Julien Cristau <jcristau@mozilla.com> [Tue, 24 Aug 2021 12:44:34 +0200] rev 47854
revlog: fix type confusion with sidedata_comp_len (issue6580)
The format string uses "i" (int) for sidedata_comp_len, so we shouldn't
be passing a pointer to Py_ssize_t to PyArg_ParseTuple. On 64-bit
big-endian, this would result in python only writing to the upper 32
bits, and things go downhill from there.
Differential Revision: https://phab.mercurial-scm.org/D11334
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 19 Aug 2021 17:47:27 +0200] rev 47853
resources: stop important a non existent FileNotFoundError
The error was hidden by the wide try/except
Differential Revision: https://phab.mercurial-scm.org/D11311
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 19 Aug 2021 17:46:46 +0200] rev 47852
template: FileNotFoundError is actually a built in exception
However it is python3 only.
Differential Revision: https://phab.mercurial-scm.org/D11310
Pierre-Yves David <pierre-yves.david@octobus.net> [Thu, 19 Aug 2021 18:04:14 +0200] rev 47851
check-code: drop the camelcase checks
Fro about 2 years we have been using CamelCase is class names. That rules gets
in the way of assigning class or exception in compatibility layers.
I think it is safe to drop it now that we started using CamelCase for some case.
My motivation for this is the need to assign `FileNotFoundError` in `pycompat`.
Differential Revision: https://phab.mercurial-scm.org/D11309
Pulkit Goyal <7895pulkit@gmail.com> [Fri, 20 Aug 2021 00:23:49 +0530] rev 47850
Added signature for changeset 53221078e0de
Pulkit Goyal <7895pulkit@gmail.com> [Fri, 20 Aug 2021 00:23:40 +0530] rev 47849
Added tag 5.9 for changeset 53221078e0de
Pierre-Yves David <pierre-yves.david@octobus.net> [Mon, 16 Aug 2021 18:39:58 +0200] rev 47848
pyoxidizer: always run the associated CI jobs
Now that all tests are passing, we can enable this by default.
Differential Revision: https://phab.mercurial-scm.org/D11306
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 21:28:36 +0200] rev 47847
pyoxidizer: skip for ZeroConf related test for now
These test are failing for obscure reason. They are fairly minor and Since they
are the only thing between us and enabling pyoxidizer testing in the CI by
default, we just shallow the bad output for now.
Differential Revision: https://phab.mercurial-scm.org/D11305
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 21:05:46 +0200] rev 47846
pyoxidizer: disable the test for disabled & broken extensions help
The test is a bit fragile and the feature is significantly broken for 3rd party
extension already. So we disable that section with a comment about why and what
is needed.
Differential Revision: https://phab.mercurial-scm.org/D11304
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 21:37:55 +0200] rev 47845
pyoxidizer: re-install PYTHONPATH behavior
Hooks and extensions can rely on PYTHONPATH value. Yet the pyoxidized binary
ignore the variable. So we manually reinstall the behavior to fix various tests
and restore the expected behavior.
This positively affects:
- test-extension.t
- test-hook.t
Differential Revision: https://phab.mercurial-scm.org/D11302
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 16:48:32 +0200] rev 47844
windows: add pytest-vcr to the dependencies
This will lead to pyoxidizer including it in its binary.
This fix test-phabricator.t
Differential Revision: https://phab.mercurial-scm.org/D11301
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 16:02:46 +0200] rev 47843
pyoxidizer: wrap the pyoxidizer script on multiple line
This is clearer and will help with future edition of the script.
Differential Revision: https://phab.mercurial-scm.org/D11300
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 15:05:25 +0200] rev 47842
pyoxidized: silence the fuzzywuzzy warning about python-Levenshtein
I don't think we need the fast implementation for the test so lets ignore it for
now.
Differential Revision: https://phab.mercurial-scm.org/D11299
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 13:16:04 +0200] rev 47841
pyoxidized: install fuzzywuzzy too
This is necessary for some of the release note tooling.
Differential Revision: https://phab.mercurial-scm.org/D11298
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 14:57:34 +0200] rev 47840
pyoxidized: adapt output of test-install.t
The python lib and template directories are in different location that requires
a different patterns.
Differential Revision: https://phab.mercurial-scm.org/D11297
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 13:02:08 +0200] rev 47839
pyoxidized: disable part of `test-install.t` related to pip and virtualenv
This part are testing manual installation of Mercurial, this is not relevant
for the pyoxidizer variant.
Differential Revision: https://phab.mercurial-scm.org/D11296
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 12:53:40 +0200] rev 47838
pyoxidized: adapt output of test-bad-extension.t
The file path in the traceback become module name.
Differential Revision: https://phab.mercurial-scm.org/D11295
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 12:49:55 +0200] rev 47837
pyoxidized: adapt output of test-flagprocessor.t
The file path in the traceback become module name.
Differential Revision: https://phab.mercurial-scm.org/D11294
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 12:42:16 +0200] rev 47836
pyoxidized: adapt output of `test-phases.t`
We are getting one less line. I am not sure why, but it seems fairly minor.
Differential Revision: https://phab.mercurial-scm.org/D11293
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 01:27:45 +0200] rev 47835
pyoxidized: add a copy of the `doc` directory in the right location
This get the hacky setup closer to how an actual install is setup.
Differential Revision: https://phab.mercurial-scm.org/D11292
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 01:26:25 +0200] rev 47834
pyoxidized: add a copy of the `contrib` directory in the right location
This get the hacky setup closer to how an actual install is setup.
Differential Revision: https://phab.mercurial-scm.org/D11291
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 01:18:34 +0200] rev 47833
pyoxidized: add a copy of the `defaultrc` file in the right location
This get the hacky setup closer to how an actual install is setup.
Differential Revision: https://phab.mercurial-scm.org/D11290
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 02:08:03 +0200] rev 47832
pyoxidized: add a copy of the `helptext` directory in the right location
This get the hacky setup closer to how an actual install is setup.
Differential Revision: https://phab.mercurial-scm.org/D11289
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 01:41:02 +0200] rev 47831
template: handle missing resource in `_readmapfile`
This fix the remaining issue in test-template-map.t.
Differential Revision: https://phab.mercurial-scm.org/D11288
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 01:10:40 +0200] rev 47830
pyoxidized: add a copy of the `templates` directory in the right location
This fix about 10 extra test bring the number of currently failing test to about 12.
Differential Revision: https://phab.mercurial-scm.org/D11287
Pierre-Yves David <pierre-yves.david@octobus.net> [Wed, 18 Aug 2021 01:07:38 +0200] rev 47829
pyoxidized: add a dedicated target in the tests
We will need to setup more items to make the "quick" version work on stable,
having a dedicated operation will allow to contains change to that operation.
Differential Revision: https://phab.mercurial-scm.org/D11286
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 17 Aug 2021 20:48:52 +0200] rev 47828
pyoxidized: adapt test-devel-warnings.t
Differential Revision: https://phab.mercurial-scm.org/D11284
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 17 Aug 2021 21:57:59 +0200] rev 47827
test-extension: adapt output to pyoxidizer
The traceback use module name instead of filename.
The rest of the test is still on fire for various other reasons. However that
specific output mismatch is now fixed.
Differential Revision: https://phab.mercurial-scm.org/D11285
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 17 Aug 2021 20:48:34 +0200] rev 47826
pyoxidized: add a `pyoxidizer` hghave keyword for line matching
Some output line can be affected by pyoxidizer, for example the source file path
are replaced by the module name. We introduce a new condition keyword to cope
with this.
Differential Revision: https://phab.mercurial-scm.org/D11283
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 17 Aug 2021 14:38:25 +0200] rev 47825
templater: swap `\` with `/` in more location
This is similar to the previous changeset, but apply that logic to
_readmapfile.
Differential Revision: https://phab.mercurial-scm.org/D11282
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 17 Aug 2021 13:22:06 +0200] rev 47824
templater: swap `\` with `/` to allow the resource logic to kicks in
Without this change our wrapper to the resource logic will fails to details the
nesting and try to pass `directory/file` entry to the resource module, leading
to a crash.
Ideally, we should track down all caller passing `\` instead of `/`, but that
is far more intrusive and we would like to have a working pyoxidized version
working for 5.9 It will be easy to add a warning to the above code to track
down "bad" caller during the 6.0 cycle.
This reduce the number of failing test with pyoxidizer from 52 to 23
Differential Revision: https://phab.mercurial-scm.org/D11281
Raphaël Gomès <rgomes@octobus.net> [Thu, 05 Aug 2021 18:38:42 +0200] rev 47823
heptapod-ci: extend timeout for pytype checking
This test routinely times out for loaded or older machines.
Differential Revision: https://phab.mercurial-scm.org/D11258
Matt Harbison <matt_harbison@yahoo.com> [Mon, 09 Aug 2021 00:49:50 -0400] rev 47822
ci: run --pyoxidized tests on Windows
They still have numerous failure, but at least we can start fixing them now.
Differential Revision: https://phab.mercurial-scm.org/D11278
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 10 Aug 2021 12:56:32 +0200] rev 47821
run-tests: introduce a --pyoxidized option
This options make it possible to use the pyoxidizer version to run the tests.
This is a first basic version that is windows only.
The test needs a working python, with Mercurial installed. However the
pyoxidizer product is "self contains" without a "usable" Python. There have been
discussion to have a fully functional `hg admin::python` command providing a
fully functional python interpreter, but nothing is of the sort is ready yet. In
In the meantime we use an hybrid approach, similar to what we do for testing
`rhg`. We install a full "normal" Mercurial, but also the pyxodizer product as
the official `hg binary`. That way, we use the pyoxidizer version or everything,
but test that needs to run python have it available, with a fully functional
Mercurial package.
This first version is pretty basic (Windows only, no --local, not
--with-pyoxidized), but it runs, various bug that we will have to fix.
Differential Revision: https://phab.mercurial-scm.org/D11277
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 10 Aug 2021 11:45:43 +0200] rev 47820
ci: add a "all" template to easily control "when" test run
When debugging some job (usually windows one) it is handy to have a way to pass
all other jobs to manual execution.
We add a "root" template to control just that.
Differential Revision: https://phab.mercurial-scm.org/D11276
Pulkit Goyal <7895pulkit@gmail.com> [Tue, 10 Aug 2021 19:19:53 +0530] rev 47819
Added signature for changeset 2813d406b036
Pulkit Goyal <7895pulkit@gmail.com> [Tue, 10 Aug 2021 19:19:47 +0530] rev 47818
Added tag 5.9rc1 for changeset 2813d406b036
Pierre-Yves David <pierre-yves.david@octobus.net> [Sat, 07 Aug 2021 16:51:47 +0200] rev 47817
issue6528: add a config option to control the fixing on the fly
This will allow people who know to be safe to avoid any performance overhead
(and other potential issue).
Differential Revision: https://phab.mercurial-scm.org/D11271
Pierre-Yves David <pierre-yves.david@octobus.net> [Sat, 07 Aug 2021 14:12:28 +0200] rev 47816
issue6528: also filter delta on the fly when applying a changegroup
This ensure that corrupted clone does not spread corruption to "fixed" version.
This might come at a performance cost, we will had a config option to control
this behavior in the next changesets.
Differential Revision: https://phab.mercurial-scm.org/D11270
Pierre-Yves David <pierre-yves.david@octobus.net> [Sat, 07 Aug 2021 14:13:53 +0200] rev 47815
filelog: open the writing context a bit earlier in `addgroup`
This is a small change made ahead of the next patch for clarification.
Differential Revision: https://phab.mercurial-scm.org/D11269
Pierre-Yves David <pierre-yves.david@octobus.net> [Sat, 07 Aug 2021 12:39:01 +0200] rev 47814
issue6528: implement _is_revision_affected_fast using callback
The delta comming from a bundle/stream does not exists in the revlog yet, so we
will need other way to retrieve the same information.
To prepare for this we split the function to use callbacks in the core logic.
Differential Revision: https://phab.mercurial-scm.org/D11268
Pierre-Yves David <pierre-yves.david@octobus.net> [Sat, 07 Aug 2021 12:38:48 +0200] rev 47813
issue6528: implement _is_revision_affected using callback
The delta comming from a bundle/stream does not exists in the revlog yet, so we
will need other way to retrieve the same information.
To prepare for this we split the function to use callbacks in the core logic.
Differential Revision: https://phab.mercurial-scm.org/D11267
Raphaël Gomès <rgomes@octobus.net> [Fri, 06 Aug 2021 12:10:36 +0200] rev 47812
debugcommands: add a `--paranoid` option to `debug-repair-issue-6528`
See justification inline.
Differential Revision: https://phab.mercurial-scm.org/D11263
Raphaël Gomès <rgomes@octobus.net> [Thu, 05 Aug 2021 17:00:03 +0200] rev 47811
repair: improve performance of detection of revisions affected by issue6528
Explanations inside the patch. I've tested this on Mozilla-Central and it's
5 times faster than the naive approach on my laptop.
Differential Revision: https://phab.mercurial-scm.org/D11262
Raphaël Gomès <rgomes@octobus.net> [Tue, 27 Jul 2021 21:45:27 +0200] rev 47810
debugcommands: introduce a debug command to repair repos affected by issue6528
This command is quite basic and slow, it will loop over the entirety of the
filelogs in the repository and check each revision for corruption, then fixes
the affected filelogs. It takes under 25 minutes for Mozilla-Central on my
not-top-of-the-line laptop, using the `--to-report` and `--from-report` options
will make this pretty tolerable to use, I think.
This change also introduces a test for the fix.
Differential Revision: https://phab.mercurial-scm.org/D11239
Matt Harbison <matt_harbison@yahoo.com> [Mon, 09 Aug 2021 19:49:57 -0400] rev 47809
contrib: switch the Windows bootstrap environment to py3.9
Use the built in `venv` module instead of `virtualenv` for simplicity, and
upgrade to a modern Mercurial that supports py3.
One issue here is that `venv` doesn't copy `python3{,Y}.dll` into the `Scripts`
subdirectory, so running the `hg.exe` that gets installed immediately fails on a
clean system because Python isn't in `PATH`. There is code in `python.exe` to
detect when it is in a venv and add the original python install to the DLL
lookup path, which we don't do in `hg.exe` yet. The simple workaround for now is
to run the `hg` script with `python.exe`. Typically `PYTHONLEGACYWINDOWSSTDIO`
must be set in the environment on Windows, but the clone process works without
it.
Differential Revision: https://phab.mercurial-scm.org/D11275
Matt Harbison <matt_harbison@yahoo.com> [Mon, 09 Aug 2021 19:36:26 -0400] rev 47808
contrib: comment out the 64-bit py38 dependency installation on Windows
Not sure what is going on here, but it appears to not install py3.8 x64 in the
usual `C:\hgdev` directory. The x32 installer works fine. I'm assuming this is
a bug in this version of the installer, but didn't look into it too much.
Differential Revision: https://phab.mercurial-scm.org/D11274
Matt Harbison <matt_harbison@yahoo.com> [Mon, 09 Aug 2021 19:32:44 -0400] rev 47807
contrib: drop python2.7 from the Windows dependency install script
While we aren't quite ready to drop py27 yet, the MS compiler is no longer
available from MS (there is a copy on github if it's really needed), and that
causes the script to fail.
Differential Revision: https://phab.mercurial-scm.org/D11273