This is an archive of the discontinued Mercurial Phabricator instance.

Differential D2380

sshpeer: return framed file object when needed
ClosedPublic

Authored by indygreg on Feb 21 2018, 5:08 PM.

Details

Reviewers
yuja
Group Reviewers
hg-reviewers
Commits
rHG043e77f3be09: sshpeer: return framed file object when needed
Summary

Currently, wireproto.wirepeer has a default implementation of
_submitbatch() and sshv1peer has a very similar implementation.
The main difference is that sshv1peer is aware of the total amount
of bytes it can read whereas the default implementation reads the
stream until no more data is returned. The default implementation
works for HTTP, since there is a known end to HTTP responses (either
Content-Length or 0 sized chunk).

This commit teaches sshv1peer to use our just-introduced "cappedreader"
class for wrapping a file object to limit the number of bytes that
can be read. We do this by introducing an argument to specify whether
the response is framed. If set, we returned a cappedreader instance
instead of the raw pipe.

_call() always has framed responses. So we set this argument
unconditionally and then .read() the entirety of the result.
Strictly speaking, we don't need to use cappedreader in this case
and can inline frame decoding/read logic. But I like when things
are consistent. The overhead should be negligible.

_callstream() and _callcompressable() are special: whether framing
is used depends on the specific command. So, we define a set
of commands that have framed response. It currently only
contains "batch."

As a result of this change, the one-off implementation of
_submitbatch() in sshv1peer can be removed since it is now
safe to .read() the response's file object until end of stream.
cappedreader takes care of not overrunning the frame.

Diff Detail

Repository
rHG Mercurial
Lint
Lint Skipped
Unit
Unit Tests Skipped

Event Timeline

indygreg created this revision.Feb 21 2018, 5:08 PM
yuja accepted this revision.Feb 22 2018, 9:05 AM
This revision is now accepted and ready to land.Feb 22 2018, 9:05 AM
Closed by commit rHG043e77f3be09: sshpeer: return framed file object when needed (authored by indygreg). · Explain WhyFeb 22 2018, 9:18 AM
This revision was automatically updated to reflect the committed changes.

Revision Contents
Changeset List

PathPackages
Mmercurial/sshpeer.py (42 lines)

Diff 5966

mercurial/sshpeer.py

# self._subprocess is unused. Keeping a handle on the process # self._subprocess is unused. Keeping a handle on the process
# holds a reference and prevents it from being garbage collected. # holds a reference and prevents it from being garbage collected.
self._subprocess = proc self._subprocess = proc
self._pipeo = stdin self._pipeo = stdin
self._pipei = stdout self._pipei = stdout
self._pipee = stderr self._pipee = stderr
self._caps = caps self._caps = caps
# Commands that have a "framed" response where the first line of the
# response contains the length of that response.
_FRAMED_COMMANDS = {
'batch',
}
# Begin of _basepeer interface. # Begin of _basepeer interface.
@util.propertycache @util.propertycache
def ui(self): def ui(self):
return self._ui return self._ui
def url(self): def url(self):
return self._url return self._url
self._cleanup() self._cleanup()
raise exception raise exception
def _cleanup(self): def _cleanup(self):
_cleanuppipes(self.ui, self._pipei, self._pipeo, self._pipee) _cleanuppipes(self.ui, self._pipei, self._pipeo, self._pipee)
__del__ = _cleanup __del__ = _cleanup
def _submitbatch(self, req): def _sendrequest(self, cmd, args, framed=False):
rsp = self._callstream("batch", cmds=wireproto.encodebatchcmds(req))
available = self._getamount()
# TODO this response parsing is probably suboptimal for large
# batches with large responses.
toread = min(available, 1024)
work = rsp.read(toread)
available -= toread
chunk = work
while chunk:
while ';' in work:
one, work = work.split(';', 1)
yield wireproto.unescapearg(one)
toread = min(available, 1024)
chunk = rsp.read(toread)
available -= toread
work += chunk
yield wireproto.unescapearg(work)
def _sendrequest(self, cmd, args):
if (self.ui.debugflag if (self.ui.debugflag
and self.ui.configbool('devel', 'debug.peer-request')): and self.ui.configbool('devel', 'debug.peer-request')):
dbg = self.ui.debug dbg = self.ui.debug
line = 'devel-peer-request: %s\n' line = 'devel-peer-request: %s\n'
dbg(line % cmd) dbg(line % cmd)
for key, value in sorted(args.items()): for key, value in sorted(args.items()):
if not isinstance(value, dict): if not isinstance(value, dict):
dbg(line % ' %s: %d bytes' % (key, len(value))) dbg(line % ' %s: %d bytes' % (key, len(value)))
if isinstance(v, dict): if isinstance(v, dict):
for dk, dv in v.iteritems(): for dk, dv in v.iteritems():
self._pipeo.write("%s %d\n" % (dk, len(dv))) self._pipeo.write("%s %d\n" % (dk, len(dv)))
self._pipeo.write(dv) self._pipeo.write(dv)
else: else:
self._pipeo.write(v) self._pipeo.write(v)
self._pipeo.flush() self._pipeo.flush()
# We know exactly how many bytes are in the response. So return a proxy
# around the raw output stream that allows reading exactly this many
# bytes. Callers then can read() without fear of overrunning the
# response.
if framed:
amount = self._getamount()
return util.cappedreader(self._pipei, amount)
return self._pipei return self._pipei
def _callstream(self, cmd, **args): def _callstream(self, cmd, **args):
args = pycompat.byteskwargs(args) args = pycompat.byteskwargs(args)
return self._sendrequest(cmd, args) return self._sendrequest(cmd, args, framed=cmd in self._FRAMED_COMMANDS)
def _callcompressable(self, cmd, **args): def _callcompressable(self, cmd, **args):
args = pycompat.byteskwargs(args) args = pycompat.byteskwargs(args)
return self._sendrequest(cmd, args) return self._sendrequest(cmd, args, framed=cmd in self._FRAMED_COMMANDS)
def _call(self, cmd, **args): def _call(self, cmd, **args):
args = pycompat.byteskwargs(args) args = pycompat.byteskwargs(args)
self._sendrequest(cmd, args) return self._sendrequest(cmd, args, framed=True).read()
return self._readframed()
def _callpush(self, cmd, fp, **args): def _callpush(self, cmd, fp, **args):
r = self._call(cmd, **args) r = self._call(cmd, **args)
if r: if r:
return '', r return '', r
for d in iter(lambda: fp.read(4096), ''): for d in iter(lambda: fp.read(4096), ''):
self._writeframed(d) self._writeframed(d)
self._writeframed("", flush=True) self._writeframed("", flush=True)