Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Tue, 09 Aug 2016 17:06:35 +0200] rev 29819
computeoutgoing: move the function from 'changegroup' to 'exchange'
Now that all users are in exchange, we can safely move the code in the
'exchange' module. This function is really about processing the argument of a
'getbundle' call, so it even makes senses to do so.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Tue, 09 Aug 2016 17:00:38 +0200] rev 29818
getchangegroup: take an 'outgoing' object as argument (API)
There is various version of this function that differ mostly by the way they
define the bundled set. The flexibility is now available in the outgoing object
itself so we move the complexity into the caller themself. This will allow use
to remove a good share of the similar function to obtains a changegroup in the
'changegroup.py' module.
An important side effect is that we stop calling 'computeoutgoing' in
'getchangegroup'. This is fine as code that needs such argument processing
is actually going through the 'exchange' module which already all this function
itself.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Tue, 09 Aug 2016 22:31:38 +0200] rev 29817
outgoing: add a 'missingroots' argument
This argument can be used instead of 'commonheads' to determine the 'outgoing'
set. We remove the outgoingbetween function as its role can now be handled by
'outgoing' itself.
I've thought of using an external function instead of making the constructor
more complicated. However, there is low hanging fruit to improve the current
code flow by storing some side products of the processing of 'missingroots'. So
in my opinion it make senses to add all this to the class.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Tue, 09 Aug 2016 15:55:44 +0200] rev 29816
outgoing: adds some default value for argument
We are about to introduce a third option to create an outgoing object:
'missingroots'. This argument will be mutually exclusive with 'commonheads' so
we implement some default value handling in preparation.
This will also help use to make more use of outgoing creation around the code
base.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Tue, 09 Aug 2016 15:26:53 +0200] rev 29815
outgoing: pass a repo object to the constructor
We are to introduce more code constructing such object in the code base. It will
be more convenient to pass a repository object, all current users already
operate at the repository level anyway. More changes to the contructor argument
are coming in later changeset.
Hannes Oldenburg <hannes.christian.oldenburg@gmail.com> [Tue, 16 Aug 2016 08:21:16 +0000] rev 29814
match: remove matchessubrepo method (API)
Since it is no more used in cmdutil.{files,remove} and scmutil.addremove
we remove this method.
Hannes Oldenburg <hannes.christian.oldenburg@gmail.com> [Tue, 16 Aug 2016 08:15:12 +0000] rev 29813
subrepo: cleanup of subrepo filematcher logic
Previously in the worst case we iterated the files in matcher twice and
had a method only for this, which reimplemented logic in subdirmatchers
constructor. So we replaced the method with a subdirmatcher.files() call.
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 14:46:24 +0900] rev 29812
pycompat: delay loading modules registered to stub
Replacement _pycompatstub designed to be compatible with our demandimporter.
try-except is replaced by version comparison because ImportError will no longer
be raised immediately.
Yuya Nishihara <yuya@tcha.org> [Tue, 16 Aug 2016 12:35:15 +0900] rev 29811
py3: import builtin wrappers automagically by code transformer
This should be less invasive than mucking builtins.
Since tokenize.untokenize() looks start/end positions of tokens, we calculates
them from the NEWLINE token of the future import.
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 12:51:21 +0900] rev 29810
py3: provide (del|get|has|set)attr wrappers that accepts bytes
These functions will be imported automagically by our code transformer.
getattr() and setattr() are widely used in our code. We wouldn't probably
want to rewrite every single call of getattr/setattr. delattr() and hasattr()
aren't that important, but they are functions of the same kind.
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 12:44:13 +0900] rev 29809
py3: check python version to enable builtins hack
Future patches will add (del|get|has|set)attr wrappers.
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 12:41:54 +0900] rev 29808
py3: move xrange alias next to import lines
Builtin functions should be available in compatibility code.
Yuya Nishihara <yuya@tcha.org> [Tue, 16 Aug 2016 17:15:54 +0900] rev 29807
check-code: allow assignment to hasattr variable
Yuya Nishihara <yuya@tcha.org> [Mon, 15 Aug 2016 16:07:55 +0900] rev 29806
debugobsolete: add formatter support (issue5134)
It appears that computing index isn't cheap if --rev is specified. That's
why "index" field is available only if --index is specified.
I've named marker.flags() as "flag" because "flags" implies a list or dict
in template world.
Thanks to Piotr Listkiewicz for the initial implementation of this patch.
Yuya Nishihara <yuya@tcha.org> [Mon, 15 Aug 2016 12:58:33 +0900] rev 29805
formatter: add function to convert dict to appropriate format
This will be used to process key-value pairs by formatter. The default
field names and format are derived from the {extras} template keyword.
Tests will be added later.
Yuya Nishihara <yuya@tcha.org> [Mon, 15 Aug 2016 17:17:39 +0900] rev 29804
check-code: make dict() pattern less invasive
'foodict(x=y)' should be allowed.
Gregory Szorc <gregory.szorc@gmail.com> [Sun, 14 Aug 2016 21:29:46 -0700] rev 29803
hgweb: tweak zlib chunking behavior
When doing streaming compression with zlib, zlib appears to emit chunks
with data after ~20-30kb on average is available. In other words, most
calls to compress() return an empty string. On the mozilla-unified repo,
only 48,433 of 921,167 (5.26%) of calls to compress() returned data.
In other words, we were sending hundreds of thousands of empty chunks
via a generator where they touched who knows how many frames (my guess
is millions). Filtering out the empty chunks from the generator
cuts down on overhead.
In addition, we were previously feeding 8kb chunks into zlib
compression. Since this function tends to emit *compressed* data after
20-30kb is available, it would take several calls before data was
produced. We increase the amount of data fed in at a time to 32kb.
This reduces the number of calls to compress() from 921,167 to
115,146. It also reduces the number of output chunks from 48,433 to
31,377. This does increase the average output chunk size by a little.
But I don't think this will matter in most scenarios.
The combination of these 2 changes appears to shave ~6s CPU time
or ~3% from a server serving the mozilla-unified repo.
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 17:07:05 +0900] rev 29802
test-gpg: run migration of v1 secret keys beforehand
This suppresses unwanted output at "hg sign".
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 17:01:33 +0900] rev 29801
test-gpg: start gpg-agent under control of the test runner
GnuPG v2 automatically starts gpg-agent. We should kill the daemon process.
Yuya Nishihara <yuya@tcha.org> [Sun, 14 Aug 2016 16:49:47 +0900] rev 29800
test-gpg: make temporary copy of GNUPGHOME
GnuPG v2 will convert v1 secret keys and create a socket under $GNUPGHOME.
This patch makes sure no state would persist.
We no longer need to verify trustdb.gpg, which was added by aae219a99a6e.
Gregory Szorc <gregory.szorc@gmail.com> [Mon, 15 Aug 2016 20:39:33 -0700] rev 29799
hgweb: document why we don't allow untrusted settings to control zlib
Added comment per discussion on mercurial-devel.
Gregory Szorc <gregory.szorc@gmail.com> [Sun, 14 Aug 2016 18:37:24 -0700] rev 29798
hgweb: profile HTTP requests
Currently, running `hg serve --profile` doesn't yield anything useful:
when the process is terminated the profiling output displays results
from the main thread, which typically spends most of its time in
select.select(). Furthermore, it has no meaningful results from
mercurial.* modules because the threads serving HTTP requests don't
actually get profiled.
This patch teaches the hgweb wsgi applications to profile individual
requests. If profiling is enabled, the profiler kicks in after
HTTP/WSGI environment processing but before Mercurial's main request
processing.
The profile results are printed to the configured profiling output.
If running `hg serve` from a shell, they will be printed to stderr,
just before the HTTP request line is logged. If profiling to a file,
we only write a single profile to the file because the file is not
opened in append mode. We could add support for appending to files
in a future patch if someone wants it.
Per request profiling doesn't work with the statprof profiler because
internally that profiler collects samples from the thread that
*initially* requested profiling be enabled. I have plans to address
this by vendoring Facebook's customized statprof and then improving
it.
Gregory Szorc <gregory.szorc@gmail.com> [Sun, 14 Aug 2016 16:03:30 -0700] rev 29797
hgweb: abstract call to hgwebdir wsgi function
The function names and behavior now matches hgweb. The reason for this
will be obvious in the next patch.
Gregory Szorc <gregory.szorc@gmail.com> [Sun, 14 Aug 2016 18:28:43 -0700] rev 29796
profiling: don't error with statprof when profiling has already started
statprof.reset() asserts if profiling has already started. So don't
call if it profiling is already running.