case. Do this in test-annotate.t for followlines revset tests and in test-hgweb-filelog.t for /log/<rev>/<file>?linerange=<from>:<to> tests.

sshpeer: try harder to snag stderr when stdout closes unexpectedly Resolves test failures on FreeBSD, but I'm not happy about the fix. A previous version of this also wrapped readline by putting the hack in the _call method on doublepipe. That was confusing for readers and wasn't necessary - just doing this on read() is sufficient to fix the bugs I'm observing. We can always come back and do readline later if needed.

test-worker: disable tests of forked workers on Windows The number of the "Traceback" lines differs on Windows because the main process does not raise SystemExit.

test-profile: allow negative time in JSON output (issue5542) I don't know why it can be a negative number, but that shouldn't be important here.

show: add basic labels to work template `hg show work` is much more usable if output is colored. This patch implements coloring via label() in a very hacky way. In a default Mercurial install, you'll see yellow node labels for all phases. Branches and bookmarks use the same formatting as the commit message. So this change doesn't help much in a default install. But if you have a custom colors defined for these things, output is much more readable. The implementation obviously needs some work. But for a minor change on a feature that isn't convered by BC, this seems like a clear win for the feature in 4.2.

show: rename "underway" to "work" Durham and I both like this better than "underway." We can add aliases and bikeshed on the name during the 4.3 cycle, as this whole extension is highly experimental.

histedit: make check for unresolved conflicts explicit (issue5545) Previously, we'd rely on the implicit check that `localrepo.commit` did. The problem is that that check only happened when the working copy was dirty. With a "clean" working copy but unresolved conflicts we'd get into a broken state. To fix that, do what rebase does and check for unresolved conflicts at the start of histedit --continue.

Added signature for changeset 616e788321cc

Added tag 4.2-rc for changeset 616e788321cc

freeze: merge default into stable for 4.2 code freeze

Merge stable with security patch.

Added signature for changeset 77eaf9539499

Added tag 4.1.3 for changeset 77eaf9539499

dispatch: protect against malicious 'hg serve --stdio' invocations (sec) Some shared-ssh installations assume that 'hg serve --stdio' is a safe command to run for minimally trusted users. Unfortunately, the messy implementation of argument parsing here meant that trying to access a repo named '--debugger' would give the user a pdb prompt, thereby sidestepping any hoped-for sandboxing. Serving repositories over HTTP(S) is unaffected. We're not currently hardening any subcommands other than 'serve'. If your service exposes other commands to users with arbitrary repository names, it is imperative that you defend against repository names of '--debugger' and anything starting with '--config'. The read-only mode of hg-ssh stopped working because it provided its hook configuration to "hg serve --stdio" via --config parameter. This is banned for security reasons now. This patch switches it to directly call ui.setconfig(). If your custom hosting infrastructure relies on passing --config to "hg serve --stdio", you'll need to find a different way to get that configuration into Mercurial, either by using ui.setconfig() as hg-ssh does in this patch, or by placing an hgrc file someplace where Mercurial will read it. mitrandir@fb.com provided some extra fixes for the dispatch code and for hg-ssh in places that I overlooked.

progress: retry ferr.flush() and .write() on EINTR (issue5532) See the inline comment how this could mitigate the issue. I couldn't reproduce the exact problem on my Linux machine, but there are at least two people who got EINTR in progress.py, and it seems file_write() of Python 2 is fundamentally broken [1]. Let's make something in on 4.2. [1]: https://hg.python.org/cpython/file/v2.7.13/Objects/fileobject.c#l1850

progress: extract stubs to restart ferr.flush() and .write() on EINTR

filemerge: optionally strip quotes from merge marker template (BC) For consistency with the other template options. Quotes are necessary if you want to preserve leading/trailing whitespace, which would be stripped by config parser.

commit: optionally strip quotes from commit template (BC) For consistency with the other template options.

graphlog: optionally strip quotes from graphnode template (BC) For consistency with the other template options.

dispatch: ignore further SIGPIPE while handling KeyboardInterrupt I got the following error by running "hg log" and quitting the pager immediately. Any output here may trigger another SIGPIPE, so only thing we can do is to swallow the exception and exit with an error status. Traceback (most recent call last): File "./hg", line 45, in <module> mercurial.dispatch.run() File "mercurial/dispatch.py", line 83, in run status = (dispatch(req) or 0) & 255 File "mercurial/dispatch.py", line 167, in dispatch req.ui.warn(_("interrupted!\n")) File "mercurial/ui.py", line 1224, in warn self.write_err(*msg, **opts) File "mercurial/ui.py", line 790, in write_err self._write_err(*msgs, **opts) File "mercurial/ui.py", line 798, in _write_err self.ferr.write(a) File "mercurial/ui.py", line 129, in _catchterm raise error.SignalInterrupt mercurial.error.SignalInterrupt Perhaps this wasn't visible before de5c9d0e02ea because the original stderr handle was restored very late.

worker: print traceback for uncaught exception unconditionally This is what a Python interpreter would do if there were no os._exit().

worker: propagate exit code to main process Follows up 86cd09bc13ba.

dispatch: print traceback in scmutil.callcatch() if --traceback specified Otherwise, traceback wouldn't be printed for a known exception occurred in worker processes.