Fri, 28 Sep 2018 10:03:32 -0700 dagop: extract descendants() from revlog module
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 28 Sep 2018 10:03:32 -0700] rev 39999
dagop: extract descendants() from revlog module This method needs to be implemented in other storage backends and is generic if we parameterize the functions for retrieving revision numbers and parent revision numbers. Let's extract it to dagop so backends don't need to reinvent the wheel. I'm not too worried about performance overhead of the extra function call because I don't expect anyone to be calling descendants() in a tight loop. Differential Revision: https://phab.mercurial-scm.org/D4794
Fri, 28 Sep 2018 09:33:05 -0700 filelog: remove checkhash() (API)
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 28 Sep 2018 09:33:05 -0700] rev 39998
filelog: remove checkhash() (API) It is unused. While a caller may want to ask the store to validate the hash of some provided text, since there are no in-core consumers of this method, let's drop it for now. Differential Revision: https://phab.mercurial-scm.org/D4793
Fri, 28 Sep 2018 09:28:38 -0700 filelog: remove revdiff() (API)
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 28 Sep 2018 09:28:38 -0700] rev 39997
filelog: remove revdiff() (API) This proxy method is no longer used. While it might be useful to query a storage backend for the delta between any 2 revisions because the store could have a delta cached and could compute it more efficiently than the caller calling revision() twice in order to compute a delta, since nothing in core is using this API now, I feel comfortable nuking it. Differential Revision: https://phab.mercurial-scm.org/D4792
Fri, 28 Sep 2018 09:46:50 -0700 localrepo: define storage backend in creation options (API)
Gregory Szorc <gregory.szorc@gmail.com> [Fri, 28 Sep 2018 09:46:50 -0700] rev 39996
localrepo: define storage backend in creation options (API) We add an experimental config option to define the storage backend for new repositories. By default, it uses "revlogv1," which maps to the current and only modern supported repository format. We add a "backend" creation option to control which backend to use. It defaults to using the value from the config option. newreporequirements() will now barf if it sees a "backend" value that isn't "revlogv1." This forces extensions to monkeypatch the function to handle requirements derivation for custom backends. In order for this to "just work," we factored out obtaining the default creation options into its own function and made callers of newreporequirements() responsible for passing in valid data. Without this, direct callers of newreporequirements() wouldn't get the proper results. Differential Revision: https://phab.mercurial-scm.org/D4791
Thu, 27 Sep 2018 09:23:17 -0700 wireprotov2: derive "required" from presence of default value
Gregory Szorc <gregory.szorc@gmail.com> [Thu, 27 Sep 2018 09:23:17 -0700] rev 39995
wireprotov2: derive "required" from presence of default value If we define a default value for all optional arguments, we no longer need to explicitly declare whether the argument is required. Instead, we can derive it from the presence of a default value callable. Differential Revision: https://phab.mercurial-scm.org/D4790
Mon, 01 Oct 2018 09:05:40 -0700 localrepo: capture repo interface factory functions as lambas
Gregory Szorc <gregory.szorc@gmail.com> [Mon, 01 Oct 2018 09:05:40 -0700] rev 39994
localrepo: capture repo interface factory functions as lambas Previously, the list would hold a reference to the original function and attempting to wrap the module function via extension wouldn't result in the wrapped function being called, completely undermining the expected behavior. Differential Revision: https://phab.mercurial-scm.org/D4821
Mon, 14 May 2018 00:43:07 +0200 extensions: new closehead module for closing arbitrary heads
Joerg Sonnenberger <joerg@bec.de> [Mon, 14 May 2018 00:43:07 +0200] rev 39993
extensions: new closehead module for closing arbitrary heads ``hg close-head`` allows closing arbitrary heads. It is equivalent to checking out the given revisions and commit an empty change with ``hg commit --close-branch``. Differential Revision: https://phab.mercurial-scm.org/D3557
Tue, 02 Oct 2018 13:12:56 -0700 cext: use modern buffer protocol in mpatch_flist()
Gregory Szorc <gregory.szorc@gmail.com> [Tue, 02 Oct 2018 13:12:56 -0700] rev 39992
cext: use modern buffer protocol in mpatch_flist() Differential Revision: https://phab.mercurial-scm.org/D4841
Tue, 02 Oct 2018 13:13:03 -0700 cext: use modern buffer protocol in patches()
Gregory Szorc <gregory.szorc@gmail.com> [Tue, 02 Oct 2018 13:13:03 -0700] rev 39991
cext: use modern buffer protocol in patches() PyObject_AsCharBuffer() is part of the "Old Buffer Protocol," which has been deprecated for years. Let's port away from it. PyBuffer_GetBuffer() must be paired with PyBuffer_Release(), hence the added "goto cleanup" in a failure case. We don't bump the extension version because the API has not changed. Differential Revision: https://phab.mercurial-scm.org/D4840
Mon, 01 Oct 2018 14:44:27 -0400 identify: when using -T, avoid unnecessary remote bookmarks query
Valentin Gatien-Baron <vgatien-baron@janestreet.com> [Mon, 01 Oct 2018 14:44:27 -0400] rev 39990
identify: when using -T, avoid unnecessary remote bookmarks query Differential Revision: https://phab.mercurial-scm.org/D4839
Mon, 01 Oct 2018 09:58:42 -0400 identify: only query remote bookmarks if needed
Valentin Gatien-Baron <vgatien-baron@janestreet.com> [Mon, 01 Oct 2018 09:58:42 -0400] rev 39989
identify: only query remote bookmarks if needed Instead of all the time when operating on a remote repo. This perf regression was introduced in 15a79ac823e8, in 4.3. This datahint method returns nothing for -Tjson, -Tpickle, -Tdebug --config ui.formatdebug=true and --config ui.formatjson, so the bookmarks won't show up. I don't know what these formatters are for. plainformatter and templateformatter work properly, and the few other uses of datahint should have the same kind of problem. There is further weirdness where "--template '{node}'" is not enough to avoid querying the bookmarks, you also need to pass --id or -q. Differential Revision: https://phab.mercurial-scm.org/D4819
Wed, 03 Oct 2018 13:59:19 +0300 py3: whitelist another passing tests caught by buildbot
Pulkit Goyal <pulkit@yandex-team.ru> [Wed, 03 Oct 2018 13:59:19 +0300] rev 39988
py3: whitelist another passing tests caught by buildbot I this indygreg recent patches made this test pass on Python 3. So thanks to him! Differential Revision: https://phab.mercurial-scm.org/D4848
Wed, 03 Oct 2018 13:55:51 +0300 manifest: remove an unused variable caught by pyflakes
Pulkit Goyal <pulkit@yandex-team.ru> [Wed, 03 Oct 2018 13:55:51 +0300] rev 39987
manifest: remove an unused variable caught by pyflakes Uses of todel were removed in D4843 (19103e68a698). This patch will make buildbots green again. Differential Revision: https://phab.mercurial-scm.org/D4847
Tue, 02 Oct 2018 22:40:01 -0400 setup: ignore message about disabling 3rd party extensions because of version
Matt Harbison <matt_harbison@yahoo.com> [Tue, 02 Oct 2018 22:40:01 -0400] rev 39986
setup: ignore message about disabling 3rd party extensions because of version I started getting into a bind recently when switching between py2 and py3 because switching requires a `make clean`, which kills __version__.py. But then when running `make local`, it picks up the local hg.exe (MSYS seems to prefix $PATH with '.'), which doesn't know its version. That causes it to emit a warning about needing at least 4.3 to load evolve, which caused setup.py to fail saying there is no working hg executable to figure out the version. If we can ignore general extension import failures, we should be able to ignore this too.
Tue, 02 Oct 2018 09:11:18 -0700 narrow: avoid overwriting a variable
Martin von Zweigbergk <martinvonz@google.com> [Tue, 02 Oct 2018 09:11:18 -0700] rev 39985
narrow: avoid overwriting a variable I don't like to overwrite variables, especially not with a different type (because it makes it harder to explain what the variable is for, and, therefore, to name it well). Differential Revision: https://phab.mercurial-scm.org/D4846
Thu, 27 Sep 2018 20:16:48 -0700 treemanifests: remove _loadalllazy in _diff()
spectral <spectral@google.com> [Thu, 27 Sep 2018 20:16:48 -0700] rev 39984
treemanifests: remove _loadalllazy in _diff() The benchmarks below use a similar setup as in ee7ee0c516ca and my other recent commits. Yes, in some cases this runs in literally 5% of the time it previously took. before = a6f8ab53 diff --git: repo | N | T | before (mean +- stdev) | after (mean +- stdev) | % of before ------+---+---+------------------------+-----------------------+------------ m-u | | | 1.301 s +- 0.004 s | 1.309 s +- 0.012 s | 100.6% m-u | | x | 1.303 s +- 0.009 s | 1.302 s +- 0.006 s | 99.9% m-u | x | | 1.308 s +- 0.006 s | 1.309 s +- 0.007 s | 100.1% m-u | x | x | 85.7 ms +- 0.6 ms | 86.0 ms +- 0.3 ms | 100.4% l-d-r | | | 197.5 ms +- 0.7 ms | 197.8 ms +- 2.2 ms | 100.2% l-d-r | | x | 199.4 ms +- 0.6 ms | 199.3 ms +- 0.9 ms | 99.9% l-d-r | x | | 86.1 ms +- 0.5 ms | 85.8 ms +- 0.9 ms | 99.7% l-d-r | x | x | 64.4 ms +- 0.4 ms | 64.4 ms +- 0.3 ms | 100.0% diff -c . --git: repo | N | T | before (mean +- stdev) | after (mean +- stdev) | % of before ------+---+---+------------------------+-----------------------+------------ m-u | | | 236.7 ms +- 1.1 ms | 236.5 ms +- 1.3 ms | 99.9% m-u | | x | 158.7 ms +- 1.0 ms | 128.0 ms +- 1.0 ms | 80.7% <-- m-u | x | | 239.7 ms +- 1.8 ms | 238.1 ms +- 1.5 ms | 99.3% m-u | x | x | 132.4 ms +- 0.9 ms | 132.3 ms +- 0.6 ms | 99.9% l-d-r | | | 81.8 ms +- 0.4 ms | 81.8 ms +- 0.3 ms | 100.0% l-d-r | | x | 3.894 s +- 0.017 s | 193.6 ms +- 0.4 ms | 5.0% <-- l-d-r | x | | 106.9 ms +- 0.4 ms | 106.6 ms +- 0.3 ms | 99.7% l-d-r | x | x | 182.7 ms +- 0.8 ms | 183.3 ms +- 0.9 ms | 100.3% rebase -r . --keep -d .^^: repo | N | T | before (mean +- stdev) | after (mean +- stdev) | % of before ------+---+---+------------------------+-----------------------+------------ m-u | | | 5.615 s +- 0.109 s | 5.562 s +- 0.015 s | 99.1% m-u | | x | 5.701 s +- 0.027 s | 5.715 s +- 0.023 s | 100.2% m-u | x | | 5.572 s +- 0.128 s | 5.613 s +- 0.182 s | 100.7% m-u | x | x | 633.3 ms +- 28.7 ms | 636.2 ms +- 13.8 ms | 100.5% l-d-r | | | 666.4 ms +- 17.0 ms | 658.5 ms +- 9.3 ms | 98.8% l-d-r | | x | 6.520 s +- 0.070 s | 6.505 s +- 0.026 s | 99.8% l-d-r | x | | 279.0 ms +- 13.0 ms | 276.5 ms +- 4.7 ms | 99.1% l-d-r | x | x | 1.636 s +- 0.058 s | 1.657 s +- 0.014 s | 101.3% status --change . --copies: repo | N | T | before (mean +- stdev) | after (mean +- stdev) | % of before ------+---+---+------------------------+-----------------------+------------ m-u | | | 218.6 ms +- 1.4 ms | 217.9 ms +- 1.5 ms | 99.7% m-u | | x | 138.5 ms +- 0.4 ms | 108.4 ms +- 0.2 ms | 78.3% <-- m-u | x | | 220.1 ms +- 1.3 ms | 219.7 ms +- 1.5 ms | 99.8% m-u | x | x | 113.2 ms +- 0.4 ms | 112.4 ms +- 0.8 ms | 99.3% l-d-r | | | 80.2 ms +- 0.3 ms | 80.6 ms +- 0.6 ms | 100.5% l-d-r | | x | 3.899 s +- 0.020 s | 194.8 ms +- 4.0 ms | 5.0% <-- l-d-r | x | | 83.4 ms +- 0.8 ms | 83.2 ms +- 0.2 ms | 99.8% l-d-r | x | x | 732.2 ms +- 4.3 ms | 194.9 ms +- 1.0 ms | 26.6% <-- status --copies: repo | N | T | before (mean +- stdev) | after (mean +- stdev) | % of before ------+---+---+------------------------+-----------------------+------------ m-u | | | 1.917 s +- 0.005 s | 1.914 s +- 0.004 s | 99.8% m-u | | x | 1.909 s +- 0.012 s | 1.934 s +- 0.004 s | 101.3% m-u | x | | 1.915 s +- 0.005 s | 1.904 s +- 0.004 s | 99.4% m-u | x | x | 94.8 ms +- 0.3 ms | 94.7 ms +- 0.2 ms | 99.9% l-d-r | | | 593.9 ms +- 1.2 ms | 594.6 ms +- 9.4 ms | 100.1% l-d-r | | x | 595.2 ms +- 3.8 ms | 597.2 ms +- 2.6 ms | 100.3% l-d-r | x | | 182.5 ms +- 1.6 ms | 182.1 ms +- 0.6 ms | 99.8% l-d-r | x | x | 149.6 ms +- 0.9 ms | 149.1 ms +- 0.8 ms | 99.7% update $rev^; ~/src/hg/hg{hg}/hg update $rev: repo | N | T | before (mean +- stdev) | after (mean +- stdev) | % of before ------+---+---+------------------------+-----------------------+------------ m-u | | | 3.121 s +- 0.007 s | 3.129 s +- 0.012 s | 100.3% m-u | | x | 2.972 s +- 0.011 s | 2.981 s +- 0.012 s | 100.3% m-u | x | | 3.144 s +- 0.014 s | 3.141 s +- 0.011 s | 99.9% m-u | x | x | 312.2 ms +- 2.4 ms | 312.3 ms +- 2.1 ms | 100.0% l-d-r | | | 444.4 ms +- 4.3 ms | 446.9 ms +- 5.3 ms | 100.6% l-d-r | | x | 9.159 s +- 0.069 s | 9.182 s +- 0.040 s | 100.3% l-d-r | x | | 254.6 ms +- 1.6 ms | 255.2 ms +- 1.6 ms | 100.2% l-d-r | x | x | 1.525 s +- 0.007 s | 1.577 s +- 0.007 s | 103.4% <--? Differential Revision: https://phab.mercurial-scm.org/D4845
Tue, 02 Oct 2018 13:41:00 -0700 treemanifests: skip extraneous check for item before calling _loadlazy
spectral <spectral@google.com> [Tue, 02 Oct 2018 13:41:00 -0700] rev 39983
treemanifests: skip extraneous check for item before calling _loadlazy Differential Revision: https://phab.mercurial-scm.org/D4844
Tue, 02 Oct 2018 13:38:26 -0700 treemanifests: make _loadchildrensetlazy just call _loadlazy
spectral <spectral@google.com> [Tue, 02 Oct 2018 13:38:26 -0700] rev 39982
treemanifests: make _loadchildrensetlazy just call _loadlazy This was suggested on the original review for _loadchildrensetlazy (D4370), it's taken me a while to get to it though. Differential Revision: https://phab.mercurial-scm.org/D4843
Tue, 02 Oct 2018 13:37:12 -0700 treemanifests: make _loadlazy tolerate item not on _lazydirs
spectral <spectral@google.com> [Tue, 02 Oct 2018 13:37:12 -0700] rev 39981
treemanifests: make _loadlazy tolerate item not on _lazydirs I'd like to clean up a few cases where we check for an item in _lazydirs before calling _loadlazy - this will remove an extraneous dict lookup and make it slightly more versatile. Differential Revision: https://phab.mercurial-scm.org/D4842
Wed, 08 Aug 2018 23:17:16 -0700 debugcommands: add a debugindexstats command
Martin von Zweigbergk <martinvonz@google.com> [Wed, 08 Aug 2018 23:17:16 -0700] rev 39980
debugcommands: add a debugindexstats command Someone went through the trouble of recording these stats, so let's make them accessible. Differential Revision: https://phab.mercurial-scm.org/D4832
Mon, 24 Sep 2018 19:23:50 +0900 rust-chg: add main program
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 19:23:50 +0900] rev 39979
rust-chg: add main program It can at least connect to a *running* command server, and execute Mercurial command. No signal handling nor daemon management is implemented yet.
Mon, 24 Sep 2018 19:06:30 +0900 rust-chg: add interface to chdir the server
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 19:06:30 +0900] rev 39978
rust-chg: add interface to chdir the server
Mon, 24 Sep 2018 18:57:54 +0900 rust-chg: add Client extensions to run cHg-specific requests
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 18:57:54 +0900] rev 39977
rust-chg: add Client extensions to run cHg-specific requests This just provides a nicer way to issue command-server requests.
Mon, 24 Sep 2018 18:33:46 +0900 rust-chg: port basic socket path handling from cHg of C
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 18:33:46 +0900] rev 39976
rust-chg: port basic socket path handling from cHg of C This is basically modeled after setcmdserveropts() of chg.c.
Mon, 24 Sep 2018 18:21:10 +0900 rust-chg: add state machine to handle "runcommand" request with cHg extension
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 18:21:10 +0900] rev 39975
rust-chg: add state machine to handle "runcommand" request with cHg extension This is modeled after tokio-hglib's RunCommand state to support the "S" channel message.
Mon, 24 Sep 2018 18:18:35 +0900 rust-chg: add callback to handle pager and shell command requests
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 18:18:35 +0900] rev 39974
rust-chg: add callback to handle pager and shell command requests This could be inlined into the ChgRunCommand state to be introduced by the next patch, but it seemed good to separate any user interactions from the IPC code.
Sat, 29 Sep 2018 21:59:07 +0900 rust-chg: add low-level function to set pager fd blocking
Yuya Nishihara <yuya@tcha.org> [Sat, 29 Sep 2018 21:59:07 +0900] rev 39973
rust-chg: add low-level function to set pager fd blocking This is necessary because the server expects stdout/stderr to be blocking, whereas we'll use async library to spawn pager, which makes pipes unblocking.
Mon, 24 Sep 2018 16:59:12 +0900 rust-chg: add future that handles "attachio" request
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 16:59:12 +0900] rev 39972
rust-chg: add future that handles "attachio" request This is the sequence to send client-side stdio and pager stdin to the server.
Mon, 24 Sep 2018 16:33:24 +0900 rust-chg: add parser for request messages sent to "S" channel
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 16:33:24 +0900] rev 39971
rust-chg: add parser for request messages sent to "S" channel The data structure is documented at chgserver.py:channeledsystem().
Mon, 24 Sep 2018 16:22:03 +0900 rust-chg: add wrapper around C function
Yuya Nishihara <yuya@tcha.org> [Mon, 24 Sep 2018 16:22:03 +0900] rev 39970
rust-chg: add wrapper around C function
(0) -30000 -10000 -3000 -1000 -300 -100 -50 -30 +30 +50 +100 +300 +1000 +3000 +10000 tip