annotate mercurial/httppeer.py @ 41978:a56487081109

store: don't read the whole fncache in memory

In large repositories with lot of files, the fncache grows more than 100 MB and
reading that whole thing into memory slows things down. Let's not read the whole
thing into memory.

This patch changes fncache loading code to read 1 MB at once. Loading 1 MB at
once saves ~1 sec on perffncacheload for our internal repository. I tried
various values such as 0.5 MB, 5 MB, 10 MB but best results were produced using
1 MB as the chunksize.

On a narrow clone with fncache around 40 MB, this patch saves ~0.04 seconds on
average on perffncacheload.

To test the code, I have coded an extension in test-fncache.t which set
chunksize to 1 byte, and the test passes with that.

Differential Revision: https://phab.mercurial-scm.org/D5296

# httppeer.py - HTTP repository proxy classes for mercurial
#
# Copyright 2005, 2006 Matt Mackall <[e-mail]>
# Copyright 2006 Vadim Gelfer <[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 errno
import io
import os
import socket
import struct
import weakref

from .i18n import _
from . import (
    bundle2,
    error,
    httpconnection,
    pycompat,
    repository,
    statichttprepo,
    url as urlmod,
    util,
    wireprotoframing,
    wireprototypes,
    wireprotov1peer,
    wireprotov2peer,
    wireprotov2server,
)
from .utils import (
    cborutil,
    interfaceutil,
    stringutil,
)

httplib = util.httplib
urlerr = util.urlerr
urlreq = util.urlreq

def encodevalueinheaders(value, header, limit):
    """Encode a string value into multiple HTTP headers.

    ``value`` will be encoded into 1 or more HTTP headers with the names
    ``header-<N>`` where ``<N>`` is an integer starting at 1. Each header
    name + value will be at most ``limit`` bytes long.

    Returns an iterable of 2-tuples consisting of header names and
    values as native strings.
    """
    # HTTP Headers are ASCII. Python 3 requires them to be unicodes,
    # not bytes. This function always takes bytes in as arguments.
    fmt = pycompat.strurl(header) + r'-%s'
    # Note: it is *NOT* a bug that the last bit here is a bytestring
    # and not a unicode: we're just getting the encoded length anyway,
    # and using an r-string to make it portable between Python 2 and 3
    # doesn't work because then the \r is a literal backslash-r
    # instead of a carriage return.
    valuelen = limit - len(fmt % r'000') - len(': \r\n')
    result = []

    n = 0
    for i in pycompat.xrange(0, len(value), valuelen):
        n += 1
        result.append((fmt % str(n), pycompat.strurl(value[i:i + valuelen])))

    return result

class _multifile(object):
    def __init__(self, *fileobjs):
        for f in fileobjs:
            if not util.safehasattr(f, 'length'):
                raise ValueError(
                    '_multifile only supports file objects that '
                    'have a length but this one does not:', type(f), f)
        self._fileobjs = fileobjs
        self._index = 0

    @property
    def length(self):
        return sum(f.length for f in self._fileobjs)

    def read(self, amt=None):
        if amt <= 0:
            return ''.join(f.read() for f in self._fileobjs)
        parts = []
        while amt and self._index < len(self._fileobjs):
            parts.append(self._fileobjs[self._index].read(amt))
            got = len(parts[-1])
            if got < amt:
                self._index += 1
            amt -= got
        return ''.join(parts)

    def seek(self, offset, whence=os.SEEK_SET):
        if whence != os.SEEK_SET:
            raise NotImplementedError(
                '_multifile does not support anything other'
                ' than os.SEEK_SET for whence on seek()')
        if offset != 0:
            raise NotImplementedError(
                '_multifile only supports seeking to start, but that '
                'could be fixed if you need it')
        for f in self._fileobjs:
            f.seek(0)
        self._index = 0

def makev1commandrequest(ui, requestbuilder, caps, capablefn,
                         repobaseurl, cmd, args):
    """Make an HTTP request to run a command for a version 1 client.

    ``caps`` is a set of known server capabilities. The value may be
    None if capabilities are not yet known.

    ``capablefn`` is a function to evaluate a capability.

    ``cmd``, ``args``, and ``data`` define the command, its arguments, and
    raw data to pass to it.
    """
    if cmd == 'pushkey':
        args['data'] = ''
    data = args.pop('data', None)
    headers = args.pop('headers', {})

    ui.debug("sending %s command\n" % cmd)
    q = [('cmd', cmd)]
    headersize = 0
    # Important: don't use self.capable() here or else you end up
    # with infinite recursion when trying to look up capabilities
    # for the first time.
    postargsok = caps is not None and 'httppostargs' in caps

    # Send arguments via POST.
    if postargsok and args:
        strargs = urlreq.urlencode(sorted(args.items()))
        if not data:
            data = strargs
        else:
            if isinstance(data, bytes):
                i = io.BytesIO(data)
                i.length = len(data)
                data = i
            argsio = io.BytesIO(strargs)
            argsio.length = len(strargs)
            data = _multifile(argsio, data)
        headers[r'X-HgArgs-Post'] = len(strargs)
    elif args:
        # Calling self.capable() can infinite loop if we are calling
        # "capabilities". But that command should never accept wire
        # protocol arguments. So this should never happen.
        assert cmd != 'capabilities'
        httpheader = capablefn('httpheader')
        if httpheader:
            headersize = int(httpheader.split(',', 1)[0])

        # Send arguments via HTTP headers.
        if headersize > 0:
            # The headers can typically carry more data than the URL.
            encargs = urlreq.urlencode(sorted(args.items()))
            for header, value in encodevalueinheaders(encargs, 'X-HgArg',
                                                      headersize):
                headers[header] = value
        # Send arguments via query string (Mercurial <1.9).
        else:
            q += sorted(args.items())

    qs = '?%s' % urlreq.urlencode(q)
    cu = "%s%s" % (repobaseurl, qs)
    size = 0
    if util.safehasattr(data, 'length'):
        size = data.length
    elif data is not None:
        size = len(data)
    if data is not None and r'Content-Type' not in headers:
        headers[r'Content-Type'] = r'application/mercurial-0.1'

    # Tell the server we accept application/mercurial-0.2 and multiple
    # compression formats if the server is capable of emitting those
    # payloads.
    # Note: Keep this set empty by default, as client advertisement of
    # protocol parameters should only occur after the handshake.
    protoparams = set()

    mediatypes = set()
    if caps is not None:
        mt = capablefn('httpmediatype')
        if mt:
            protoparams.add('0.1')
            mediatypes = set(mt.split(','))

        protoparams.add('partial-pull')

    if '0.2tx' in mediatypes:
        protoparams.add('0.2')

    if '0.2tx' in mediatypes and capablefn('compression'):
        # We /could/ compare supported compression formats and prune
        # non-mutually supported or error if nothing is mutually supported.
        # For now, send the full list to the server and have it error.
        comps = [e.wireprotosupport().name for e in
                 util.compengines.supportedwireengines(util.CLIENTROLE)]
        protoparams.add('comp=%s' % ','.join(comps))

    if protoparams:
        protoheaders = encodevalueinheaders(' '.join(sorted(protoparams)),
                                            'X-HgProto',
                                            headersize or 1024)
        for header, value in protoheaders:
            headers[header] = value

    varyheaders = []
    for header in headers:
        if header.lower().startswith(r'x-hg'):
            varyheaders.append(header)

    if varyheaders:
        headers[r'Vary'] = r','.join(sorted(varyheaders))

    req = requestbuilder(pycompat.strurl(cu), data, headers)

    if data is not None:
        ui.debug("sending %d bytes\n" % size)
        req.add_unredirected_header(r'Content-Length', r'%d' % size)

    return req, cu, qs

def _reqdata(req):
    """Get request data, if any. If no data, returns None."""
    if pycompat.ispy3:
        return req.data
    if not req.has_data():
        return None
    return req.get_data()

def sendrequest(ui, opener, req):
    """Send a prepared HTTP request.

    Returns the response object.
    """
    dbg = ui.debug
    if (ui.debugflag
        and ui.configbool('devel', 'debug.peer-request')):
        line = 'devel-peer-request: %s\n'
        dbg(line % '%s %s' % (pycompat.bytesurl(req.get_method()),
                              pycompat.bytesurl(req.get_full_url())))
        hgargssize = None

        for header, value in sorted(req.header_items()):
            header = pycompat.bytesurl(header)
            value = pycompat.bytesurl(value)
            if header.startswith('X-hgarg-'):
                if hgargssize is None:
                    hgargssize = 0
                hgargssize += len(value)
            else:
                dbg(line % '  %s %s' % (header, value))

        if hgargssize is not None:
            dbg(line % '  %d bytes of commands arguments in headers'
                % hgargssize)
        data = _reqdata(req)
        if data is not None:
            length = getattr(data, 'length', None)
            if length is None:
                length = len(data)
            dbg(line % '  %d bytes of data' % length)

        start = util.timer()

    res = None
    try:
        res = opener.open(req)
    except urlerr.httperror as inst:
        if inst.code == 401:
            raise error.Abort(_('authorization failed'))
        raise
    except httplib.HTTPException as inst:
        ui.debug('http error requesting %s\n' %
                 util.hidepassword(req.get_full_url()))
        ui.traceback()
        raise IOError(None, inst)
    finally:
        if ui.debugflag and ui.configbool('devel', 'debug.peer-request'):
            code = res.code if res else -1
            dbg(line % '  finished in %.4f seconds (%d)'
                % (util.timer() - start, code))

    # Insert error handlers for common I/O failures.
    urlmod.wrapresponse(res)

    return res

class RedirectedRepoError(error.RepoError):
    def __init__(self, msg, respurl):
        super(RedirectedRepoError, self).__init__(msg)
        self.respurl = respurl

def parsev1commandresponse(ui, baseurl, requrl, qs, resp, compressible,
                           allowcbor=False):
    # record the url we got redirected to
    redirected = False
    respurl = pycompat.bytesurl(resp.geturl())
    if respurl.endswith(qs):
        respurl = respurl[:-len(qs)]
        qsdropped = False
    else:
        qsdropped = True

    if baseurl.rstrip('/') != respurl.rstrip('/'):
        redirected = True
        if not ui.quiet:
            ui.warn(_('real URL is %s\n') % respurl)

    try:
        proto = pycompat.bytesurl(resp.getheader(r'content-type', r''))
    except AttributeError:
        proto = pycompat.bytesurl(resp.headers.get(r'content-type', r''))

    safeurl = util.hidepassword(baseurl)
    if proto.startswith('application/hg-error'):
        raise error.OutOfBandError(resp.read())

    # Pre 1.0 versions of Mercurial used text/plain and
    # application/hg-changegroup. We don't support such old servers.
    if not proto.startswith('application/mercurial-'):
        ui.debug("requested URL: '%s'\n" % util.hidepassword(requrl))
        msg = _("'%s' does not appear to be an hg repository:\n"
                "---%%<--- (%s)\n%s\n---%%<---\n") % (
            safeurl, proto or 'no content-type', resp.read(1024))

        # Some servers may strip the query string from the redirect. We
        # raise a special error type so callers can react to this specially.
        if redirected and qsdropped:
            raise RedirectedRepoError(msg, respurl)
        else:
            raise error.RepoError(msg)

    try:
        subtype = proto.split('-', 1)[1]

        # Unless we end up supporting CBOR in the legacy wire protocol,
        # this should ONLY be encountered for the initial capabilities
        # request during handshake.
        if subtype == 'cbor':
            if allowcbor:
                return respurl, proto, resp
            else:
                raise error.RepoError(_('unexpected CBOR response from '
                                        'server'))

        version_info = tuple([int(n) for n in subtype.split('.')])
    except ValueError:
        raise error.RepoError(_("'%s' sent a broken Content-Type "
                                "header (%s)") % (safeurl, proto))

    # TODO consider switching to a decompression reader that uses
    # generators.
    if version_info == (0, 1):
        if compressible:
            resp = util.compengines['zlib'].decompressorreader(resp)

    elif version_info == (0, 2):
        # application/mercurial-0.2 always identifies the compression
        # engine in the payload header.
        elen = struct.unpack('B', util.readexactly(resp, 1))[0]
        ename = util.readexactly(resp, elen)
        engine = util.compengines.forwiretype(ename)

        resp = engine.decompressorreader(resp)
    else:
        raise error.RepoError(_("'%s' uses newer protocol %s") %
                              (safeurl, subtype))

    return respurl, proto, resp

class httppeer(wireprotov1peer.wirepeer):
    def __init__(self, ui, path, url, opener, requestbuilder, caps):
        self.ui = ui
        self._path = path
        self._url = url
        self._caps = caps
        self._urlopener = opener
        self._requestbuilder = requestbuilder

    def __del__(self):
        for h in self._urlopener.handlers:
            h.close()
            getattr(h, "close_all", lambda: None)()

    # Begin of ipeerconnection interface.

    def url(self):
        return self._path

    def local(self):
        return None

    def peer(self):
        return self

    def canpush(self):
        return True

    def close(self):
        try:
            reqs, sent, recv = (self._urlopener.requestscount,
                                self._urlopener.sentbytescount,
                                self._urlopener.receivedbytescount)
        except AttributeError:
            return
        self.ui.note(_('(sent %d HTTP requests and %d bytes; '
                       'received %d bytes in responses)\n') %
                     (reqs, sent, recv))

    # End of ipeerconnection interface.

    # Begin of ipeercommands interface.

    def capabilities(self):
        return self._caps

    # End of ipeercommands interface.

    def _callstream(self, cmd, _compressible=False, **args):
        args = pycompat.byteskwargs(args)

        req, cu, qs = makev1commandrequest(self.ui, self._requestbuilder,
                                           self._caps, self.capable,
                                           self._url, cmd, args)

        resp = sendrequest(self.ui, self._urlopener, req)

        self._url, ct, resp = parsev1commandresponse(self.ui, self._url, cu, qs,
                                                     resp, _compressible)

        return resp

    def _call(self, cmd, **args):
        fp = self._callstream(cmd, **args)
        try:
            return fp.read()
        finally:
            # if using keepalive, allow connection to be reused
            fp.close()

    def _callpush(self, cmd, cg, **args):
        # have to stream bundle to a temp file because we do not have
        # http 1.1 chunked transfer.

        types = self.capable('unbundle')
        try:
            types = types.split(',')
        except AttributeError:
            # servers older than d1b16a746db6 will send 'unbundle' as a
            # boolean capability. They only support headerless/uncompressed
            # bundles.
            types = [""]
        for x in types:
            if x in bundle2.bundletypes:
                type = x
                break

        tempname = bundle2.writebundle(self.ui, cg, None, type)
        fp = httpconnection.httpsendfile(self.ui, tempname, "rb")
        headers = {r'Content-Type': r'application/mercurial-0.1'}

        try:
            r = self._call(cmd, data=fp, headers=headers, **args)
            vals = r.split('\n', 1)
            if len(vals) < 2:
                raise error.ResponseError(_("unexpected response:"), r)
            return vals
        except urlerr.httperror:
            # Catch and re-raise these so we don't try and treat them
            # like generic socket errors. They lack any values in
            # .args on Python 3 which breaks our socket.error block.
            raise
        except socket.error as err:
            if err.args[0] in (errno.ECONNRESET, errno.EPIPE):
                raise error.Abort(_('push failed: %s') % err.args[1])
            raise error.Abort(err.args[1])
        finally:
            fp.close()
            os.unlink(tempname)

    def _calltwowaystream(self, cmd, fp, **args):
        fh = None
        fp_ = None
        filename = None
        try:
            # dump bundle to disk
            fd, filename = pycompat.mkstemp(prefix="hg-bundle-", suffix=".hg")
            fh = os.fdopen(fd, r"wb")
            d = fp.read(4096)
            while d:
                fh.write(d)
                d = fp.read(4096)
            fh.close()
            # start http push
            fp_ = httpconnection.httpsendfile(self.ui, filename, "rb")
            headers = {r'Content-Type': r'application/mercurial-0.1'}
            return self._callstream(cmd, data=fp_, headers=headers, **args)
        finally:
            if fp_ is not None:
                fp_.close()
            if fh is not None:
                fh.close()
                os.unlink(filename)

    def _callcompressable(self, cmd, **args):
        return self._callstream(cmd, _compressible=True, **args)

    def _abort(self, exception):
        raise exception

def sendv2request(ui, opener, requestbuilder, apiurl, permission, requests,
                  redirect):
    wireprotoframing.populatestreamencoders()

    uiencoders = ui.configlist(b'experimental', b'httppeer.v2-encoder-order')

    if uiencoders:
        encoders = []

        for encoder in uiencoders:
            if encoder not in wireprotoframing.STREAM_ENCODERS:
                ui.warn(_(b'wire protocol version 2 encoder referenced in '
                          b'config (%s) is not known; ignoring\n') % encoder)
            else:
                encoders.append(encoder)

    else:
        encoders = wireprotoframing.STREAM_ENCODERS_ORDER

    reactor = wireprotoframing.clientreactor(ui,
                                             hasmultiplesend=False,
                                             buffersends=True,
                                             clientcontentencoders=encoders)

    handler = wireprotov2peer.clienthandler(ui, reactor,
                                            opener=opener,
                                            requestbuilder=requestbuilder)

    url = '%s/%s' % (apiurl, permission)

    if len(requests) > 1:
        url += '/multirequest'
    else:
        url += '/%s' % requests[0][0]

    ui.debug('sending %d commands\n' % len(requests))
    for command, args, f in requests:
        ui.debug('sending command %s: %s\n' % (
            command, stringutil.pprint(args, indent=2)))
        assert not list(handler.callcommand(command, args, f,
                                            redirect=redirect))

    # TODO stream this.
    body = b''.join(map(bytes, handler.flushcommands()))

    # TODO modify user-agent to reflect v2
    headers = {
        r'Accept': wireprotov2server.FRAMINGTYPE,
        r'Content-Type': wireprotov2server.FRAMINGTYPE,
    }

    req = requestbuilder(pycompat.strurl(url), body, headers)
    req.add_unredirected_header(r'Content-Length', r'%d' % len(body))

    try:
        res = opener.open(req)
    except urlerr.httperror as e:
        if e.code == 401:
            raise error.Abort(_('authorization failed'))

        raise
    except httplib.HTTPException as e:
        ui.traceback()
        raise IOError(None, e)

    return handler, res

class queuedcommandfuture(pycompat.futures.Future):
    """Wraps result() on command futures to trigger submission on call."""

    def result(self, timeout=None):
        if self.done():
            return pycompat.futures.Future.result(self, timeout)

        self._peerexecutor.sendcommands()

        # sendcommands() will restore the original __class__ and self.result
        # will resolve to Future.result.
        return self.result(timeout)

@interfaceutil.implementer(repository.ipeercommandexecutor)
class httpv2executor(object):
    def __init__(self, ui, opener, requestbuilder, apiurl, descriptor,
                 redirect):
        self._ui = ui
        self._opener = opener
        self._requestbuilder = requestbuilder
        self._apiurl = apiurl
        self._descriptor = descriptor
        self._redirect = redirect
        self._sent = False
        self._closed = False
        self._neededpermissions = set()
        self._calls = []
        self._futures = weakref.WeakSet()
        self._responseexecutor = None
        self._responsef = None

    def __enter__(self):
        return self

    def __exit__(self, exctype, excvalue, exctb):
        self.close()

    def callcommand(self, command, args):
        if self._sent:
            raise error.ProgrammingError('callcommand() cannot be used after '
                                         'commands are sent')

        if self._closed:
            raise error.ProgrammingError('callcommand() cannot be used after '
                                         'close()')

        # The service advertises which commands are available. So if we attempt
        # to call an unknown command or pass an unknown argument, we can screen
        # for this.
        if command not in self._descriptor['commands']:
            raise error.ProgrammingError(
                'wire protocol command %s is not available' % command)

        cmdinfo = self._descriptor['commands'][command]
        unknownargs = set(args.keys()) - set(cmdinfo.get('args', {}))

        if unknownargs:
            raise error.ProgrammingError(
                'wire protocol command %s does not accept argument: %s' % (
                    command, ', '.join(sorted(unknownargs))))

        self._neededpermissions |= set(cmdinfo['permissions'])

        # TODO we /could/ also validate types here, since the API descriptor
        # includes types...

        f = pycompat.futures.Future()

        # Monkeypatch it so result() triggers sendcommands(), otherwise result()
        # could deadlock.
        f.__class__ = queuedcommandfuture
        f._peerexecutor = self

        self._futures.add(f)
        self._calls.append((command, args, f))

        return f

    def sendcommands(self):
        if self._sent:
            return

        if not self._calls:
            return

        self._sent = True

        # Unhack any future types so caller sees a clean type and so we
        # break reference cycle.
        for f in self._futures:
            if isinstance(f, queuedcommandfuture):
                f.__class__ = pycompat.futures.Future
                f._peerexecutor = None

        # Mark the future as running and filter out cancelled futures.
        calls = [(command, args, f)
                 for command, args, f in self._calls
                 if f.set_running_or_notify_cancel()]

        # Clear out references, prevent improper object usage.
        self._calls = None

        if not calls:
            return

        permissions = set(self._neededpermissions)

        if 'push' in permissions and 'pull' in permissions:
            permissions.remove('pull')

        if len(permissions) > 1:
            raise error.RepoError(_('cannot make request requiring multiple '
                                    'permissions: %s') %
                                  _(', ').join(sorted(permissions)))

        permission = {
            'push': 'rw',
            'pull': 'ro',
        }[permissions.pop()]

        handler, resp = sendv2request(
            self._ui, self._opener, self._requestbuilder, self._apiurl,
            permission, calls, self._redirect)

        # TODO we probably want to validate the HTTP code, media type, etc.

        self._responseexecutor = pycompat.futures.ThreadPoolExecutor(1)
        self._responsef = self._responseexecutor.submit(self._handleresponse,
                                                        handler, resp)

    def close(self):
        if self._closed:
            return

        self.sendcommands()

        self._closed = True

        if not self._responsef:
            return

        # TODO ^C here may not result in immediate program termination.

        try:
            self._responsef.result()
        finally:
            self._responseexecutor.shutdown(wait=True)
            self._responsef = None
            self._responseexecutor = None

            # If any of our futures are still in progress, mark them as
            # errored, otherwise a result() could wait indefinitely.
            for f in self._futures:
                if not f.done():
                    f.set_exception(error.ResponseError(
                        _('unfulfilled command response')))

            self._futures = None

    def _handleresponse(self, handler, resp):
        # Called in a thread to read the response.

        while handler.readdata(resp):
            pass

@interfaceutil.implementer(repository.ipeerv2)
class httpv2peer(object):
    def __init__(self, ui, repourl, apipath, opener, requestbuilder,
                 apidescriptor):
        self.ui = ui
        self.apidescriptor = apidescriptor

        if repourl.endswith('/'):
            repourl = repourl[:-1]

        self._url = repourl
        self._apipath = apipath
        self._apiurl = '%s/%s' % (repourl, apipath)
        self._opener = opener
        self._requestbuilder = requestbuilder

        self._redirect = wireprotov2peer.supportedredirects(ui, apidescriptor)

    # Start of ipeerconnection.

    def url(self):
        return self._url

    def local(self):
        return None

    def peer(self):
        return self

    def canpush(self):
        # TODO change once implemented.
        return False

    def close(self):
        self.ui.note(_('(sent %d HTTP requests and %d bytes; '
                       'received %d bytes in responses)\n') %
                     (self._opener.requestscount,
                      self._opener.sentbytescount,
                      self._opener.receivedbytescount))

    # End of ipeerconnection.

    # Start of ipeercapabilities.

    def capable(self, name):
        # The capabilities used internally historically map to capabilities
        # advertised from the "capabilities" wire protocol command. However,
        # version 2 of that command works differently.

        # Maps to commands that are available.
        if name in ('branchmap', 'getbundle', 'known', 'lookup', 'pushkey'):
            return True

        # Other concepts.
        if name in ('bundle2'):
            return True

        # Alias command-* to presence of command of that name.
        if name.startswith('command-'):
            return name[len('command-'):] in self.apidescriptor['commands']

        return False

    def requirecap(self, name, purpose):
        if self.capable(name):
            return

        raise error.CapabilityError(
            _('cannot %s; client or remote repository does not support the '
              '\'%s\' capability') % (purpose, name))

    # End of ipeercapabilities.

    def _call(self, name, **args):
        with self.commandexecutor() as e:
            return e.callcommand(name, args).result()

    def commandexecutor(self):
        return httpv2executor(self.ui, self._opener, self._requestbuilder,
                              self._apiurl, self.apidescriptor, self._redirect)

# Registry of API service names to metadata about peers that handle it.
#
# The following keys are meaningful:
#
# init
#    Callable receiving (ui, repourl, servicepath, opener, requestbuilder,
#                        apidescriptor) to create a peer.
#
# priority
#    Integer priority for the service. If we could choose from multiple
#    services, we choose the one with the highest priority.
API_PEERS = {
    wireprototypes.HTTP_WIREPROTO_V2: {
        'init': httpv2peer,
        'priority': 50,
    },
}

def performhandshake(ui, url, opener, requestbuilder):
    # The handshake is a request to the capabilities command.

    caps = None
    def capable(x):
        raise error.ProgrammingError('should not be called')

    args = {}

    # The client advertises support for newer protocols by adding an
    # X-HgUpgrade-* header with a list of supported APIs and an
    # X-HgProto-* header advertising which serializing formats it supports.
    # We only support the HTTP version 2 transport and CBOR responses for
    # now.
    advertisev2 = ui.configbool('experimental', 'httppeer.advertise-v2')

    if advertisev2:
        args['headers'] = {
            r'X-HgProto-1': r'cbor',
        }

        args['headers'].update(
            encodevalueinheaders(' '.join(sorted(API_PEERS)),
                                 'X-HgUpgrade',
                                 # We don't know the header limit this early.
                                 # So make it small.
                                 1024))

    req, requrl, qs = makev1commandrequest(ui, requestbuilder, caps,
                                           capable, url, 'capabilities',
                                           args)
    resp = sendrequest(ui, opener, req)

    # The server may redirect us to the repo root, stripping the
    # ?cmd=capabilities query string from the URL. The server would likely
    # return HTML in this case and ``parsev1commandresponse()`` would raise.
    # We catch this special case and re-issue the capabilities request against
    # the new URL.
    #
    # We should ideally not do this, as a redirect that drops the query
    # string from the URL is arguably a server bug. (Garbage in, garbage out).
    # However,  Mercurial clients for several years appeared to handle this
    # issue without behavior degradation. And according to issue 5860, it may
    # be a longstanding bug in some server implementations. So we allow a
    # redirect that drops the query string to "just work."
    try:
        respurl, ct, resp = parsev1commandresponse(ui, url, requrl, qs, resp,
                                                   compressible=False,
                                                   allowcbor=advertisev2)
    except RedirectedRepoError as e:
        req, requrl, qs = makev1commandrequest(ui, requestbuilder, caps,
                                               capable, e.respurl,
                                               'capabilities', args)
        resp = sendrequest(ui, opener, req)
        respurl, ct, resp = parsev1commandresponse(ui, url, requrl, qs, resp,
                                                   compressible=False,
                                                   allowcbor=advertisev2)

    try:
        rawdata = resp.read()
    finally:
        resp.close()

    if not ct.startswith('application/mercurial-'):
        raise error.ProgrammingError('unexpected content-type: %s' % ct)

    if advertisev2:
        if ct == 'application/mercurial-cbor':
            try:
                info = cborutil.decodeall(rawdata)[0]
            except cborutil.CBORDecodeError:
                raise error.Abort(_('error decoding CBOR from remote server'),
                                  hint=_('try again and consider contacting '
                                         'the server operator'))

        # We got a legacy response. That's fine.
        elif ct in ('application/mercurial-0.1', 'application/mercurial-0.2'):
            info = {
                'v1capabilities': set(rawdata.split())
            }

        else:
            raise error.RepoError(
                _('unexpected response type from server: %s') % ct)
    else:
        info = {
            'v1capabilities': set(rawdata.split())
        }

    return respurl, info

def makepeer(ui, path, opener=None, requestbuilder=urlreq.request):
    """Construct an appropriate HTTP peer instance.

    ``opener`` is an ``url.opener`` that should be used to establish
    connections, perform HTTP requests.

    ``requestbuilder`` is the type used for constructing HTTP requests.
    It exists as an argument so extensions can override the default.
    """
    u = util.url(path)
    if u.query or u.fragment:
        raise error.Abort(_('unsupported URL component: "%s"') %
                          (u.query or u.fragment))

    # urllib cannot handle URLs with embedded user or passwd.
    url, authinfo = u.authinfo()
    ui.debug('using %s\n' % url)

    opener = opener or urlmod.opener(ui, authinfo)

    respurl, info = performhandshake(ui, url, opener, requestbuilder)

    # Given the intersection of APIs that both we and the server support,
    # sort by their advertised priority and pick the first one.
    #
    # TODO consider making this request-based and interface driven. For
    # example, the caller could say "I want a peer that does X." It's quite
    # possible that not all peers would do that. Since we know the service
    # capabilities, we could filter out services not meeting the
    # requirements. Possibly by consulting the interfaces defined by the
    # peer type.
    apipeerchoices = set(info.get('apis', {}).keys()) & set(API_PEERS.keys())

    preferredchoices = sorted(apipeerchoices,
                              key=lambda x: API_PEERS[x]['priority'],
                              reverse=True)

    for service in preferredchoices:
        apipath = '%s/%s' % (info['apibase'].rstrip('/'), service)

        return API_PEERS[service]['init'](ui, respurl, apipath, opener,
                                          requestbuilder,
                                          info['apis'][service])

    # Failed to construct an API peer. Fall back to legacy.
    return httppeer(ui, path, respurl, opener, requestbuilder,
                    info['v1capabilities'])

def instance(ui, path, create, intents=None, createopts=None):
    if create:
        raise error.Abort(_('cannot create new http repository'))
    try:
        if path.startswith('https:') and not urlmod.has_https:
            raise error.Abort(_('Python support for SSL and HTTPS '
                                'is not installed'))

        inst = makepeer(ui, path)

        return inst
    except error.RepoError as httpexception:
        try:
            r = statichttprepo.instance(ui, "static-" + path, create)
            ui.note(_('(falling back to static-http)\n'))
            return r
        except error.RepoError:
            raise httpexception # use the original http RepoError instead