Bryan O'Sullivan <bryano@fb.com> [Tue, 18 Dec 2012 17:15:13 -0800] rev 18097
posix: move server side of unix domain sockets out of inotify
We also turn the unix domain socket into a class, so that we have a
sensible place to hang its logically related attributes and behaviour.
We'll shortly want to reuse this in other code.
Bryan O'Sullivan <bryano@fb.com> [Tue, 18 Dec 2012 17:33:32 -0800] rev 18096
inotify: don't fall over just because of a dangling symlink
Previously, the inotify server failed to start if .hg/inotify.sock was
a symlink that pointed to a non-existent path. This behaviour does not
seem to make any sense.
Now, if we encounter a broken symlink, we unlink it and continue.
Bryan O'Sullivan <bryano@fb.com> [Wed, 19 Dec 2012 10:40:34 -0800] rev 18095
test-inotify: test symlink indirection for unix sockets
The inotify code performs a delicate dance to work around the 108-byte
limit on unix domain socket path names on Linux.
This change sets us up to safely refactor that code without breaking
it. (It is redundant with part of test-inotify-issue1208.t, but we will
shortly make that test go away.)
Adrian Buehlmann <adrian@cadifra.com> [Wed, 19 Dec 2012 10:02:43 +0100] rev 18094
test-pathencode: compare current pathencoding implementations
We already have two implementations of the pathencoding (C and
Python) and this test can perfectly well be used to probabilistically
test them instead of just wasting CPU cycles and test time.
Siddharth Agarwal <sid0@fb.com> [Mon, 17 Dec 2012 20:51:21 -0800] rev 18093
rebase: use lazy ancestor membership testing
For a repository with over 400,000 commits, rebasing one revision near tip,
this avoids one walk up the DAG, speeding the operation up by around 0.8
seconds.
Siddharth Agarwal <sid0@fb.com> [Mon, 17 Dec 2012 20:43:37 -0800] rev 18092
localrepo: use lazy ancestor membership testing
For a repository with over 400,000 commits, rebasing one revision near tip,
this avoids two treks up the DAG, speeding the operation up by around 1.6
seconds.
Siddharth Agarwal <sid0@fb.com> [Tue, 18 Dec 2012 12:47:20 -0800] rev 18091
ancestor: add lazy membership testing to lazyancestors
This also makes the perfancestorset command use lazy membership testing. In a
linear repository with over 400,000 commits, without this patch, hg
perfancestorset takes 0.80 seconds no matter how far behind we're looking.
With this patch, hg perfancestorset -- X takes:
Rev X Time
-1 0.00s
-4000 0.01s
-20000 0.04s
-80000 0.17s
-200000 0.43s
-300000 0.69s
0 0.88s
Thus, for revisions close to tip, we're up to several orders of magnitude
faster. At 0 we're around 10% slower.
Siddharth Agarwal <sid0@fb.com> [Tue, 18 Dec 2012 10:14:01 -0800] rev 18090
revlog: move ancestor generation out to a new class
This refactoring is to prepare for implementing lazy membership.
Bryan O'Sullivan <bryano@fb.com> [Mon, 17 Dec 2012 15:57:02 -0800] rev 18089
ignore: process hgignore files in deterministic order
Previously, we processed them in whatever order the dict iterator gave us.
Bryan O'Sullivan <bryano@fb.com> [Mon, 17 Dec 2012 15:52:44 -0800] rev 18088
ignore: only read an ignore file once
Bryan O'Sullivan <bryano@fb.com> [Mon, 17 Dec 2012 16:23:37 -0800] rev 18087
ignore: refactor ignore into two functions
This prepares us for eventually being able to hash the list of patterns
in use.
Pierre-Yves David <pierre-yves.david@logilab.fr> [Mon, 10 Dec 2012 18:12:41 +0100] rev 18086
clfilter: fix `nodemap` usage in `getbundle`
With the current implementation, `changelog.nodemap` is not filtered. So some
filtered changeset in common are not filtered by `n in nodemap`. This leads to
crash lower in the stack when the bundle generation try to access those node on
a filtered changelog.
Pierre-Yves David <pierre-yves.david@logilab.fr> [Mon, 17 Dec 2012 17:00:24 +0100] rev 18085
test: fix truncated comment in test
The push is made to ensure repository are related. The comment in the initial
changeset got truncated somehow.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Mon, 17 Dec 2012 18:09:41 +0100] rev 18084
clfilter: ensure context raise RepoLookupError when the revision is filtered
Currently the code path of `changectx(filteredrepo, rev)` call
`filteredrepo.changelog.node(rev)`. When `rev` is filtered this raise an
unhandled `IndexError`. This case now raise a `RepoLookupError` as other
error case do.
Siddharth Agarwal <sid0@fb.com> [Mon, 17 Dec 2012 15:08:37 -0800] rev 18083
revlog: remove incancestors since it is no longer used
Siddharth Agarwal <sid0@fb.com> [Mon, 17 Dec 2012 15:08:06 -0800] rev 18082
transplant: replace incancestors uses with ancestors
Siddharth Agarwal <sid0@fb.com> [Mon, 17 Dec 2012 15:13:51 -0800] rev 18081
revlog.ancestors: add support for including revs
This is in preparation for an upcoming refactoring. This also fixes a bug in
incancestors, where if an element of revs was an ancestor of another it would
be generated twice.
Siddharth Agarwal <sid0@fb.com> [Fri, 14 Dec 2012 10:23:18 -0800] rev 18080
perf: add command to test performance of membership in ancestor set
The new command, perfancestorset, takes an argument denoting which revset to
test the membership of.
Currently this runs through all the ancestors and converts them into a set.
The primary purpose of having this is to compare this approach, currently used
in several places, against the upcoming lazy approach.
Siddharth Agarwal <sid0@fb.com> [Tue, 11 Dec 2012 14:47:33 -0800] rev 18079
ancestor: move missingancestors doctest out into a separate file
This is in preparation for upcoming patches which will reuse the same graph
for tests.
Kevin Bullock <kbullock@ringworld.org> [Sun, 16 Dec 2012 23:02:54 -0600] rev 18078
merge with crew-stable
Tim Henigan <tim.henigan@gmail.com> [Thu, 29 Nov 2012 08:44:54 -0500] rev 18077
dirstate: remove obsolete comment from setbranch
This comment should have been removed in b74361cf7c0a, when the call
to scmutil.checknewlabel was removed.
Idan Kamara <idankk86@gmail.com> [Sat, 15 Dec 2012 20:19:07 +0200] rev 18076
dirstate: don't rename branch file if writing it failed
Kevin Bullock <kbullock@ringworld.org> [Sun, 16 Dec 2012 22:00:38 -0600] rev 18075
commands: 'hg bookmark NAME' should work even with ui.strict=True
Before this patch, enabling strict command processing (ui.strict=True)
meant that 'hg bookmark NAME', as referenced several places in the
documentation, would not work. This adds 'bookmark' as an explicit alias
to 'bookmarks'.
Mads Kiilerich <mads@kiilerich.com> [Sun, 16 Dec 2012 19:41:56 +0100] rev 18074
tests: improve test-mv-cp-st-diff.t readability
Pierre-Yves David <pierre-yves.david@logilab.fr> [Sat, 10 Nov 2012 02:19:20 +0100] rev 18073
obsolete: refuse to push divergent changeset
As other troubles `unstable` and `bumped`. Followup patches may simplify the
push code with unification of "obsolescence troubles" handling.
Pierre-Yves David <pierre-yves.david@logilab.fr> [Wed, 12 Dec 2012 03:20:49 +0100] rev 18072
obsolete: add a divergent method on context
The same we have `unstable` and `bumped`. Convenient method to access troubles
information in general may land later.
This get actual use and testing in the next changesets.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Wed, 12 Dec 2012 03:12:55 +0100] rev 18071
obsolete: add revset and test for divergent changesets
This changesets add a new `divergent()` revset similar to `unstable()` and
`bumped()` one. Introducting this revset allows actuall test of the divergent
detection.
Pierre-Yves David <pierre-yves.david@logilab.fr> [Wed, 12 Dec 2012 03:19:30 +0100] rev 18070
obsolete: detect divergent changesets
Divergent changeset are final successors (non obsolete) of a changeset who
compete with another set of final successors for this same changeset.
For example if you have two obsolescence markers A -> B and A -> C, B and C are
both "divergent" because they compete to be the one true successors of A.
Public revision can't be divergent.
This function is used and tested in the next changeset.
Pierre-Yves David <pierre-yves.david@logilab.fr> [Sat, 10 Nov 2012 01:56:59 +0100] rev 18069
obsolete: drop successors sets which are subset of another one
If both "(B,)" and "(B, C)" are successors set of "A", "(B,)" is dropped.
We won't be interrested in detection such divergence scenario.
Pierre-Yves David <pierre-yves.david@logilab.fr> [Thu, 13 Dec 2012 15:38:43 +0100] rev 18068
obsolete: compute successors set
Successors set are an important part of obsolescence. It is necessary to detect
and solve divergence situation. This changeset add a core function to compute
them, a debug command to audit them and solid test on the concept.
Check function docstring for details about the concept.
Mads Kiilerich <mads@kiilerich.com> [Sun, 09 Dec 2012 00:25:21 +0100] rev 18067
mq: don't fail when removing a patch without patch file from series file
Mads Kiilerich <madski@unity3d.com> [Thu, 13 Dec 2012 19:19:06 +0100] rev 18066
largefiles: align rm warnings with warnings used in core
Mads Kiilerich <madski@unity3d.com> [Thu, 13 Dec 2012 21:34:55 +0100] rev 18065
merge with stable
Levi Bard <levi@unity3d.com> [Mon, 10 Dec 2012 14:58:42 +0100] rev 18064
largefiles: commit directories that only contain largefiles (issue3548)
If we pass a directory to commit whose only commitable files
are largefiles, the core commit code aborts before finding
the largefiles.
So we do the following:
For directories that only have largefiles as matches,
we explicitly add the largefiles to the matchlist and remove
the directory.
In other cases, we leave the match list unmodified.
Siddharth Agarwal <sid0@fb.com> [Fri, 07 Dec 2012 10:37:43 -0800] rev 18063
revset.children: ignore rev numbers that are too low
This replaces unnecessary parentrevs() calls with calculating min(parentset).
Even though the min operation is O(size of parentset), since parentrevs is
relatively expensive, this tradeoff almost always works in our favour. In a
repository with over 400,000 changesets, hg perfrevset "children(X)" takes:
Set X Before After
-1 0.51s 0.06s
-1000: 0.55s 0.08s
-10000: 0.56s 0.10s
-100000: 0.60s 0.25s
-100000:-99000 0.55s 0.19s
0:100000 0.60s 0.61s
all() 0.72s 0.74s
The relative performance is similar for Mercurial's own repository -- several
times faster in most cases, slightly slower for revisions close to 0 and
all().
Siddharth Agarwal <sid0@fb.com> [Thu, 06 Dec 2012 11:21:11 -0800] rev 18062
perf: add a command to measure revset performance
Bryan O'Sullivan <bos@serpentine.com> [Wed, 12 Dec 2012 15:17:18 -0800] rev 18061
run-tests: fix whitespace nonsense
Bryan O'Sullivan <bos@serpentine.com> [Wed, 12 Dec 2012 15:12:28 -0800] rev 18060
merge with mpm
Bryan O'Sullivan <bryano@fb.com> [Wed, 12 Dec 2012 14:52:58 -0800] rev 18059
run-tests: on windows, put correct python at front of PATH
The older approach of trying to copy the python executable into the test
directory was doomed to fail.
There remains one weakness with this approach: if you've run "make local",
tests may pick up the wrong extension DLLs from inside the source tree. I
don't know why this happens.
A reasonable workaround for now is to test either using --local or with
a working directory that does not contain built DLLs.
Siddharth Agarwal <sid0@fb.com> [Tue, 11 Dec 2012 15:38:42 -0800] rev 18058
run-tests: fix exename on Windows
Bryan O'Sullivan <bryano@fb.com> [Tue, 11 Dec 2012 15:13:23 -0800] rev 18057
run-tests: support running tests in parallel on windows
Previously, we used os.spawnvp, which doesn't exist on Windows, and
isn't needed anyway (the command line begins with an absolute path).
We also need a slightly more convoluted way to wait for processes
without specifying an order on Windows, as it lacks os.wait.
Bryan O'Sullivan <bryano@fb.com> [Tue, 11 Dec 2012 13:44:00 -0800] rev 18056
makefile: allow local builds to work on windows/mingw32
Mads Kiilerich <madski@unity3d.com> [Sun, 09 Dec 2012 23:33:16 +0100] rev 18055
check-code: make 'missing whitespace in assignment' more aggressive
New warnings:
> a.b=ab
missing whitespace in assignment
(the pattern did not accept '.' on the left hand side)
> a=a
missing whitespace in assignment
(the right hand side pattern never matched a single character)
> a=a + 7
missing whitespace in assignment
(the pattern only matched one character after the identifier following =)
Mads Kiilerich <madski@unity3d.com> [Sun, 09 Dec 2012 23:33:16 +0100] rev 18054
check-code: there must also be whitespace between ')' and operator
The check pattern only checked for whitespace between keyword and operator.
Now it also warns:
> x = f(),7
missing whitespace after ,
> x = f()+7
missing whitespace in expression
Mads Kiilerich <madski@unity3d.com> [Sun, 09 Dec 2012 23:33:16 +0100] rev 18053
rm: drop misleading 'use -f' hint for the rm --after 'not removing' warning
A warning mentioning 'forgetting' or 'recording delete' would be more correct
than 'not removing' but also more confusing.
Mads Kiilerich <madski@unity3d.com> [Sun, 09 Dec 2012 23:33:16 +0100] rev 18052
contrib: state explicitly how hgtest.vim can be installed
Mads Kiilerich <madski@unity3d.com> [Sun, 09 Dec 2012 23:33:16 +0100] rev 18051
tests: kill daemons early, making breaking at "Accept" prompt safe
Bryan O'Sullivan <bryano@fb.com> [Mon, 10 Dec 2012 12:14:55 -0800] rev 18050
run-tests: use correct python safely under --jobs
Bryan O'Sullivan <bryano@fb.com> [Mon, 10 Dec 2012 12:09:02 -0800] rev 18049
run-tests: check for the correct python when starting
Bryan O'Sullivan <bryano@fb.com> [Mon, 10 Dec 2012 12:07:42 -0800] rev 18048
run-tests: use correct python when run with --local
Bryan O'Sullivan <bryano@fb.com> [Mon, 10 Dec 2012 11:18:03 -0800] rev 18047
synthrepo: do not crash if a list is empty
Angel Ezquerra <angel.ezquerra@gmail.com> [Tue, 27 Nov 2012 22:24:02 +0100] rev 18046
hgwebdir: do not show RSS and Atom links for plain directories
Up until now the templates that show RSS and Atom feeds on the "repository
lists" (i.e. gitweb and monoblue) showed them for all entries, including regular
folders. Clicking on those "folder RSS" links would result in an error page
being shown.
This patch hides those links for regular folders.
Angel Ezquerra <angel.ezquerra@gmail.com> [Tue, 04 Dec 2012 00:37:50 +0100] rev 18045
hgweb: add branches RSS and Atom feeds
There were no RSS nor Atom feeds for the branches page. Different hgweb
templates linked to different feeds on their branches page (some linked to the
tags feed, some to the log feed and some to the unexisting branches feed).
Kevin Bullock <kbullock@ringworld.org> [Fri, 07 Dec 2012 12:03:37 -0600] rev 18044
push: reunite comment with the line of code it describes
Kevin Bullock <kbullock@ringworld.org> [Fri, 07 Dec 2012 11:30:38 -0600] rev 18043
bookmarks: spelling correction in docstring
Kevin Bullock <kbullock@ringworld.org> [Thu, 06 Dec 2012 22:07:44 -0600] rev 18042
merge: fix mistake in moved _checkcollision call from 5881d5b7552f
Matt Mackall <mpm@selenic.com> [Thu, 06 Dec 2012 16:56:44 -0600] rev 18041
windows: correctly pass a mode to S_IFMT in statfiles
Siddharth Agarwal <sid0@fb.com> [Wed, 05 Dec 2012 14:33:15 -0800] rev 18040
strip: make query to get new bookmark target cheaper
The current query to get the new bookmark target for stripped revisions
involves multiple walks up the DAG, and is really expensive, taking over 2.5
seconds on a repository with over 400,000 changesets even if just one
changeset is being stripped.
A slightly simplified version of the current query is
max(heads(::<tostrip> - <tostrip>))
We make two observations here.
1. For any set s, max(heads(s)) == max(s). That is because revision numbers
define a topological order, so that the element with the highest revision
number in s will not have any children in s.
2. For any set s, max(::s - s) == max(parents(s) - s). In other words, the
ancestor of s with the highest revision number not in s is a parent of one
of the revs in s. Why? Because if it were an ancestor but not a parent of s,
it would have a descendant that would be a parent of s. This descendant
would have a higher revision number, leading to a contradiction.
Combining these two observations, we rewrite the revset query as
max(parents(<tostrip>) - <tostrip>)
The time complexity is now linear in the number of changesets being stripped.
For the above repository, the query now takes 0.1 seconds when one changeset
is stripped. This speeds up operations that use repair.strip, like the rebase
and strip commands.
David Schleimer <dschleimer@fb.com> [Tue, 04 Dec 2012 12:54:18 -0800] rev 18039
graft: explicit current node tracking
This changes graft to explicitly track the progression of commits it
makes, and updates it's idea of the current node based on it's last
commit, rather than from the working copy parent. This should have no
effect on the value of current since we were reading the working copy
parent immediately after commiting to it.
The motivation for this change is that a subsequent patch will break
the current node and working copy relationship. Splitting this out
into a separate patch will make that one more readible.
David Schleimer <dschleimer@fb.com> [Tue, 04 Dec 2012 12:54:18 -0800] rev 18038
graft: move commit info building
This moves the logic for generating the commit metadata ahead of the
merge operation. The only purposae of this patch is to make
subsequent patches easier to read, and there should be no behavior
changes.