zphricz <zack.hricz@oculus.com> [Wed, 21 Dec 2016 13:51:34 -0800] rev 30653
fsmonitor: fix exception message scraping
pywatchman.CommandError formats its error message such that
'unable to resolve root' is not a prefix. This change fixes that by
instead just searching for it as a substring.
David Soria Parra <davidsp@fb.com> [Wed, 21 Dec 2016 13:44:15 -0800] rev 30652
tests: add magic string to bypass check-commit
Allow bypassing test-check-commit runs by specifying '# no-check-commit'
in the description. This should be avoided but is useful for upstream
imports such as pywatchman which will cause check-code to fail otherwise.
Gregory Szorc <gregory.szorc@gmail.com> [Sun, 18 Dec 2016 12:53:20 -0800] rev 30651
convert: config option for git rename limit
By default, Git applies rename and copy detection to 400 files. The
diff.renamelimit config option and -l argument to diff commands can
override this.
As part of converting some repositories in the wild, I was hitting
the default limit. Unfortunately, the warnings that Git prints in this
scenario are swallowed because the process running functionality in
common.py redirects stderr to /dev/null by default. This seems like
a bug, but a bug for another day.
This commit establishes a config option to send the rename limit
through to `git diff-tree`. The added tests demonstrate a too-low
rename limit doesn't result in copy metadata being recorded.
Yuya Nishihara <yuya@tcha.org> [Thu, 22 Dec 2016 01:09:45 +0900] rev 30650
chgserver: backout changeset dfb19aed409e (per discussion)
On Wed, 21 Dec 2016 15:39:05 +0000, Jun Wu wrote:
> Actually, patch 1 is unnecessary if we go with the "ui._runpager" approach.
> Maybe someone can drop it without adding too many markers.
Jun Wu <quark@fb.com> [Fri, 16 Dec 2016 14:48:37 +0000] rev 30649
chgserver: override runcommand
Next patches will customize chgserver's runcommand. So let's override it.
The docstring is temporarily missing and will be filled later.
Jun Wu <quark@fb.com> [Fri, 16 Dec 2016 14:46:34 +0000] rev 30648
chgserver: store csystem separately
Previously, the "system" channel is inside the ui object. In the future, chg
will let dispatch to create a new ui object from scratch, to maximize
compatibility. And chgserver will use a "uisetup" like an extension to wrap
ui.system. To be able to do that cleanly, the system channel needs to be
accessed directly.
Pulkit Goyal <7895pulkit@gmail.com> [Mon, 19 Dec 2016 02:26:41 +0530] rev 30647
py3: replace sys.platform with pycompat.sysplatform (part 2 of 2)
Pulkit Goyal <7895pulkit@gmail.com> [Mon, 19 Dec 2016 02:15:24 +0530] rev 30646
py3: replace sys.platform with pycompat.sysplatform (part 1 of 2)
sys.platform returns unicode on python 3 world. Our code base has most of the
things bytes because of the transformer. So we have a bytes version of this as
pycompat.sysplatform. This series of 2 patches replaces occurences of
sys.platform with pycompat.sysplatform.
Pulkit Goyal <7895pulkit@gmail.com> [Mon, 19 Dec 2016 00:28:12 +0530] rev 30645
py3: replace os.name with pycompat.osname (part 2 of 2)
Pulkit Goyal <7895pulkit@gmail.com> [Mon, 19 Dec 2016 00:16:52 +0530] rev 30644
py3: replace os.name with pycompat.osname (part 1 of 2)
os.name returns unicodes on py3 and we have pycompat.osname which returns
bytes. This series of 2 patches will change every ocurrence of os.name with
pycompat.osname.
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 02:08:59 +0530] rev 30643
py3: replace os.environ with encoding.environ (part 5 of 5)
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 02:06:00 +0530] rev 30642
py3: replace os.environ with encoding.environ (part 4 of 5)
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 01:54:36 +0530] rev 30641
py3: replace os.environ with encoding.environ (part 3 of 5)
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 01:46:39 +0530] rev 30640
py3: replace os.environ with encoding.environ (part 2 of 5)
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 01:34:41 +0530] rev 30639
py3: replace os.environ with encoding.environ (part 1 of 5)
os.environ is a dictionary which has string elements on Python 3. We have
encoding.environ which take care of all these things. This is the first patch
of 5 patch series which tend to replace the occurences of os.environ with
encoding.environ as using os.environ will result in unusual behaviour.
David Soria Parra <davidsp@fb.com> [Tue, 20 Dec 2016 20:28:41 -0800] rev 30638
tests: exclude bundled pywatchman from check-code test
pywatchman is imported from upstream and therefore fails to pass
linting. We have added 'no-check-code' manually to every file in the
past. This is cumbersome and modifies upstream sources.
David Soria Parra <davidsp@fb.com> [Tue, 20 Dec 2016 09:23:50 -0800] rev 30637
convert: parse perforce data on-demand
We are using read-only attributes that parse the perforce data on
demand. We are reading the data only once whenever an attribute is
requested and use it throughout the import process. This is equivalent
to the previous behavior, but we are avoiding reading from perforce when
we initialize the object, but instead run it during the actual import
process, when the first attribute is requested (usually getheads(), see
`convertcmd.convert`).
David Soria Parra <davidsp@fb.com> [Tue, 20 Dec 2016 09:23:50 -0800] rev 30636
convert: return calculated values from parse() instead of manpulating state
David Soria Parra <davidsp@fb.com> [Tue, 20 Dec 2016 09:23:50 -0800] rev 30635
convert: move localname state to function scope
David Soria Parra <davidsp@fb.com> [Tue, 20 Dec 2016 09:23:50 -0800] rev 30634
convert: use return value in parse_view() instead of manipulating state
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Mon, 19 Dec 2016 04:25:18 +0100] rev 30633
changegroup: simplify logic around enabling changegroup 03
There was multiple spot that took care of adding '03' as supported changegroup
version for different condition. We gather them all in one location for
simplicity.
The 'supportedincomingversions' function is now doing nothing, but I kept it
around because it looks like a great hooking point for extension.
(Note that we should probably just get changegroup3 out of experimental now, But
that would be a patch with a much wider scope).
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Mon, 19 Dec 2016 04:29:33 +0100] rev 30632
changegroup: pass 'repo' to allsupportedversions
In the next changesets, we will introduce more logic directly related to the
repository to decide what version have to be supported. So we now directly pass
the repo object instead of just ui.
Pierre-Yves David <pierre-yves.david@ens-lyon.org> [Mon, 19 Dec 2016 04:31:13 +0100] rev 30631
changegroup: simplify 'allsupportedversions' logic
Discarding '03' to add it back is a bit strange. Instead we only discard it when
needed.
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 01:17:12 +0530] rev 30630
py3: replace os.altsep with pycompat.altsep
All the occurences of os.altsep are replaced with pycompat.altsep which
returns bytes.
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 00:52:05 +0530] rev 30629
py3: have a bytes version of sys.platform
sys.platform returns unicodes on Python 3. This patch adds up
pycompat.sysplatform which returns bytes.
Pulkit Goyal <7895pulkit@gmail.com> [Sun, 18 Dec 2016 00:44:21 +0530] rev 30628
py3: have a bytes version of os.altsep
os.altsep returns unicodes on Python 3. We need a bytes version hence added
pycompat.altsep.
Pulkit Goyal <7895pulkit@gmail.com> [Sat, 17 Dec 2016 23:55:25 +0530] rev 30627
py3: make sure encoding.encoding is a bytes variable
encoding.encoding returns unicodes when locale.getpreferredencoding() is used
to get the preferred encoding. This patch fixes that.
Jun Wu <quark@fb.com> [Mon, 19 Dec 2016 22:15:00 +0000] rev 30626
chg: remove locks
See the previous two patches for the reason. The advantage is a simplified
code base and better throughput when starting multiple servers with multiple
confighashes. The disadvantage is starting multiple servers in parallel with
a single confighash will waste some CPU time, which is probably fine in
common use-cases.
This makes it easier to switch to relative paths to support long unix domain
socket paths.
Jun Wu <quark@fb.com> [Mon, 19 Dec 2016 22:09:49 +0000] rev 30625
chg: start server at a unique address
See the previous patch for motivation. Previously, the server is started at
a globally shared address. This patch appends pid to the address so it
becomes unique.
Note: with Linux pid namespace, the address may be non-unique, but it does
not affect correctness of chg - chg client will receive an redirection and
that's it.
Jun Wu <quark@fb.com> [Mon, 19 Dec 2016 22:07:41 +0000] rev 30624
chgserver: truncate base address at "." for hash address
Previously, the hash address is just appending "-$HASH" to base address.
This patch makes it truncate the basename address at "." before appending
"-$HASH".
This makes it possible to spawn new servers in a racy situation and the
client could be sure the server it connects is the new server just spawned.
This is a step towards removing the lock.
One of the functionalities of the lock is to make sure the connect will
connect to a server it just created:
1. start server --address foo
2. connect to foo # wish "foo" is the server just started
With this change, the client could do:
1. start server --address foo.tmp$PID
2. connect to foo.tmp$PID # is the server just started
(note: if it is not, it does not affect correctness - linux pid
namespace is not a concern here)
3. rename foo.tmp$PID to foo
Another functionality of the lock is to avoid starting multiple servers with
a same confighash in parallel. But that also prevents starting multiple
servers with different confighashes in parallel.