Yuya Nishihara <yuya@tcha.org> [Wed, 15 Apr 2020 23:11:55 +0900] rev 44712
templatekw: cache mergestate even if merge is not ongoing
While playing with eBPF, I noticed .hg/merge/state{,2} files were tried
to open() for each revision. That's not healthy. Let's cache the "inactive"
state as well.
Yuya Nishihara <yuya@tcha.org> [Wed, 15 Apr 2020 19:24:21 +0900] rev 44711
revset: remove explicit sort() from unstable()
The instability should be dealt with by baseset.__init__(), and the returned
smartset is ordered by 'subset &' anyway.
Anton Shestakov <av6@dwimlabs.net> [Fri, 10 Apr 2020 22:23:44 +0800] rev 44710
revset: implement a simple 'foo#generations' expression
The result of this expression is equivalent to 'foo#g[:]': all reachable
ancestors and descendants of a revset foo.
While not very useful functionality on its own, its implementation can be
tested and revset.relations dict can be easily used by extensions to add other
non-subscript relations.
Differential Revision: https://phab.mercurial-scm.org/D8394
Anton Shestakov <av6@dwimlabs.net> [Fri, 10 Apr 2020 22:22:09 +0800] rev 44709
revset: rename generationsrel() to generationssubrel()
This highlights that this is an implementation of 'x#y[z]' relation subscript
operator as opposed to 'x#y' relation operator.
Differential Revision: https://phab.mercurial-scm.org/D8393
Augie Fackler <augie@google.com> [Thu, 02 Apr 2020 16:02:21 -0400] rev 44708
manifest: leave a TODO where we may have more work for sha1 portability
Differential Revision: https://phab.mercurial-scm.org/D8375
Augie Fackler <augie@google.com> [Wed, 08 Jan 2020 11:33:01 -0500] rev 44707
manifest: add test for long hashes
Differential Revision: https://phab.mercurial-scm.org/D8373
Augie Fackler <augie@google.com> [Thu, 02 Apr 2020 16:24:34 -0400] rev 44706
manifest: remove a final 40-byte assumption from pure-python parser
Differential Revision: https://phab.mercurial-scm.org/D8372
Augie Fackler <augie@google.com> [Thu, 02 Apr 2020 16:01:36 -0400] rev 44705
manifest: fix yet another 20-byte-hash assumption
Differential Revision: https://phab.mercurial-scm.org/D8371
Augie Fackler <augie@google.com> [Thu, 02 Apr 2020 15:57:07 -0400] rev 44704
manifest: fix another pure-parsing 20-byte assumption
Differential Revision: https://phab.mercurial-scm.org/D8370
Augie Fackler <augie@google.com> [Thu, 02 Apr 2020 15:52:10 -0400] rev 44703
manifest: teach treemanifest about long hashes
Differential Revision: https://phab.mercurial-scm.org/D8369
Augie Fackler <augie@google.com> [Tue, 07 Jan 2020 13:47:29 -0500] rev 44702
manifest: start removing 40-byte hash restrictions from C code
Differential Revision: https://phab.mercurial-scm.org/D8368
Augie Fackler <augie@google.com> [Tue, 07 Jan 2020 11:25:13 -0500] rev 44701
manifest: remove a 20-byte-hash assumption from pure manifest parsing
This is almost certainly slower, but it shouldn't be by much. We need
to do this for the eventual hash transition, no matter what that looks like.
Differential Revision: https://phab.mercurial-scm.org/D8367
Matt Harbison <matt_harbison@yahoo.com> [Mon, 13 Apr 2020 12:18:29 -0400] rev 44700
revset: mark `extinct()` experimental
Differential Revision: https://phab.mercurial-scm.org/D8407
Matt Harbison <matt_harbison@yahoo.com> [Mon, 13 Apr 2020 12:17:52 -0400] rev 44699
revset: mark `successors()` experimental
Differential Revision: https://phab.mercurial-scm.org/D8406
Matt Harbison <matt_harbison@yahoo.com> [Mon, 13 Apr 2020 12:13:22 -0400] rev 44698
revset: mark `obsolete()` experimental
I assume not marking it was unintentional, since the obsolete concept it still
experimental.
Differential Revision: https://phab.mercurial-scm.org/D8405
Pierre-Yves David <pierre-yves.david@octobus.net> [Tue, 14 Apr 2020 16:09:38 +0200] rev 44697
nodemap: access the mmap opener option using bytes
The key is bytes and need to be accessed s such.
Differential Revision: https://phab.mercurial-scm.org/D8413