wireprotoserver: convert ErrorResponse to bytes

Caught by pytype:

     File "/mnt/c/Users/Matt/hg/mercurial/wireprotoserver.py", line 236, in handlewsgirequest: Function _bytestr.__init__ was called with the wrong arguments [wrong-arg-types]
              Expected: (self, ints: Iterable[int])
       Actually passed: (self, ints: mercurial.hgweb.common.ErrorResponse)
       The following methods aren't implemented on mercurial.hgweb.common.ErrorResponse:
       __iter__
     File "/mnt/c/Users/Matt/hg/mercurial/wireprotoserver.py", line 239, in handlewsgirequest: Function _bytestr.__init__ was called with the wrong arguments [wrong-arg-types]
              Expected: (self, ints: Iterable[int])
       Actually passed: (self, ints: mercurial.hgweb.common.ErrorResponse)
       The following methods aren't implemented on mercurial.hgweb.common.ErrorResponse:
       __iter__
     File "/mnt/c/Users/Matt/hg/mercurial/wireprotov2server.py", line 91, in handlehttpv2request: Function _bytestr.__init__ was called with the wrong arguments [wrong-arg-types]
              Expected: (self, ints: Iterable[int])
       Actually passed: (self, ints: mercurial.hgweb.common.ErrorResponse)
       The following methods aren't implemented on mercurial.hgweb.common.ErrorResponse:
       __iter__

Differential Revision: https://phab.mercurial-scm.org/D10182

mercurial/wireprotoserver.py

@@ -1,856 +1,859 b''
 # Copyright 21 May 2005 - (c) 2005 Jake Edge <[e-mail]>
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import contextlib
 import struct
 import threading
 
 from .i18n import _
 from . import (
     encoding,
     error,
     pycompat,
     util,
     wireprototypes,
     wireprotov1server,
     wireprotov2server,
 )
 from .interfaces import util as interfaceutil
 from .utils import (
     cborutil,
     compression,
+    stringutil,
 )
 
 stringio = util.stringio
 
 urlerr = util.urlerr
 urlreq = util.urlreq
 
 HTTP_OK = 200
 
 HGTYPE = b'application/mercurial-0.1'
 HGTYPE2 = b'application/mercurial-0.2'
 HGERRTYPE = b'application/hg-error'
 
 SSHV1 = wireprototypes.SSHV1
 SSHV2 = wireprototypes.SSHV2
 
 
 def decodevaluefromheaders(req, headerprefix):
     """Decode a long value from multiple HTTP request headers.
 
     Returns the value as a bytes, not a str.
     """
     chunks = []
     i = 1
     while True:
         v = req.headers.get(b'%s-%d' % (headerprefix, i))
         if v is None:
             break
         chunks.append(pycompat.bytesurl(v))
         i += 1
 
     return b''.join(chunks)
 
 
 @interfaceutil.implementer(wireprototypes.baseprotocolhandler)
 class httpv1protocolhandler(object):
     def __init__(self, req, ui, checkperm):
         self._req = req
         self._ui = ui
         self._checkperm = checkperm
         self._protocaps = None
 
     @property
     def name(self):
         return b'http-v1'
 
     def getargs(self, args):
         knownargs = self._args()
         data = {}
         keys = args.split()
         for k in keys:
             if k == b'*':
                 star = {}
                 for key in knownargs.keys():
                     if key != b'cmd' and key not in keys:
                         star[key] = knownargs[key][0]
                 data[b'*'] = star
             else:
                 data[k] = knownargs[k][0]
         return [data[k] for k in keys]
 
     def _args(self):
         args = self._req.qsparams.asdictoflists()
         postlen = int(self._req.headers.get(b'X-HgArgs-Post', 0))
         if postlen:
             args.update(
                 urlreq.parseqs(
                     self._req.bodyfh.read(postlen), keep_blank_values=True
                 )
             )
             return args
 
         argvalue = decodevaluefromheaders(self._req, b'X-HgArg')
         args.update(urlreq.parseqs(argvalue, keep_blank_values=True))
         return args
 
     def getprotocaps(self):
         if self._protocaps is None:
             value = decodevaluefromheaders(self._req, b'X-HgProto')
             self._protocaps = set(value.split(b' '))
         return self._protocaps
 
     def getpayload(self):
         # Existing clients *always* send Content-Length.
         length = int(self._req.headers[b'Content-Length'])
 
         # If httppostargs is used, we need to read Content-Length
         # minus the amount that was consumed by args.
         length -= int(self._req.headers.get(b'X-HgArgs-Post', 0))
         return util.filechunkiter(self._req.bodyfh, limit=length)
 
     @contextlib.contextmanager
     def mayberedirectstdio(self):
         oldout = self._ui.fout
         olderr = self._ui.ferr
 
         out = util.stringio()
 
         try:
             self._ui.fout = out
             self._ui.ferr = out
             yield out
         finally:
             self._ui.fout = oldout
             self._ui.ferr = olderr
 
     def client(self):
         return b'remote:%s:%s:%s' % (
             self._req.urlscheme,
             urlreq.quote(self._req.remotehost or b''),
             urlreq.quote(self._req.remoteuser or b''),
         )
 
     def addcapabilities(self, repo, caps):
         caps.append(b'batch')
 
         caps.append(
             b'httpheader=%d' % repo.ui.configint(b'server', b'maxhttpheaderlen')
         )
         if repo.ui.configbool(b'experimental', b'httppostargs'):
             caps.append(b'httppostargs')
 
         # FUTURE advertise 0.2rx once support is implemented
         # FUTURE advertise minrx and mintx after consulting config option
         caps.append(b'httpmediatype=0.1rx,0.1tx,0.2tx')
 
         compengines = wireprototypes.supportedcompengines(
             repo.ui, compression.SERVERROLE
         )
         if compengines:
             comptypes = b','.join(
                 urlreq.quote(e.wireprotosupport().name) for e in compengines
             )
             caps.append(b'compression=%s' % comptypes)
 
         return caps
 
     def checkperm(self, perm):
         return self._checkperm(perm)
 
 
 # This method exists mostly so that extensions like remotefilelog can
 # disable a kludgey legacy method only over http. As of early 2018,
 # there are no other known users, so with any luck we can discard this
 # hook if remotefilelog becomes a first-party extension.
 def iscmd(cmd):
     return cmd in wireprotov1server.commands
 
 
 def handlewsgirequest(rctx, req, res, checkperm):
     """Possibly process a wire protocol request.
 
     If the current request is a wire protocol request, the request is
     processed by this function.
 
     ``req`` is a ``parsedrequest`` instance.
     ``res`` is a ``wsgiresponse`` instance.
 
     Returns a bool indicating if the request was serviced. If set, the caller
     should stop processing the request, as a response has already been issued.
     """
     # Avoid cycle involving hg module.
     from .hgweb import common as hgwebcommon
 
     repo = rctx.repo
 
     # HTTP version 1 wire protocol requests are denoted by a "cmd" query
     # string parameter. If it isn't present, this isn't a wire protocol
     # request.
     if b'cmd' not in req.qsparams:
         return False
 
     cmd = req.qsparams[b'cmd']
 
     # The "cmd" request parameter is used by both the wire protocol and hgweb.
     # While not all wire protocol commands are available for all transports,
     # if we see a "cmd" value that resembles a known wire protocol command, we
     # route it to a protocol handler. This is better than routing possible
     # wire protocol requests to hgweb because it prevents hgweb from using
     # known wire protocol commands and it is less confusing for machine
     # clients.
     if not iscmd(cmd):
         return False
 
     # The "cmd" query string argument is only valid on the root path of the
     # repo. e.g. ``/?cmd=foo``, ``/repo?cmd=foo``. URL paths within the repo
     # like ``/blah?cmd=foo`` are not allowed. So don't recognize the request
     # in this case. We send an HTTP 404 for backwards compatibility reasons.
     if req.dispatchpath:
         res.status = hgwebcommon.statusmessage(404)
         res.headers[b'Content-Type'] = HGTYPE
         # TODO This is not a good response to issue for this request. This
         # is mostly for BC for now.
         res.setbodybytes(b'0\n%s\n' % b'Not Found')
         return True
 
     proto = httpv1protocolhandler(
         req, repo.ui, lambda perm: checkperm(rctx, req, perm)
     )
 
     # The permissions checker should be the only thing that can raise an
     # ErrorResponse. It is kind of a layer violation to catch an hgweb
     # exception here. So consider refactoring into a exception type that
     # is associated with the wire protocol.
     try:
         _callhttp(repo, req, res, proto, cmd)
     except hgwebcommon.ErrorResponse as e:
         for k, v in e.headers:
             res.headers[k] = v
-        res.status = hgwebcommon.statusmessage(e.code, pycompat.bytestr(e))
+        res.status = hgwebcommon.statusmessage(
+            e.code, stringutil.forcebytestr(e)
+        )
         # TODO This response body assumes the failed command was
         # "unbundle." That assumption is not always valid.
-        res.setbodybytes(b'0\n%s\n' % pycompat.bytestr(e))
+        res.setbodybytes(b'0\n%s\n' % stringutil.forcebytestr(e))
 
     return True
 
 
 def _availableapis(repo):
     apis = set()
 
     # Registered APIs are made available via config options of the name of
     # the protocol.
     for k, v in API_HANDLERS.items():
         section, option = v[b'config']
         if repo.ui.configbool(section, option):
             apis.add(k)
 
     return apis
 
 
 def handlewsgiapirequest(rctx, req, res, checkperm):
     """Handle requests to /api/*."""
     assert req.dispatchparts[0] == b'api'
 
     repo = rctx.repo
 
     # This whole URL space is experimental for now. But we want to
     # reserve the URL space. So, 404 all URLs if the feature isn't enabled.
     if not repo.ui.configbool(b'experimental', b'web.apiserver'):
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'Experimental API server endpoint not enabled'))
         return
 
     # The URL space is /api/<protocol>/*. The structure of URLs under varies
     # by <protocol>.
 
     availableapis = _availableapis(repo)
 
     # Requests to /api/ list available APIs.
     if req.dispatchparts == [b'api']:
         res.status = b'200 OK'
         res.headers[b'Content-Type'] = b'text/plain'
         lines = [
             _(
                 b'APIs can be accessed at /api/<name>, where <name> can be '
                 b'one of the following:\n'
             )
         ]
         if availableapis:
             lines.extend(sorted(availableapis))
         else:
             lines.append(_(b'(no available APIs)\n'))
         res.setbodybytes(b'\n'.join(lines))
         return
 
     proto = req.dispatchparts[1]
 
     if proto not in API_HANDLERS:
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(
             _(b'Unknown API: %s\nKnown APIs: %s')
             % (proto, b', '.join(sorted(availableapis)))
         )
         return
 
     if proto not in availableapis:
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'API %s not enabled\n') % proto)
         return
 
     API_HANDLERS[proto][b'handler'](
         rctx, req, res, checkperm, req.dispatchparts[2:]
     )
 
 
 # Maps API name to metadata so custom API can be registered.
 # Keys are:
 #
 # config
 #    Config option that controls whether service is enabled.
 # handler
 #    Callable receiving (rctx, req, res, checkperm, urlparts) that is called
 #    when a request to this API is received.
 # apidescriptor
 #    Callable receiving (req, repo) that is called to obtain an API
 #    descriptor for this service. The response must be serializable to CBOR.
 API_HANDLERS = {
     wireprotov2server.HTTP_WIREPROTO_V2: {
         b'config': (b'experimental', b'web.api.http-v2'),
         b'handler': wireprotov2server.handlehttpv2request,
         b'apidescriptor': wireprotov2server.httpv2apidescriptor,
     },
 }
 
 
 def _httpresponsetype(ui, proto, prefer_uncompressed):
     """Determine the appropriate response type and compression settings.
 
     Returns a tuple of (mediatype, compengine, engineopts).
     """
     # Determine the response media type and compression engine based
     # on the request parameters.
 
     if b'0.2' in proto.getprotocaps():
         # All clients are expected to support uncompressed data.
         if prefer_uncompressed:
             return HGTYPE2, compression._noopengine(), {}
 
         # Now find an agreed upon compression format.
         compformats = wireprotov1server.clientcompressionsupport(proto)
         for engine in wireprototypes.supportedcompengines(
             ui, compression.SERVERROLE
         ):
             if engine.wireprotosupport().name in compformats:
                 opts = {}
                 level = ui.configint(b'server', b'%slevel' % engine.name())
                 if level is not None:
                     opts[b'level'] = level
 
                 return HGTYPE2, engine, opts
 
         # No mutually supported compression format. Fall back to the
         # legacy protocol.
 
     # Don't allow untrusted settings because disabling compression or
     # setting a very high compression level could lead to flooding
     # the server's network or CPU.
     opts = {b'level': ui.configint(b'server', b'zliblevel')}
     return HGTYPE, util.compengines[b'zlib'], opts
 
 
 def processcapabilitieshandshake(repo, req, res, proto):
     """Called during a ?cmd=capabilities request.
 
     If the client is advertising support for a newer protocol, we send
     a CBOR response with information about available services. If no
     advertised services are available, we don't handle the request.
     """
     # Fall back to old behavior unless the API server is enabled.
     if not repo.ui.configbool(b'experimental', b'web.apiserver'):
         return False
 
     clientapis = decodevaluefromheaders(req, b'X-HgUpgrade')
     protocaps = decodevaluefromheaders(req, b'X-HgProto')
     if not clientapis or not protocaps:
         return False
 
     # We currently only support CBOR responses.
     protocaps = set(protocaps.split(b' '))
     if b'cbor' not in protocaps:
         return False
 
     descriptors = {}
 
     for api in sorted(set(clientapis.split()) & _availableapis(repo)):
         handler = API_HANDLERS[api]
 
         descriptorfn = handler.get(b'apidescriptor')
         if not descriptorfn:
             continue
 
         descriptors[api] = descriptorfn(req, repo)
 
     v1caps = wireprotov1server.dispatch(repo, proto, b'capabilities')
     assert isinstance(v1caps, wireprototypes.bytesresponse)
 
     m = {
         # TODO allow this to be configurable.
         b'apibase': b'api/',
         b'apis': descriptors,
         b'v1capabilities': v1caps.data,
     }
 
     res.status = b'200 OK'
     res.headers[b'Content-Type'] = b'application/mercurial-cbor'
     res.setbodybytes(b''.join(cborutil.streamencode(m)))
 
     return True
 
 
 def _callhttp(repo, req, res, proto, cmd):
     # Avoid cycle involving hg module.
     from .hgweb import common as hgwebcommon
 
     def genversion2(gen, engine, engineopts):
         # application/mercurial-0.2 always sends a payload header
         # identifying the compression engine.
         name = engine.wireprotosupport().name
         assert 0 < len(name) < 256
         yield struct.pack(b'B', len(name))
         yield name
 
         for chunk in gen:
             yield chunk
 
     def setresponse(code, contenttype, bodybytes=None, bodygen=None):
         if code == HTTP_OK:
             res.status = b'200 Script output follows'
         else:
             res.status = hgwebcommon.statusmessage(code)
 
         res.headers[b'Content-Type'] = contenttype
 
         if bodybytes is not None:
             res.setbodybytes(bodybytes)
         if bodygen is not None:
             res.setbodygen(bodygen)
 
     if not wireprotov1server.commands.commandavailable(cmd, proto):
         setresponse(
             HTTP_OK,
             HGERRTYPE,
             _(
                 b'requested wire protocol command is not available over '
                 b'HTTP'
             ),
         )
         return
 
     proto.checkperm(wireprotov1server.commands[cmd].permission)
 
     # Possibly handle a modern client wanting to switch protocols.
     if cmd == b'capabilities' and processcapabilitieshandshake(
         repo, req, res, proto
     ):
 
         return
 
     rsp = wireprotov1server.dispatch(repo, proto, cmd)
 
     if isinstance(rsp, bytes):
         setresponse(HTTP_OK, HGTYPE, bodybytes=rsp)
     elif isinstance(rsp, wireprototypes.bytesresponse):
         setresponse(HTTP_OK, HGTYPE, bodybytes=rsp.data)
     elif isinstance(rsp, wireprototypes.streamreslegacy):
         setresponse(HTTP_OK, HGTYPE, bodygen=rsp.gen)
     elif isinstance(rsp, wireprototypes.streamres):
         gen = rsp.gen
 
         # This code for compression should not be streamres specific. It
         # is here because we only compress streamres at the moment.
         mediatype, engine, engineopts = _httpresponsetype(
             repo.ui, proto, rsp.prefer_uncompressed
         )
         gen = engine.compressstream(gen, engineopts)
 
         if mediatype == HGTYPE2:
             gen = genversion2(gen, engine, engineopts)
 
         setresponse(HTTP_OK, mediatype, bodygen=gen)
     elif isinstance(rsp, wireprototypes.pushres):
         rsp = b'%d\n%s' % (rsp.res, rsp.output)
         setresponse(HTTP_OK, HGTYPE, bodybytes=rsp)
     elif isinstance(rsp, wireprototypes.pusherr):
         rsp = b'0\n%s\n' % rsp.res
         res.drain = True
         setresponse(HTTP_OK, HGTYPE, bodybytes=rsp)
     elif isinstance(rsp, wireprototypes.ooberror):
         setresponse(HTTP_OK, HGERRTYPE, bodybytes=rsp.message)
     else:
         raise error.ProgrammingError(b'hgweb.protocol internal failure', rsp)
 
 
 def _sshv1respondbytes(fout, value):
     """Send a bytes response for protocol version 1."""
     fout.write(b'%d\n' % len(value))
     fout.write(value)
     fout.flush()
 
 
 def _sshv1respondstream(fout, source):
     write = fout.write
     for chunk in source.gen:
         write(chunk)
     fout.flush()
 
 
 def _sshv1respondooberror(fout, ferr, rsp):
     ferr.write(b'%s\n-\n' % rsp)
     ferr.flush()
     fout.write(b'\n')
     fout.flush()
 
 
 @interfaceutil.implementer(wireprototypes.baseprotocolhandler)
 class sshv1protocolhandler(object):
     """Handler for requests services via version 1 of SSH protocol."""
 
     def __init__(self, ui, fin, fout):
         self._ui = ui
         self._fin = fin
         self._fout = fout
         self._protocaps = set()
 
     @property
     def name(self):
         return wireprototypes.SSHV1
 
     def getargs(self, args):
         data = {}
         keys = args.split()
         for n in pycompat.xrange(len(keys)):
             argline = self._fin.readline()[:-1]
             arg, l = argline.split()
             if arg not in keys:
                 raise error.Abort(_(b"unexpected parameter %r") % arg)
             if arg == b'*':
                 star = {}
                 for k in pycompat.xrange(int(l)):
                     argline = self._fin.readline()[:-1]
                     arg, l = argline.split()
                     val = self._fin.read(int(l))
                     star[arg] = val
                 data[b'*'] = star
             else:
                 val = self._fin.read(int(l))
                 data[arg] = val
         return [data[k] for k in keys]
 
     def getprotocaps(self):
         return self._protocaps
 
     def getpayload(self):
         # We initially send an empty response. This tells the client it is
         # OK to start sending data. If a client sees any other response, it
         # interprets it as an error.
         _sshv1respondbytes(self._fout, b'')
 
         # The file is in the form:
         #
         # <chunk size>\n<chunk>
         # ...
         # 0\n
         count = int(self._fin.readline())
         while count:
             yield self._fin.read(count)
             count = int(self._fin.readline())
 
     @contextlib.contextmanager
     def mayberedirectstdio(self):
         yield None
 
     def client(self):
         client = encoding.environ.get(b'SSH_CLIENT', b'').split(b' ', 1)[0]
         return b'remote:ssh:' + client
 
     def addcapabilities(self, repo, caps):
         if self.name == wireprototypes.SSHV1:
             caps.append(b'protocaps')
         caps.append(b'batch')
         return caps
 
     def checkperm(self, perm):
         pass
 
 
 class sshv2protocolhandler(sshv1protocolhandler):
     """Protocol handler for version 2 of the SSH protocol."""
 
     @property
     def name(self):
         return wireprototypes.SSHV2
 
     def addcapabilities(self, repo, caps):
         return caps
 
 
 def _runsshserver(ui, repo, fin, fout, ev):
     # This function operates like a state machine of sorts. The following
     # states are defined:
     #
     # protov1-serving
     #    Server is in protocol version 1 serving mode. Commands arrive on
     #    new lines. These commands are processed in this state, one command
     #    after the other.
     #
     # protov2-serving
     #    Server is in protocol version 2 serving mode.
     #
     # upgrade-initial
     #    The server is going to process an upgrade request.
     #
     # upgrade-v2-filter-legacy-handshake
     #    The protocol is being upgraded to version 2. The server is expecting
     #    the legacy handshake from version 1.
     #
     # upgrade-v2-finish
     #    The upgrade to version 2 of the protocol is imminent.
     #
     # shutdown
     #    The server is shutting down, possibly in reaction to a client event.
     #
     # And here are their transitions:
     #
     # protov1-serving -> shutdown
     #    When server receives an empty request or encounters another
     #    error.
     #
     # protov1-serving -> upgrade-initial
     #    An upgrade request line was seen.
     #
     # upgrade-initial -> upgrade-v2-filter-legacy-handshake
     #    Upgrade to version 2 in progress. Server is expecting to
     #    process a legacy handshake.
     #
     # upgrade-v2-filter-legacy-handshake -> shutdown
     #    Client did not fulfill upgrade handshake requirements.
     #
     # upgrade-v2-filter-legacy-handshake -> upgrade-v2-finish
     #    Client fulfilled version 2 upgrade requirements. Finishing that
     #    upgrade.
     #
     # upgrade-v2-finish -> protov2-serving
     #    Protocol upgrade to version 2 complete. Server can now speak protocol
     #    version 2.
     #
     # protov2-serving -> protov1-serving
     #    Ths happens by default since protocol version 2 is the same as
     #    version 1 except for the handshake.
 
     state = b'protov1-serving'
     proto = sshv1protocolhandler(ui, fin, fout)
     protoswitched = False
 
     while not ev.is_set():
         if state == b'protov1-serving':
             # Commands are issued on new lines.
             request = fin.readline()[:-1]
 
             # Empty lines signal to terminate the connection.
             if not request:
                 state = b'shutdown'
                 continue
 
             # It looks like a protocol upgrade request. Transition state to
             # handle it.
             if request.startswith(b'upgrade '):
                 if protoswitched:
                     _sshv1respondooberror(
                         fout,
                         ui.ferr,
                         b'cannot upgrade protocols multiple times',
                     )
                     state = b'shutdown'
                     continue
 
                 state = b'upgrade-initial'
                 continue
 
             available = wireprotov1server.commands.commandavailable(
                 request, proto
             )
 
             # This command isn't available. Send an empty response and go
             # back to waiting for a new command.
             if not available:
                 _sshv1respondbytes(fout, b'')
                 continue
 
             rsp = wireprotov1server.dispatch(repo, proto, request)
             repo.ui.fout.flush()
             repo.ui.ferr.flush()
 
             if isinstance(rsp, bytes):
                 _sshv1respondbytes(fout, rsp)
             elif isinstance(rsp, wireprototypes.bytesresponse):
                 _sshv1respondbytes(fout, rsp.data)
             elif isinstance(rsp, wireprototypes.streamres):
                 _sshv1respondstream(fout, rsp)
             elif isinstance(rsp, wireprototypes.streamreslegacy):
                 _sshv1respondstream(fout, rsp)
             elif isinstance(rsp, wireprototypes.pushres):
                 _sshv1respondbytes(fout, b'')
                 _sshv1respondbytes(fout, b'%d' % rsp.res)
             elif isinstance(rsp, wireprototypes.pusherr):
                 _sshv1respondbytes(fout, rsp.res)
             elif isinstance(rsp, wireprototypes.ooberror):
                 _sshv1respondooberror(fout, ui.ferr, rsp.message)
             else:
                 raise error.ProgrammingError(
                     b'unhandled response type from '
                     b'wire protocol command: %s' % rsp
                 )
 
         # For now, protocol version 2 serving just goes back to version 1.
         elif state == b'protov2-serving':
             state = b'protov1-serving'
             continue
 
         elif state == b'upgrade-initial':
             # We should never transition into this state if we've switched
             # protocols.
             assert not protoswitched
             assert proto.name == wireprototypes.SSHV1
 
             # Expected: upgrade <token> <capabilities>
             # If we get something else, the request is malformed. It could be
             # from a future client that has altered the upgrade line content.
             # We treat this as an unknown command.
             try:
                 token, caps = request.split(b' ')[1:]
             except ValueError:
                 _sshv1respondbytes(fout, b'')
                 state = b'protov1-serving'
                 continue
 
             # Send empty response if we don't support upgrading protocols.
             if not ui.configbool(b'experimental', b'sshserver.support-v2'):
                 _sshv1respondbytes(fout, b'')
                 state = b'protov1-serving'
                 continue
 
             try:
                 caps = urlreq.parseqs(caps)
             except ValueError:
                 _sshv1respondbytes(fout, b'')
                 state = b'protov1-serving'
                 continue
 
             # We don't see an upgrade request to protocol version 2. Ignore
             # the upgrade request.
             wantedprotos = caps.get(b'proto', [b''])[0]
             if SSHV2 not in wantedprotos:
                 _sshv1respondbytes(fout, b'')
                 state = b'protov1-serving'
                 continue
 
             # It looks like we can honor this upgrade request to protocol 2.
             # Filter the rest of the handshake protocol request lines.
             state = b'upgrade-v2-filter-legacy-handshake'
             continue
 
         elif state == b'upgrade-v2-filter-legacy-handshake':
             # Client should have sent legacy handshake after an ``upgrade``
             # request. Expected lines:
             #
             #    hello
             #    between
             #    pairs 81
             #    0000...-0000...
 
             ok = True
             for line in (b'hello', b'between', b'pairs 81'):
                 request = fin.readline()[:-1]
 
                 if request != line:
                     _sshv1respondooberror(
                         fout,
                         ui.ferr,
                         b'malformed handshake protocol: missing %s' % line,
                     )
                     ok = False
                     state = b'shutdown'
                     break
 
             if not ok:
                 continue
 
             request = fin.read(81)
             if request != b'%s-%s' % (b'0' * 40, b'0' * 40):
                 _sshv1respondooberror(
                     fout,
                     ui.ferr,
                     b'malformed handshake protocol: '
                     b'missing between argument value',
                 )
                 state = b'shutdown'
                 continue
 
             state = b'upgrade-v2-finish'
             continue
 
         elif state == b'upgrade-v2-finish':
             # Send the upgrade response.
             fout.write(b'upgraded %s %s\n' % (token, SSHV2))
             servercaps = wireprotov1server.capabilities(repo, proto)
             rsp = b'capabilities: %s' % servercaps.data
             fout.write(b'%d\n%s\n' % (len(rsp), rsp))
             fout.flush()
 
             proto = sshv2protocolhandler(ui, fin, fout)
             protoswitched = True
 
             state = b'protov2-serving'
             continue
 
         elif state == b'shutdown':
             break
 
         else:
             raise error.ProgrammingError(
                 b'unhandled ssh server state: %s' % state
             )
 
 
 class sshserver(object):
     def __init__(self, ui, repo, logfh=None):
         self._ui = ui
         self._repo = repo
         self._fin, self._fout = ui.protectfinout()
 
         # Log write I/O to stdout and stderr if configured.
         if logfh:
             self._fout = util.makeloggingfileobject(
                 logfh, self._fout, b'o', logdata=True
             )
             ui.ferr = util.makeloggingfileobject(
                 logfh, ui.ferr, b'e', logdata=True
             )
 
     def serve_forever(self):
         self.serveuntil(threading.Event())
         self._ui.restorefinout(self._fin, self._fout)
 
     def serveuntil(self, ev):
         """Serve until a threading.Event is set."""
         _runsshserver(self._ui, self._repo, self._fin, self._fout, ev)

mercurial/wireprotov2server.py

@@ -1,1613 +1,1615 b''
 # Copyright 21 May 2005 - (c) 2005 Jake Edge <[e-mail]>
 # Copyright 2005-2007 Matt Mackall <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2 or any later version.
 
 from __future__ import absolute_import
 
 import collections
 import contextlib
 
 from .i18n import _
 from .node import (
     hex,
     nullid,
 )
 from . import (
     discovery,
     encoding,
     error,
     match as matchmod,
     narrowspec,
     pycompat,
     streamclone,
     templatefilters,
     util,
     wireprotoframing,
     wireprototypes,
 )
 from .interfaces import util as interfaceutil
 from .utils import (
     cborutil,
     hashutil,
     stringutil,
 )
 
 FRAMINGTYPE = b'application/mercurial-exp-framing-0006'
 
 HTTP_WIREPROTO_V2 = wireprototypes.HTTP_WIREPROTO_V2
 
 COMMANDS = wireprototypes.commanddict()
 
 # Value inserted into cache key computation function. Change the value to
 # force new cache keys for every command request. This should be done when
 # there is a change to how caching works, etc.
 GLOBAL_CACHE_VERSION = 1
 
 
 def handlehttpv2request(rctx, req, res, checkperm, urlparts):
     from .hgweb import common as hgwebcommon
 
     # URL space looks like: <permissions>/<command>, where <permission> can
     # be ``ro`` or ``rw`` to signal read-only or read-write, respectively.
 
     # Root URL does nothing meaningful... yet.
     if not urlparts:
         res.status = b'200 OK'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'HTTP version 2 API handler'))
         return
 
     if len(urlparts) == 1:
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(
             _(b'do not know how to process %s\n') % req.dispatchpath
         )
         return
 
     permission, command = urlparts[0:2]
 
     if permission not in (b'ro', b'rw'):
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'unknown permission: %s') % permission)
         return
 
     if req.method != b'POST':
         res.status = b'405 Method Not Allowed'
         res.headers[b'Allow'] = b'POST'
         res.setbodybytes(_(b'commands require POST requests'))
         return
 
     # At some point we'll want to use our own API instead of recycling the
     # behavior of version 1 of the wire protocol...
     # TODO return reasonable responses - not responses that overload the
     # HTTP status line message for error reporting.
     try:
         checkperm(rctx, req, b'pull' if permission == b'ro' else b'push')
     except hgwebcommon.ErrorResponse as e:
-        res.status = hgwebcommon.statusmessage(e.code, pycompat.bytestr(e))
+        res.status = hgwebcommon.statusmessage(
+            e.code, stringutil.forcebytestr(e)
+        )
         for k, v in e.headers:
             res.headers[k] = v
         res.setbodybytes(b'permission denied')
         return
 
     # We have a special endpoint to reflect the request back at the client.
     if command == b'debugreflect':
         _processhttpv2reflectrequest(rctx.repo.ui, rctx.repo, req, res)
         return
 
     # Extra commands that we handle that aren't really wire protocol
     # commands. Think extra hard before making this hackery available to
     # extension.
     extracommands = {b'multirequest'}
 
     if command not in COMMANDS and command not in extracommands:
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'unknown wire protocol command: %s\n') % command)
         return
 
     repo = rctx.repo
     ui = repo.ui
 
     proto = httpv2protocolhandler(req, ui)
 
     if (
         not COMMANDS.commandavailable(command, proto)
         and command not in extracommands
     ):
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'invalid wire protocol command: %s') % command)
         return
 
     # TODO consider cases where proxies may add additional Accept headers.
     if req.headers.get(b'Accept') != FRAMINGTYPE:
         res.status = b'406 Not Acceptable'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(
             _(b'client MUST specify Accept header with value: %s\n')
             % FRAMINGTYPE
         )
         return
 
     if req.headers.get(b'Content-Type') != FRAMINGTYPE:
         res.status = b'415 Unsupported Media Type'
         # TODO we should send a response with appropriate media type,
         # since client does Accept it.
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(
             _(b'client MUST send Content-Type header with value: %s\n')
             % FRAMINGTYPE
         )
         return
 
     _processhttpv2request(ui, repo, req, res, permission, command, proto)
 
 
 def _processhttpv2reflectrequest(ui, repo, req, res):
     """Reads unified frame protocol request and dumps out state to client.
 
     This special endpoint can be used to help debug the wire protocol.
 
     Instead of routing the request through the normal dispatch mechanism,
     we instead read all frames, decode them, and feed them into our state
     tracker. We then dump the log of all that activity back out to the
     client.
     """
     # Reflection APIs have a history of being abused, accidentally disclosing
     # sensitive data, etc. So we have a config knob.
     if not ui.configbool(b'experimental', b'web.api.debugreflect'):
         res.status = b'404 Not Found'
         res.headers[b'Content-Type'] = b'text/plain'
         res.setbodybytes(_(b'debugreflect service not available'))
         return
 
     # We assume we have a unified framing protocol request body.
 
     reactor = wireprotoframing.serverreactor(ui)
     states = []
 
     while True:
         frame = wireprotoframing.readframe(req.bodyfh)
 
         if not frame:
             states.append(b'received: <no frame>')
             break
 
         states.append(
             b'received: %d %d %d %s'
             % (frame.typeid, frame.flags, frame.requestid, frame.payload)
         )
 
         action, meta = reactor.onframerecv(frame)
         states.append(templatefilters.json((action, meta)))
 
     action, meta = reactor.oninputeof()
     meta[b'action'] = action
     states.append(templatefilters.json(meta))
 
     res.status = b'200 OK'
     res.headers[b'Content-Type'] = b'text/plain'
     res.setbodybytes(b'\n'.join(states))
 
 
 def _processhttpv2request(ui, repo, req, res, authedperm, reqcommand, proto):
     """Post-validation handler for HTTPv2 requests.
 
     Called when the HTTP request contains unified frame-based protocol
     frames for evaluation.
     """
     # TODO Some HTTP clients are full duplex and can receive data before
     # the entire request is transmitted. Figure out a way to indicate support
     # for that so we can opt into full duplex mode.
     reactor = wireprotoframing.serverreactor(ui, deferoutput=True)
     seencommand = False
 
     outstream = None
 
     while True:
         frame = wireprotoframing.readframe(req.bodyfh)
         if not frame:
             break
 
         action, meta = reactor.onframerecv(frame)
 
         if action == b'wantframe':
             # Need more data before we can do anything.
             continue
         elif action == b'runcommand':
             # Defer creating output stream because we need to wait for
             # protocol settings frames so proper encoding can be applied.
             if not outstream:
                 outstream = reactor.makeoutputstream()
 
             sentoutput = _httpv2runcommand(
                 ui,
                 repo,
                 req,
                 res,
                 authedperm,
                 reqcommand,
                 reactor,
                 outstream,
                 meta,
                 issubsequent=seencommand,
             )
 
             if sentoutput:
                 return
 
             seencommand = True
 
         elif action == b'error':
             # TODO define proper error mechanism.
             res.status = b'200 OK'
             res.headers[b'Content-Type'] = b'text/plain'
             res.setbodybytes(meta[b'message'] + b'\n')
             return
         else:
             raise error.ProgrammingError(
                 b'unhandled action from frame processor: %s' % action
             )
 
     action, meta = reactor.oninputeof()
     if action == b'sendframes':
         # We assume we haven't started sending the response yet. If we're
         # wrong, the response type will raise an exception.
         res.status = b'200 OK'
         res.headers[b'Content-Type'] = FRAMINGTYPE
         res.setbodygen(meta[b'framegen'])
     elif action == b'noop':
         pass
     else:
         raise error.ProgrammingError(
             b'unhandled action from frame processor: %s' % action
         )
 
 
 def _httpv2runcommand(
     ui,
     repo,
     req,
     res,
     authedperm,
     reqcommand,
     reactor,
     outstream,
     command,
     issubsequent,
 ):
     """Dispatch a wire protocol command made from HTTPv2 requests.
 
     The authenticated permission (``authedperm``) along with the original
     command from the URL (``reqcommand``) are passed in.
     """
     # We already validated that the session has permissions to perform the
     # actions in ``authedperm``. In the unified frame protocol, the canonical
     # command to run is expressed in a frame. However, the URL also requested
     # to run a specific command. We need to be careful that the command we
     # run doesn't have permissions requirements greater than what was granted
     # by ``authedperm``.
     #
     # Our rule for this is we only allow one command per HTTP request and
     # that command must match the command in the URL. However, we make
     # an exception for the ``multirequest`` URL. This URL is allowed to
     # execute multiple commands. We double check permissions of each command
     # as it is invoked to ensure there is no privilege escalation.
     # TODO consider allowing multiple commands to regular command URLs
     # iff each command is the same.
 
     proto = httpv2protocolhandler(req, ui, args=command[b'args'])
 
     if reqcommand == b'multirequest':
         if not COMMANDS.commandavailable(command[b'command'], proto):
             # TODO proper error mechanism
             res.status = b'200 OK'
             res.headers[b'Content-Type'] = b'text/plain'
             res.setbodybytes(
                 _(b'wire protocol command not available: %s')
                 % command[b'command']
             )
             return True
 
         # TODO don't use assert here, since it may be elided by -O.
         assert authedperm in (b'ro', b'rw')
         wirecommand = COMMANDS[command[b'command']]
         assert wirecommand.permission in (b'push', b'pull')
 
         if authedperm == b'ro' and wirecommand.permission != b'pull':
             # TODO proper error mechanism
             res.status = b'403 Forbidden'
             res.headers[b'Content-Type'] = b'text/plain'
             res.setbodybytes(
                 _(b'insufficient permissions to execute command: %s')
                 % command[b'command']
             )
             return True
 
         # TODO should we also call checkperm() here? Maybe not if we're going
         # to overhaul that API. The granted scope from the URL check should
         # be good enough.
 
     else:
         # Don't allow multiple commands outside of ``multirequest`` URL.
         if issubsequent:
             # TODO proper error mechanism
             res.status = b'200 OK'
             res.headers[b'Content-Type'] = b'text/plain'
             res.setbodybytes(
                 _(b'multiple commands cannot be issued to this URL')
             )
             return True
 
         if reqcommand != command[b'command']:
             # TODO define proper error mechanism
             res.status = b'200 OK'
             res.headers[b'Content-Type'] = b'text/plain'
             res.setbodybytes(_(b'command in frame must match command in URL'))
             return True
 
     res.status = b'200 OK'
     res.headers[b'Content-Type'] = FRAMINGTYPE
 
     try:
         objs = dispatch(repo, proto, command[b'command'], command[b'redirect'])
 
         action, meta = reactor.oncommandresponsereadyobjects(
             outstream, command[b'requestid'], objs
         )
 
     except error.WireprotoCommandError as e:
         action, meta = reactor.oncommanderror(
             outstream, command[b'requestid'], e.message, e.messageargs
         )
 
     except Exception as e:
         action, meta = reactor.onservererror(
             outstream,
             command[b'requestid'],
             _(b'exception when invoking command: %s')
             % stringutil.forcebytestr(e),
         )
 
     if action == b'sendframes':
         res.setbodygen(meta[b'framegen'])
         return True
     elif action == b'noop':
         return False
     else:
         raise error.ProgrammingError(
             b'unhandled event from reactor: %s' % action
         )
 
 
 def getdispatchrepo(repo, proto, command):
     viewconfig = repo.ui.config(b'server', b'view')
     return repo.filtered(viewconfig)
 
 
 def dispatch(repo, proto, command, redirect):
     """Run a wire protocol command.
 
     Returns an iterable of objects that will be sent to the client.
     """
     repo = getdispatchrepo(repo, proto, command)
 
     entry = COMMANDS[command]
     func = entry.func
     spec = entry.args
 
     args = proto.getargs(spec)
 
     # There is some duplicate boilerplate code here for calling the command and
     # emitting objects. It is either that or a lot of indented code that looks
     # like a pyramid (since there are a lot of code paths that result in not
     # using the cacher).
     callcommand = lambda: func(repo, proto, **pycompat.strkwargs(args))
 
     # Request is not cacheable. Don't bother instantiating a cacher.
     if not entry.cachekeyfn:
         for o in callcommand():
             yield o
         return
 
     if redirect:
         redirecttargets = redirect[b'targets']
         redirecthashes = redirect[b'hashes']
     else:
         redirecttargets = []
         redirecthashes = []
 
     cacher = makeresponsecacher(
         repo,
         proto,
         command,
         args,
         cborutil.streamencode,
         redirecttargets=redirecttargets,
         redirecthashes=redirecthashes,
     )
 
     # But we have no cacher. Do default handling.
     if not cacher:
         for o in callcommand():
             yield o
         return
 
     with cacher:
         cachekey = entry.cachekeyfn(
             repo, proto, cacher, **pycompat.strkwargs(args)
         )
 
         # No cache key or the cacher doesn't like it. Do default handling.
         if cachekey is None or not cacher.setcachekey(cachekey):
             for o in callcommand():
                 yield o
             return
 
         # Serve it from the cache, if possible.
         cached = cacher.lookup()
 
         if cached:
             for o in cached[b'objs']:
                 yield o
             return
 
         # Else call the command and feed its output into the cacher, allowing
         # the cacher to buffer/mutate objects as it desires.
         for o in callcommand():
             for o in cacher.onobject(o):
                 yield o
 
         for o in cacher.onfinished():
             yield o
 
 
 @interfaceutil.implementer(wireprototypes.baseprotocolhandler)
 class httpv2protocolhandler(object):
     def __init__(self, req, ui, args=None):
         self._req = req
         self._ui = ui
         self._args = args
 
     @property
     def name(self):
         return HTTP_WIREPROTO_V2
 
     def getargs(self, args):
         # First look for args that were passed but aren't registered on this
         # command.
         extra = set(self._args) - set(args)
         if extra:
             raise error.WireprotoCommandError(
                 b'unsupported argument to command: %s'
                 % b', '.join(sorted(extra))
             )
 
         # And look for required arguments that are missing.
         missing = {a for a in args if args[a][b'required']} - set(self._args)
 
         if missing:
             raise error.WireprotoCommandError(
                 b'missing required arguments: %s' % b', '.join(sorted(missing))
             )
 
         # Now derive the arguments to pass to the command, taking into
         # account the arguments specified by the client.
         data = {}
         for k, meta in sorted(args.items()):
             # This argument wasn't passed by the client.
             if k not in self._args:
                 data[k] = meta[b'default']()
                 continue
 
             v = self._args[k]
 
             # Sets may be expressed as lists. Silently normalize.
             if meta[b'type'] == b'set' and isinstance(v, list):
                 v = set(v)
 
             # TODO consider more/stronger type validation.
 
             data[k] = v
 
         return data
 
     def getprotocaps(self):
         # Protocol capabilities are currently not implemented for HTTP V2.
         return set()
 
     def getpayload(self):
         raise NotImplementedError
 
     @contextlib.contextmanager
     def mayberedirectstdio(self):
         raise NotImplementedError
 
     def client(self):
         raise NotImplementedError
 
     def addcapabilities(self, repo, caps):
         return caps
 
     def checkperm(self, perm):
         raise NotImplementedError
 
 
 def httpv2apidescriptor(req, repo):
     proto = httpv2protocolhandler(req, repo.ui)
 
     return _capabilitiesv2(repo, proto)
 
 
 def _capabilitiesv2(repo, proto):
     """Obtain the set of capabilities for version 2 transports.
 
     These capabilities are distinct from the capabilities for version 1
     transports.
     """
     caps = {
         b'commands': {},
         b'framingmediatypes': [FRAMINGTYPE],
         b'pathfilterprefixes': set(narrowspec.VALID_PREFIXES),
     }
 
     for command, entry in COMMANDS.items():
         args = {}
 
         for arg, meta in entry.args.items():
             args[arg] = {
                 # TODO should this be a normalized type using CBOR's
                 # terminology?
                 b'type': meta[b'type'],
                 b'required': meta[b'required'],
             }
 
             if not meta[b'required']:
                 args[arg][b'default'] = meta[b'default']()
 
             if meta[b'validvalues']:
                 args[arg][b'validvalues'] = meta[b'validvalues']
 
         # TODO this type of check should be defined in a per-command callback.
         if (
             command == b'rawstorefiledata'
             and not streamclone.allowservergeneration(repo)
         ):
             continue
 
         caps[b'commands'][command] = {
             b'args': args,
             b'permissions': [entry.permission],
         }
 
         if entry.extracapabilitiesfn:
             extracaps = entry.extracapabilitiesfn(repo, proto)
             caps[b'commands'][command].update(extracaps)
 
     caps[b'rawrepoformats'] = sorted(repo.requirements & repo.supportedformats)
 
     targets = getadvertisedredirecttargets(repo, proto)
     if targets:
         caps[b'redirect'] = {
             b'targets': [],
             b'hashes': [b'sha256', b'sha1'],
         }
 
         for target in targets:
             entry = {
                 b'name': target[b'name'],
                 b'protocol': target[b'protocol'],
                 b'uris': target[b'uris'],
             }
 
             for key in (b'snirequired', b'tlsversions'):
                 if key in target:
                     entry[key] = target[key]
 
             caps[b'redirect'][b'targets'].append(entry)
 
     return proto.addcapabilities(repo, caps)
 
 
 def getadvertisedredirecttargets(repo, proto):
     """Obtain a list of content redirect targets.
 
     Returns a list containing potential redirect targets that will be
     advertised in capabilities data. Each dict MUST have the following
     keys:
 
     name
        The name of this redirect target. This is the identifier clients use
        to refer to a target. It is transferred as part of every command
        request.
 
     protocol
        Network protocol used by this target. Typically this is the string
        in front of the ``://`` in a URL. e.g. ``https``.
 
     uris
        List of representative URIs for this target. Clients can use the
        URIs to test parsing for compatibility or for ordering preference
        for which target to use.
 
     The following optional keys are recognized:
 
     snirequired
        Bool indicating if Server Name Indication (SNI) is required to
        connect to this target.
 
     tlsversions
        List of bytes indicating which TLS versions are supported by this
        target.
 
     By default, clients reflect the target order advertised by servers
     and servers will use the first client-advertised target when picking
     a redirect target. So targets should be advertised in the order the
     server prefers they be used.
     """
     return []
 
 
 def wireprotocommand(
     name,
     args=None,
     permission=b'push',
     cachekeyfn=None,
     extracapabilitiesfn=None,
 ):
     """Decorator to declare a wire protocol command.
 
     ``name`` is the name of the wire protocol command being provided.
 
     ``args`` is a dict defining arguments accepted by the command. Keys are
     the argument name. Values are dicts with the following keys:
 
        ``type``
           The argument data type. Must be one of the following string
           literals: ``bytes``, ``int``, ``list``, ``dict``, ``set``,
           or ``bool``.
 
        ``default``
           A callable returning the default value for this argument. If not
           specified, ``None`` will be the default value.
 
        ``example``
           An example value for this argument.
 
        ``validvalues``
           Set of recognized values for this argument.
 
     ``permission`` defines the permission type needed to run this command.
     Can be ``push`` or ``pull``. These roughly map to read-write and read-only,
     respectively. Default is to assume command requires ``push`` permissions
     because otherwise commands not declaring their permissions could modify
     a repository that is supposed to be read-only.
 
     ``cachekeyfn`` defines an optional callable that can derive the
     cache key for this request.
 
     ``extracapabilitiesfn`` defines an optional callable that defines extra
     command capabilities/parameters that are advertised next to the command
     in the capabilities data structure describing the server. The callable
     receives as arguments the repository and protocol objects. It returns
     a dict of extra fields to add to the command descriptor.
 
     Wire protocol commands are generators of objects to be serialized and
     sent to the client.
 
     If a command raises an uncaught exception, this will be translated into
     a command error.
 
     All commands can opt in to being cacheable by defining a function
     (``cachekeyfn``) that is called to derive a cache key. This function
     receives the same arguments as the command itself plus a ``cacher``
     argument containing the active cacher for the request and returns a bytes
     containing the key in a cache the response to this command may be cached
     under.
     """
     transports = {
         k for k, v in wireprototypes.TRANSPORTS.items() if v[b'version'] == 2
     }
 
     if permission not in (b'push', b'pull'):
         raise error.ProgrammingError(
             b'invalid wire protocol permission; '
             b'got %s; expected "push" or "pull"' % permission
         )
 
     if args is None:
         args = {}
 
     if not isinstance(args, dict):
         raise error.ProgrammingError(
             b'arguments for version 2 commands must be declared as dicts'
         )
 
     for arg, meta in args.items():
         if arg == b'*':
             raise error.ProgrammingError(
                 b'* argument name not allowed on version 2 commands'
             )
 
         if not isinstance(meta, dict):
             raise error.ProgrammingError(
                 b'arguments for version 2 commands '
                 b'must declare metadata as a dict'
             )
 
         if b'type' not in meta:
             raise error.ProgrammingError(
                 b'%s argument for command %s does not '
                 b'declare type field' % (arg, name)
             )
 
         if meta[b'type'] not in (
             b'bytes',
             b'int',
             b'list',
             b'dict',
             b'set',
             b'bool',
         ):
             raise error.ProgrammingError(
                 b'%s argument for command %s has '
                 b'illegal type: %s' % (arg, name, meta[b'type'])
             )
 
         if b'example' not in meta:
             raise error.ProgrammingError(
                 b'%s argument for command %s does not '
                 b'declare example field' % (arg, name)
             )
 
         meta[b'required'] = b'default' not in meta
 
         meta.setdefault(b'default', lambda: None)
         meta.setdefault(b'validvalues', None)
 
     def register(func):
         if name in COMMANDS:
             raise error.ProgrammingError(
                 b'%s command already registered for version 2' % name
             )
 
         COMMANDS[name] = wireprototypes.commandentry(
             func,
             args=args,
             transports=transports,
             permission=permission,
             cachekeyfn=cachekeyfn,
             extracapabilitiesfn=extracapabilitiesfn,
         )
 
         return func
 
     return register
 
 
 def makecommandcachekeyfn(command, localversion=None, allargs=False):
     """Construct a cache key derivation function with common features.
 
     By default, the cache key is a hash of:
 
     * The command name.
     * A global cache version number.
     * A local cache version number (passed via ``localversion``).
     * All the arguments passed to the command.
     * The media type used.
     * Wire protocol version string.
     * The repository path.
     """
     if not allargs:
         raise error.ProgrammingError(
             b'only allargs=True is currently supported'
         )
 
     if localversion is None:
         raise error.ProgrammingError(b'must set localversion argument value')
 
     def cachekeyfn(repo, proto, cacher, **args):
         spec = COMMANDS[command]
 
         # Commands that mutate the repo can not be cached.
         if spec.permission == b'push':
             return None
 
         # TODO config option to disable caching.
 
         # Our key derivation strategy is to construct a data structure
         # holding everything that could influence cacheability and to hash
         # the CBOR representation of that. Using CBOR seems like it might
         # be overkill. However, simpler hashing mechanisms are prone to
         # duplicate input issues. e.g. if you just concatenate two values,
         # "foo"+"bar" is identical to "fo"+"obar". Using CBOR provides
         # "padding" between values and prevents these problems.
 
         # Seed the hash with various data.
         state = {
             # To invalidate all cache keys.
             b'globalversion': GLOBAL_CACHE_VERSION,
             # More granular cache key invalidation.
             b'localversion': localversion,
             # Cache keys are segmented by command.
             b'command': command,
             # Throw in the media type and API version strings so changes
             # to exchange semantics invalid cache.
             b'mediatype': FRAMINGTYPE,
             b'version': HTTP_WIREPROTO_V2,
             # So same requests for different repos don't share cache keys.
             b'repo': repo.root,
         }
 
         # The arguments passed to us will have already been normalized.
         # Default values will be set, etc. This is important because it
         # means that it doesn't matter if clients send an explicit argument
         # or rely on the default value: it will all normalize to the same
         # set of arguments on the server and therefore the same cache key.
         #
         # Arguments by their very nature must support being encoded to CBOR.
         # And the CBOR encoder is deterministic. So we hash the arguments
         # by feeding the CBOR of their representation into the hasher.
         if allargs:
             state[b'args'] = pycompat.byteskwargs(args)
 
         cacher.adjustcachekeystate(state)
 
         hasher = hashutil.sha1()
         for chunk in cborutil.streamencode(state):
             hasher.update(chunk)
 
         return pycompat.sysbytes(hasher.hexdigest())
 
     return cachekeyfn
 
 
 def makeresponsecacher(
     repo, proto, command, args, objencoderfn, redirecttargets, redirecthashes
 ):
     """Construct a cacher for a cacheable command.
 
     Returns an ``iwireprotocolcommandcacher`` instance.
 
     Extensions can monkeypatch this function to provide custom caching
     backends.
     """
     return None
 
 
 def resolvenodes(repo, revisions):
     """Resolve nodes from a revisions specifier data structure."""
     cl = repo.changelog
     clhasnode = cl.hasnode
 
     seen = set()
     nodes = []
 
     if not isinstance(revisions, list):
         raise error.WireprotoCommandError(
             b'revisions must be defined as an array'
         )
 
     for spec in revisions:
         if b'type' not in spec:
             raise error.WireprotoCommandError(
                 b'type key not present in revision specifier'
             )
 
         typ = spec[b'type']
 
         if typ == b'changesetexplicit':
             if b'nodes' not in spec:
                 raise error.WireprotoCommandError(
                     b'nodes key not present in changesetexplicit revision '
                     b'specifier'
                 )
 
             for node in spec[b'nodes']:
                 if node not in seen:
                     nodes.append(node)
                     seen.add(node)
 
         elif typ == b'changesetexplicitdepth':
             for key in (b'nodes', b'depth'):
                 if key not in spec:
                     raise error.WireprotoCommandError(
                         b'%s key not present in changesetexplicitdepth revision '
                         b'specifier',
                         (key,),
                     )
 
             for rev in repo.revs(
                 b'ancestors(%ln, %s)', spec[b'nodes'], spec[b'depth'] - 1
             ):
                 node = cl.node(rev)
 
                 if node not in seen:
                     nodes.append(node)
                     seen.add(node)
 
         elif typ == b'changesetdagrange':
             for key in (b'roots', b'heads'):
                 if key not in spec:
                     raise error.WireprotoCommandError(
                         b'%s key not present in changesetdagrange revision '
                         b'specifier',
                         (key,),
                     )
 
             if not spec[b'heads']:
                 raise error.WireprotoCommandError(
                     b'heads key in changesetdagrange cannot be empty'
                 )
 
             if spec[b'roots']:
                 common = [n for n in spec[b'roots'] if clhasnode(n)]
             else:
                 common = [nullid]
 
             for n in discovery.outgoing(repo, common, spec[b'heads']).missing:
                 if n not in seen:
                     nodes.append(n)
                     seen.add(n)
 
         else:
             raise error.WireprotoCommandError(
                 b'unknown revision specifier type: %s', (typ,)
             )
 
     return nodes
 
 
 @wireprotocommand(b'branchmap', permission=b'pull')
 def branchmapv2(repo, proto):
     yield {
         encoding.fromlocal(k): v
         for k, v in pycompat.iteritems(repo.branchmap())
     }
 
 
 @wireprotocommand(b'capabilities', permission=b'pull')
 def capabilitiesv2(repo, proto):
     yield _capabilitiesv2(repo, proto)
 
 
 @wireprotocommand(
     b'changesetdata',
     args={
         b'revisions': {
             b'type': b'list',
             b'example': [
                 {
                     b'type': b'changesetexplicit',
                     b'nodes': [b'abcdef...'],
                 }
             ],
         },
         b'fields': {
             b'type': b'set',
             b'default': set,
             b'example': {b'parents', b'revision'},
             b'validvalues': {b'bookmarks', b'parents', b'phase', b'revision'},
         },
     },
     permission=b'pull',
 )
 def changesetdata(repo, proto, revisions, fields):
     # TODO look for unknown fields and abort when they can't be serviced.
     # This could probably be validated by dispatcher using validvalues.
 
     cl = repo.changelog
     outgoing = resolvenodes(repo, revisions)
     publishing = repo.publishing()
 
     if outgoing:
         repo.hook(b'preoutgoing', throw=True, source=b'serve')
 
     yield {
         b'totalitems': len(outgoing),
     }
 
     # The phases of nodes already transferred to the client may have changed
     # since the client last requested data. We send phase-only records
     # for these revisions, if requested.
     # TODO actually do this. We'll probably want to emit phase heads
     # in the ancestry set of the outgoing revisions. This will ensure
     # that phase updates within that set are seen.
     if b'phase' in fields:
         pass
 
     nodebookmarks = {}
     for mark, node in repo._bookmarks.items():
         nodebookmarks.setdefault(node, set()).add(mark)
 
     # It is already topologically sorted by revision number.
     for node in outgoing:
         d = {
             b'node': node,
         }
 
         if b'parents' in fields:
             d[b'parents'] = cl.parents(node)
 
         if b'phase' in fields:
             if publishing:
                 d[b'phase'] = b'public'
             else:
                 ctx = repo[node]
                 d[b'phase'] = ctx.phasestr()
 
         if b'bookmarks' in fields and node in nodebookmarks:
             d[b'bookmarks'] = sorted(nodebookmarks[node])
             del nodebookmarks[node]
 
         followingmeta = []
         followingdata = []
 
         if b'revision' in fields:
             revisiondata = cl.revision(node)
             followingmeta.append((b'revision', len(revisiondata)))
             followingdata.append(revisiondata)
 
         # TODO make it possible for extensions to wrap a function or register
         # a handler to service custom fields.
 
         if followingmeta:
             d[b'fieldsfollowing'] = followingmeta
 
         yield d
 
         for extra in followingdata:
             yield extra
 
     # If requested, send bookmarks from nodes that didn't have revision
     # data sent so receiver is aware of any bookmark updates.
     if b'bookmarks' in fields:
         for node, marks in sorted(pycompat.iteritems(nodebookmarks)):
             yield {
                 b'node': node,
                 b'bookmarks': sorted(marks),
             }
 
 
 class FileAccessError(Exception):
     """Represents an error accessing a specific file."""
 
     def __init__(self, path, msg, args):
         self.path = path
         self.msg = msg
         self.args = args
 
 
 def getfilestore(repo, proto, path):
     """Obtain a file storage object for use with wire protocol.
 
     Exists as a standalone function so extensions can monkeypatch to add
     access control.
     """
     # This seems to work even if the file doesn't exist. So catch
     # "empty" files and return an error.
     fl = repo.file(path)
 
     if not len(fl):
         raise FileAccessError(path, b'unknown file: %s', (path,))
 
     return fl
 
 
 def emitfilerevisions(repo, path, revisions, linknodes, fields):
     for revision in revisions:
         d = {
             b'node': revision.node,
         }
 
         if b'parents' in fields:
             d[b'parents'] = [revision.p1node, revision.p2node]
 
         if b'linknode' in fields:
             d[b'linknode'] = linknodes[revision.node]
 
         followingmeta = []
         followingdata = []
 
         if b'revision' in fields:
             if revision.revision is not None:
                 followingmeta.append((b'revision', len(revision.revision)))
                 followingdata.append(revision.revision)
             else:
                 d[b'deltabasenode'] = revision.basenode
                 followingmeta.append((b'delta', len(revision.delta)))
                 followingdata.append(revision.delta)
 
         if followingmeta:
             d[b'fieldsfollowing'] = followingmeta
 
         yield d
 
         for extra in followingdata:
             yield extra
 
 
 def makefilematcher(repo, pathfilter):
     """Construct a matcher from a path filter dict."""
 
     # Validate values.
     if pathfilter:
         for key in (b'include', b'exclude'):
             for pattern in pathfilter.get(key, []):
                 if not pattern.startswith((b'path:', b'rootfilesin:')):
                     raise error.WireprotoCommandError(
                         b'%s pattern must begin with `path:` or `rootfilesin:`; '
                         b'got %s',
                         (key, pattern),
                     )
 
     if pathfilter:
         matcher = matchmod.match(
             repo.root,
             b'',
             include=pathfilter.get(b'include', []),
             exclude=pathfilter.get(b'exclude', []),
         )
     else:
         matcher = matchmod.match(repo.root, b'')
 
     # Requested patterns could include files not in the local store. So
     # filter those out.
     return repo.narrowmatch(matcher)
 
 
 @wireprotocommand(
     b'filedata',
     args={
         b'haveparents': {
             b'type': b'bool',
             b'default': lambda: False,
             b'example': True,
         },
         b'nodes': {
             b'type': b'list',
             b'example': [b'0123456...'],
         },
         b'fields': {
             b'type': b'set',
             b'default': set,
             b'example': {b'parents', b'revision'},
             b'validvalues': {b'parents', b'revision', b'linknode'},
         },
         b'path': {
             b'type': b'bytes',
             b'example': b'foo.txt',
         },
     },
     permission=b'pull',
     # TODO censoring a file revision won't invalidate the cache.
     # Figure out a way to take censoring into account when deriving
     # the cache key.
     cachekeyfn=makecommandcachekeyfn(b'filedata', 1, allargs=True),
 )
 def filedata(repo, proto, haveparents, nodes, fields, path):
     # TODO this API allows access to file revisions that are attached to
     # secret changesets. filesdata does not have this problem. Maybe this
     # API should be deleted?
 
     try:
         # Extensions may wish to access the protocol handler.
         store = getfilestore(repo, proto, path)
     except FileAccessError as e:
         raise error.WireprotoCommandError(e.msg, e.args)
 
     clnode = repo.changelog.node
     linknodes = {}
 
     # Validate requested nodes.
     for node in nodes:
         try:
             store.rev(node)
         except error.LookupError:
             raise error.WireprotoCommandError(
                 b'unknown file node: %s', (hex(node),)
             )
 
         # TODO by creating the filectx against a specific file revision
         # instead of changeset, linkrev() is always used. This is wrong for
         # cases where linkrev() may refer to a hidden changeset. But since this
         # API doesn't know anything about changesets, we're not sure how to
         # disambiguate the linknode. Perhaps we should delete this API?
         fctx = repo.filectx(path, fileid=node)
         linknodes[node] = clnode(fctx.introrev())
 
     revisions = store.emitrevisions(
         nodes,
         revisiondata=b'revision' in fields,
         assumehaveparentrevisions=haveparents,
     )
 
     yield {
         b'totalitems': len(nodes),
     }
 
     for o in emitfilerevisions(repo, path, revisions, linknodes, fields):
         yield o
 
 
 def filesdatacapabilities(repo, proto):
     batchsize = repo.ui.configint(
         b'experimental', b'server.filesdata.recommended-batch-size'
     )
     return {
         b'recommendedbatchsize': batchsize,
     }
 
 
 @wireprotocommand(
     b'filesdata',
     args={
         b'haveparents': {
             b'type': b'bool',
             b'default': lambda: False,
             b'example': True,
         },
         b'fields': {
             b'type': b'set',
             b'default': set,
             b'example': {b'parents', b'revision'},
             b'validvalues': {
                 b'firstchangeset',
                 b'linknode',
                 b'parents',
                 b'revision',
             },
         },
         b'pathfilter': {
             b'type': b'dict',
             b'default': lambda: None,
             b'example': {b'include': [b'path:tests']},
         },
         b'revisions': {
             b'type': b'list',
             b'example': [
                 {
                     b'type': b'changesetexplicit',
                     b'nodes': [b'abcdef...'],
                 }
             ],
         },
     },
     permission=b'pull',
     # TODO censoring a file revision won't invalidate the cache.
     # Figure out a way to take censoring into account when deriving
     # the cache key.
     cachekeyfn=makecommandcachekeyfn(b'filesdata', 1, allargs=True),
     extracapabilitiesfn=filesdatacapabilities,
 )
 def filesdata(repo, proto, haveparents, fields, pathfilter, revisions):
     # TODO This should operate on a repo that exposes obsolete changesets. There
     # is a race between a client making a push that obsoletes a changeset and
     # another client fetching files data for that changeset. If a client has a
     # changeset, it should probably be allowed to access files data for that
     # changeset.
 
     outgoing = resolvenodes(repo, revisions)
     filematcher = makefilematcher(repo, pathfilter)
 
     # path -> {fnode: linknode}
     fnodes = collections.defaultdict(dict)
 
     # We collect the set of relevant file revisions by iterating the changeset
     # revisions and either walking the set of files recorded in the changeset
     # or by walking the manifest at that revision. There is probably room for a
     # storage-level API to request this data, as it can be expensive to compute
     # and would benefit from caching or alternate storage from what revlogs
     # provide.
     for node in outgoing:
         ctx = repo[node]
         mctx = ctx.manifestctx()
         md = mctx.read()
 
         if haveparents:
             checkpaths = ctx.files()
         else:
             checkpaths = md.keys()
 
         for path in checkpaths:
             fnode = md[path]
 
             if path in fnodes and fnode in fnodes[path]:
                 continue
 
             if not filematcher(path):
                 continue
 
             fnodes[path].setdefault(fnode, node)
 
     yield {
         b'totalpaths': len(fnodes),
         b'totalitems': sum(len(v) for v in fnodes.values()),
     }
 
     for path, filenodes in sorted(fnodes.items()):
         try:
             store = getfilestore(repo, proto, path)
         except FileAccessError as e:
             raise error.WireprotoCommandError(e.msg, e.args)
 
         yield {
             b'path': path,
             b'totalitems': len(filenodes),
         }
 
         revisions = store.emitrevisions(
             filenodes.keys(),
             revisiondata=b'revision' in fields,
             assumehaveparentrevisions=haveparents,
         )
 
         for o in emitfilerevisions(repo, path, revisions, filenodes, fields):
             yield o
 
 
 @wireprotocommand(
     b'heads',
     args={
         b'publiconly': {
             b'type': b'bool',
             b'default': lambda: False,
             b'example': False,
         },
     },
     permission=b'pull',
 )
 def headsv2(repo, proto, publiconly):
     if publiconly:
         repo = repo.filtered(b'immutable')
 
     yield repo.heads()
 
 
 @wireprotocommand(
     b'known',
     args={
         b'nodes': {
             b'type': b'list',
             b'default': list,
             b'example': [b'deadbeef'],
         },
     },
     permission=b'pull',
 )
 def knownv2(repo, proto, nodes):
     result = b''.join(b'1' if n else b'0' for n in repo.known(nodes))
     yield result
 
 
 @wireprotocommand(
     b'listkeys',
     args={
         b'namespace': {
             b'type': b'bytes',
             b'example': b'ns',
         },
     },
     permission=b'pull',
 )
 def listkeysv2(repo, proto, namespace):
     keys = repo.listkeys(encoding.tolocal(namespace))
     keys = {
         encoding.fromlocal(k): encoding.fromlocal(v)
         for k, v in pycompat.iteritems(keys)
     }
 
     yield keys
 
 
 @wireprotocommand(
     b'lookup',
     args={
         b'key': {
             b'type': b'bytes',
             b'example': b'foo',
         },
     },
     permission=b'pull',
 )
 def lookupv2(repo, proto, key):
     key = encoding.tolocal(key)
 
     # TODO handle exception.
     node = repo.lookup(key)
 
     yield node
 
 
 def manifestdatacapabilities(repo, proto):
     batchsize = repo.ui.configint(
         b'experimental', b'server.manifestdata.recommended-batch-size'
     )
 
     return {
         b'recommendedbatchsize': batchsize,
     }
 
 
 @wireprotocommand(
     b'manifestdata',
     args={
         b'nodes': {
             b'type': b'list',
             b'example': [b'0123456...'],
         },
         b'haveparents': {
             b'type': b'bool',
             b'default': lambda: False,
             b'example': True,
         },
         b'fields': {
             b'type': b'set',
             b'default': set,
             b'example': {b'parents', b'revision'},
             b'validvalues': {b'parents', b'revision'},
         },
         b'tree': {
             b'type': b'bytes',
             b'example': b'',
         },
     },
     permission=b'pull',
     cachekeyfn=makecommandcachekeyfn(b'manifestdata', 1, allargs=True),
     extracapabilitiesfn=manifestdatacapabilities,
 )
 def manifestdata(repo, proto, haveparents, nodes, fields, tree):
     store = repo.manifestlog.getstorage(tree)
 
     # Validate the node is known and abort on unknown revisions.
     for node in nodes:
         try:
             store.rev(node)
         except error.LookupError:
             raise error.WireprotoCommandError(b'unknown node: %s', (node,))
 
     revisions = store.emitrevisions(
         nodes,
         revisiondata=b'revision' in fields,
         assumehaveparentrevisions=haveparents,
     )
 
     yield {
         b'totalitems': len(nodes),
     }
 
     for revision in revisions:
         d = {
             b'node': revision.node,
         }
 
         if b'parents' in fields:
             d[b'parents'] = [revision.p1node, revision.p2node]
 
         followingmeta = []
         followingdata = []
 
         if b'revision' in fields:
             if revision.revision is not None:
                 followingmeta.append((b'revision', len(revision.revision)))
                 followingdata.append(revision.revision)
             else:
                 d[b'deltabasenode'] = revision.basenode
                 followingmeta.append((b'delta', len(revision.delta)))
                 followingdata.append(revision.delta)
 
         if followingmeta:
             d[b'fieldsfollowing'] = followingmeta
 
         yield d
 
         for extra in followingdata:
             yield extra
 
 
 @wireprotocommand(
     b'pushkey',
     args={
         b'namespace': {
             b'type': b'bytes',
             b'example': b'ns',
         },
         b'key': {
             b'type': b'bytes',
             b'example': b'key',
         },
         b'old': {
             b'type': b'bytes',
             b'example': b'old',
         },
         b'new': {
             b'type': b'bytes',
             b'example': b'new',
         },
     },
     permission=b'push',
 )
 def pushkeyv2(repo, proto, namespace, key, old, new):
     # TODO handle ui output redirection
     yield repo.pushkey(
         encoding.tolocal(namespace),
         encoding.tolocal(key),
         encoding.tolocal(old),
         encoding.tolocal(new),
     )
 
 
 @wireprotocommand(
     b'rawstorefiledata',
     args={
         b'files': {
             b'type': b'list',
             b'example': [b'changelog', b'manifestlog'],
         },
         b'pathfilter': {
             b'type': b'list',
             b'default': lambda: None,
             b'example': {b'include': [b'path:tests']},
         },
     },
     permission=b'pull',
 )
 def rawstorefiledata(repo, proto, files, pathfilter):
     if not streamclone.allowservergeneration(repo):
         raise error.WireprotoCommandError(b'stream clone is disabled')
 
     # TODO support dynamically advertising what store files "sets" are
     # available. For now, we support changelog, manifestlog, and files.
     files = set(files)
     allowedfiles = {b'changelog', b'manifestlog'}
 
     unsupported = files - allowedfiles
     if unsupported:
         raise error.WireprotoCommandError(
             b'unknown file type: %s', (b', '.join(sorted(unsupported)),)
         )
 
     with repo.lock():
         topfiles = list(repo.store.topfiles())
 
     sendfiles = []
     totalsize = 0
 
     # TODO this is a bunch of storage layer interface abstractions because
     # it assumes revlogs.
     for name, encodedname, size in topfiles:
         if b'changelog' in files and name.startswith(b'00changelog'):
             pass
         elif b'manifestlog' in files and name.startswith(b'00manifest'):
             pass
         else:
             continue
 
         sendfiles.append((b'store', name, size))
         totalsize += size
 
     yield {
         b'filecount': len(sendfiles),
         b'totalsize': totalsize,
     }
 
     for location, name, size in sendfiles:
         yield {
             b'location': location,
             b'path': name,
             b'size': size,
         }
 
         # We have to use a closure for this to ensure the context manager is
         # closed only after sending the final chunk.
         def getfiledata():
             with repo.svfs(name, b'rb', auditpath=False) as fh:
                 for chunk in util.filechunkiter(fh, limit=size):
                     yield chunk
 
         yield wireprototypes.indefinitebytestringresponse(getfiledata())