tests: add optional setsockopt() output on Python 3 Tests still don't pass. But this gets us a little closer. Differential Revision: https://phab.mercurial-scm.org/D5699

wireprotov2peer: rewrite character traversal to use slices Otherwise on Python 3 we iterate over integers instead of a bytes instance and the comparison fails. Differential Revision: https://phab.mercurial-scm.org/D5698

wireprotov2server: don't attempt to cast command name I'm not sure why this was added. The command name should already be bytes. Differential Revision: https://phab.mercurial-scm.org/D5697

wireprotov2server: use pycompat.strkwargs when calling cachekeyfn Differential Revision: https://phab.mercurial-scm.org/D5696

diffstat: make --git work properly on renames (issue6025) `$ hg diff --stat --git` shows only the new filename on renames. I added the old filename also to the output to make it identical with the output of `$ git diff --stat`. Differential Revision: https://phab.mercurial-scm.org/D5628

cmdutil: drop obsolete comment about updating to '.' Obsolete since cf68e2649e0a (status: advertise --abort instead of 'update -C .' to abort a merge, 2018-08-09) and 5b04b6204931 (status: advertise --abort instead of 'update -C .' to abort graft, 2018-08-09). Differential Revision: https://phab.mercurial-scm.org/D5694

mq: slightly simplify check for patched working copy Differential Revision: https://phab.mercurial-scm.org/D5693

localrepo: slightly simplify check for removed parents in _rollback() Differential Revision: https://phab.mercurial-scm.org/D5692

transplant: use bailifchanged() instead of reimplementing it Differential Revision: https://phab.mercurial-scm.org/D5691

revpair: clarify check for empty revset Differential Revision: https://phab.mercurial-scm.org/D5688

revpair: simplify revpair by always relying on smartset.first/last I thinkt the code was written the way it was because it comes from 97b2f26dfc43 (revpair: smartset compatibility, 2014-03-20) and the first/last methods came only later, in 228b0aafb1ce (smartset: add first and last methods, 2014-10-06). Differential Revision: https://phab.mercurial-scm.org/D5687

grep: use set instead of dict with dummy value Differential Revision: https://phab.mercurial-scm.org/D5633

grep: don't look up copy info unless --follow is given If no --follow was given, then the "copy" variable will become None. In that case we would still look up the copy information from the filelog and then ignore it. Let's avoid even looking it up. Differential Revision: https://phab.mercurial-scm.org/D5620

hghave: let OSError with ENOENT through like any other Before this patch, if we get an OSError with ENOENT, we would not re-raise it and would instead run into an undefined variable ("p") soon thereafter. Differential Revision: https://phab.mercurial-scm.org/D5631

cleanup: delete lots of unused local variables These were found by IntelliJ. There are many more, but these seemed pretty safe. Differential Revision: https://phab.mercurial-scm.org/D5629

localrepo: use context manager for lock and transaction in commitctx() Differential Revision: https://phab.mercurial-scm.org/D5624

localrepo: use context manager for locks in commit() Differential Revision: https://phab.mercurial-scm.org/D5623

localrepo: use context manager for transaction in commit() A side-effect is that the transaction is released a little earlier, so the "commit message saved in ..." messages now comes after the "rollback completed" message. That seems slightly better to me. Differential Revision: https://phab.mercurial-scm.org/D5622

commit: print "commit message saved in" message also if bookmark update fails The bookmark update is done within the same transaction, so if fails, the commit will also be rolled back. I just happened to notice this while messing with this code; we have not had reports of crashing bookmark updates from our users. The same applies to the clearing of the merge state, although I don't think that's currently tied into the transaction. Differential Revision: https://phab.mercurial-scm.org/D5621

remotefilelog: remove pointless return statement from constructor Differential Revision: https://phab.mercurial-scm.org/D5630

revset: support ranges in #generations relation

tests: fix module-import warnings in test-update-atomic.t These were caught by test-check-module-imports.t on Python 3. This should also make python3-buildbot green again. Differential Revision: https://phab.mercurial-scm.org/D5674

py3: use dict.items() instead of dict.iteritems() in tests/test-lfs-server.t dict.iteritems() is not present on Python 3. Differential Revision: https://phab.mercurial-scm.org/D5673

py3: add b'' prefixes in tests/test-lfs-serve.t # skip-blame because just b'' prefixes Differential Revision: https://phab.mercurial-scm.org/D5672

py3: add b'' prefixes in doc/check-seclevel.py # skip-blame because just b'' prefixes Differential Revision: https://phab.mercurial-scm.org/D5670

tests: remove some over-globbing in test-dispatch.t Hilariously, this was exposed by Python 3 work. Differential Revision: https://phab.mercurial-scm.org/D5689

py3: port test-parseindex.t to Python 3 Differential Revision: https://phab.mercurial-scm.org/D5686

py3: almost fix test-trusted.py The test now only fails due to output format changes on the repr of mercurial.error.ParseError. I'm not really sure how to approach that for now. Differential Revision: https://phab.mercurial-scm.org/D5685

merge with stable

rust-cpython: raising error.WdirUnsupported The Graph implementation of hg-cpython returns the appropriate error upon encounter with the working directory special revision number, and this gives us in particular a code path to test from test-rust-ancestors.py In the current implementation, the exception is actually raised from the iterator instantiation; we are nonetheless consuming the iterator in the test with `list()` in order not to depend on implementation details.