Boris Feld <boris.feld@octobus.net> [Wed, 29 Nov 2017 23:20:52 -0500] rev 35243
test: fix common-pattern for pure variant
The $USUAL_COMPRESSIONS$ value that was taken was not compatible with the pure
variant systems as zlib seems to not be available in these case.
Differential Revision: https://phab.mercurial-scm.org/D1562
Matt Harbison <matt_harbison@yahoo.com> [Sat, 02 Dec 2017 20:03:28 -0500] rev 35242
tests: add a substitution for EADDRINUSE/WSAEADDRINUSE messages
I suspect some more of these are globbed out, so this is a bit of future
proofing.
Matt Harbison <matt_harbison@yahoo.com> [Sat, 02 Dec 2017 20:10:58 -0500] rev 35241
tests: add a substitution for ECONNRESET/WSAECONNRESET messages
Matt Harbison <matt_harbison@yahoo.com> [Sat, 02 Dec 2017 20:38:23 -0500] rev 35240
tests: add a substitution for ENOTDIR/ERROR_PATH_NOT_FOUND messages
Matt Harbison <matt_harbison@yahoo.com> [Sat, 02 Dec 2017 19:33:34 -0500] rev 35239
tests: add a substitution for ENOENT/ERROR_FILE_NOT_FOUND messages
Automatic replacement seems better than trying to figure out a check-code rule.
I didn't bother looking to see why the error message and file name is reversed
in the annotate and histedit tests, based on Windows or not.
I originally had this as a list of tuples, conditional on the platform. But
there are a couple of 'No such file or directory' messages emitted by Mercurial
itself, so unconditional is required for stability. There are also several
variants of what I assume is 'connection refused' and 'unknown host' in
test-clone.t and test-clonebundles.t for Docker, FreeBSD jails, etc. Yes, these
are handled by (re) tags, but maybe it would be better to capture those strings
in order to avoid whack-a-mole in future tests. All of this points to using a
dictionary containing one or more strings-to-be-replaced values.
Gregory Szorc <gregory.szorc@gmail.com> [Sun, 03 Dec 2017 20:55:35 -0800] rev 35238
setup: only write some autogenerated files if they change
Without this change, setup.py always writes some files on every
invocation. This prevents some builds from being a no-op when they
should. And, since times can sneak into generated .pyc files,
this prevents file content from being deterministic between
builds.
As part of the refactor, we treat file content as bytes.
The only potential regression from this would be if some tool
is looking at mtimes of the changed files to determine if
further action should be taken. But I don't think anything
critically important is keyed off the mtimes of these specific files.
Differential Revision: https://phab.mercurial-scm.org/D1580
Augie Fackler <augie@google.com> [Mon, 04 Dec 2017 15:30:30 -0500] rev 35237
python3: whitelist many more passing tests
Differential Revision: https://phab.mercurial-scm.org/D1584