mercurial/hgweb/protocol.py
author Gregory Szorc <gregory.szorc@gmail.com>
Sun, 07 Aug 2016 18:09:58 -0700
changeset 29748 5e2365698d44
parent 28883 032c4c2f802a
child 29788 b1809f5d7630
permissions -rw-r--r--
hgweb: config option to control zlib compression level Before this patch, the HTTP transport protocol would always zlib compress certain responses (notably "getbundle" wire protocol commands) at zlib compression level 6. zlib can be a massive CPU resource sink for servers. Some server operators may wish to reduce server-side CPU requirements while requiring more bandwidth. This is common on corporate intranets, for example. Others may wish to use more CPU but reduce bandwidth. This patch introduces a config option to allow server operators to control the zlib compression level. On the "mozilla-unified" generaldelta repository, setting this value to "0" (disable compression) results in server-side CPU utilization for a `hg clone` going from ~180s to ~124s CPU time on my i7-6700K. A level of "1" (which increases the transfer size from ~1,074 MB at level 6 to ~1,222 MB) utilizes ~132s CPU time.
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
5598
d534ba1c4eb4 separate the wire protocol commands from the user interface commands
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents:
diff changeset
     1
#
d534ba1c4eb4 separate the wire protocol commands from the user interface commands
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents:
diff changeset
     2
# Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net>
d534ba1c4eb4 separate the wire protocol commands from the user interface commands
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents:
diff changeset
     3
# Copyright 2005-2007 Matt Mackall <mpm@selenic.com>
d534ba1c4eb4 separate the wire protocol commands from the user interface commands
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents:
diff changeset
     4
#
8225
46293a0c7e9f updated license to be explicit about GPL version 2
Martin Geisler <mg@lazybytes.net>
parents: 8109
diff changeset
     5
# This software may be used and distributed according to the terms of the
10263
25e572394f5c Update license to GPLv2+
Matt Mackall <mpm@selenic.com>
parents: 9713
diff changeset
     6
# GNU General Public License version 2 or any later version.
5598
d534ba1c4eb4 separate the wire protocol commands from the user interface commands
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents:
diff changeset
     7
27046
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
     8
from __future__ import absolute_import
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
     9
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    10
import cgi
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    11
import zlib
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    12
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    13
from .common import (
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    14
    HTTP_OK,
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    15
)
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    16
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    17
from .. import (
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    18
    util,
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    19
    wireproto,
37fcfe52c68c hgweb: use absolute_import
Yuya Nishihara <yuya@tcha.org>
parents: 20903
diff changeset
    20
)
28861
86db5cb55d46 pycompat: switch to util.stringio for py3 compat
timeless <timeless@mozdev.org>
parents: 28530
diff changeset
    21
stringio = util.stringio
5963
5be210afe1b8 hgweb: explicitly check if requested command exists
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 5915
diff changeset
    22
28883
032c4c2f802a pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents: 28861
diff changeset
    23
urlerr = util.urlerr
032c4c2f802a pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents: 28861
diff changeset
    24
urlreq = util.urlreq
032c4c2f802a pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents: 28861
diff changeset
    25
5993
948a41e77902 hgweb: explicit response status
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 5963
diff changeset
    26
HGTYPE = 'application/mercurial-0.1'
15017
f4522df38c65 wireproto: add out-of-band error class to allow remote repo to report errors
Andrew Pritchard <andrewp@fogcreek.com>
parents: 14614
diff changeset
    27
HGERRTYPE = 'application/hg-error'
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    28
20903
8d477543882b wireproto: introduce an abstractserverproto class
Pierre-Yves David <pierre-yves.david@fb.com>
parents: 18352
diff changeset
    29
class webproto(wireproto.abstractserverproto):
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    30
    def __init__(self, req, ui):
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    31
        self.req = req
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    32
        self.response = ''
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    33
        self.ui = ui
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    34
    def getargs(self, args):
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    35
        knownargs = self._args()
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    36
        data = {}
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    37
        keys = args.split()
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    38
        for k in keys:
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    39
            if k == '*':
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    40
                star = {}
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    41
                for key in knownargs.keys():
13721
3458c15ab2f0 wireproto: fix handling of '*' args for HTTP and SSH
Peter Arrenbrecht <peter.arrenbrecht@gmail.com>
parents: 12704
diff changeset
    42
                    if key != 'cmd' and key not in keys:
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    43
                        star[key] = knownargs[key][0]
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    44
                data['*'] = star
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    45
            else:
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    46
                data[k] = knownargs[k][0]
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    47
        return [data[k] for k in keys]
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    48
    def _args(self):
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    49
        args = self.req.form.copy()
28530
fd2acc5046f6 http: support sending hgargs via POST body instead of in GET or headers
Augie Fackler <augie@google.com>
parents: 27046
diff changeset
    50
        postlen = int(self.req.env.get('HTTP_X_HGARGS_POST', 0))
fd2acc5046f6 http: support sending hgargs via POST body instead of in GET or headers
Augie Fackler <augie@google.com>
parents: 27046
diff changeset
    51
        if postlen:
fd2acc5046f6 http: support sending hgargs via POST body instead of in GET or headers
Augie Fackler <augie@google.com>
parents: 27046
diff changeset
    52
            args.update(cgi.parse_qs(
fd2acc5046f6 http: support sending hgargs via POST body instead of in GET or headers
Augie Fackler <augie@google.com>
parents: 27046
diff changeset
    53
                self.req.read(postlen), keep_blank_values=True))
fd2acc5046f6 http: support sending hgargs via POST body instead of in GET or headers
Augie Fackler <augie@google.com>
parents: 27046
diff changeset
    54
            return args
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    55
        chunks = []
14094
d10c6835497e http: minor tweaks to long arg handling
Matt Mackall <mpm@selenic.com>
parents: 14093
diff changeset
    56
        i = 1
14494
1ffeeb91c55d check-code: flag 0/1 used as constant Boolean expression
Martin Geisler <mg@lazybytes.net>
parents: 14094
diff changeset
    57
        while True:
14094
d10c6835497e http: minor tweaks to long arg handling
Matt Mackall <mpm@selenic.com>
parents: 14093
diff changeset
    58
            h = self.req.env.get('HTTP_X_HGARG_' + str(i))
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    59
            if h is None:
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    60
                break
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    61
            chunks += [h]
14094
d10c6835497e http: minor tweaks to long arg handling
Matt Mackall <mpm@selenic.com>
parents: 14093
diff changeset
    62
            i += 1
14093
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    63
        args.update(cgi.parse_qs(''.join(chunks), keep_blank_values=True))
ce99d887585f httprepo: long arguments support (issue2126)
Steven Brown <StevenGBrown@gmail.com>
parents: 13721
diff changeset
    64
        return args
11621
e46a8b2331a6 protocol: shuffle server methods to group send methods
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11618
diff changeset
    65
    def getfile(self, fp):
e46a8b2331a6 protocol: shuffle server methods to group send methods
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11618
diff changeset
    66
        length = int(self.req.env['CONTENT_LENGTH'])
e46a8b2331a6 protocol: shuffle server methods to group send methods
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11618
diff changeset
    67
        for s in util.filechunkiter(self.req, limit=length):
e46a8b2331a6 protocol: shuffle server methods to group send methods
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11618
diff changeset
    68
            fp.write(s)
e46a8b2331a6 protocol: shuffle server methods to group send methods
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11618
diff changeset
    69
    def redirect(self):
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    70
        self.oldio = self.ui.fout, self.ui.ferr
28861
86db5cb55d46 pycompat: switch to util.stringio for py3 compat
timeless <timeless@mozdev.org>
parents: 28530
diff changeset
    71
        self.ui.ferr = self.ui.fout = stringio()
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    72
    def restore(self):
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    73
        val = self.ui.fout.getvalue()
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    74
        self.ui.ferr, self.ui.fout = self.oldio
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    75
        return val
11623
31d0a6d50ee2 protocol: extract compression from streaming mechanics
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11622
diff changeset
    76
    def groupchunks(self, cg):
29748
5e2365698d44 hgweb: config option to control zlib compression level
Gregory Szorc <gregory.szorc@gmail.com>
parents: 28883
diff changeset
    77
        z = zlib.compressobj(self.ui.configint('server', 'zliblevel', -1))
14494
1ffeeb91c55d check-code: flag 0/1 used as constant Boolean expression
Martin Geisler <mg@lazybytes.net>
parents: 14094
diff changeset
    78
        while True:
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    79
            chunk = cg.read(4096)
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    80
            if not chunk:
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    81
                break
11623
31d0a6d50ee2 protocol: extract compression from streaming mechanics
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11622
diff changeset
    82
            yield z.compress(chunk)
31d0a6d50ee2 protocol: extract compression from streaming mechanics
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11622
diff changeset
    83
        yield z.flush()
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    84
    def _client(self):
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    85
        return 'remote:%s:%s:%s' % (
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    86
            self.req.env.get('wsgi.url_scheme') or 'http',
28883
032c4c2f802a pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents: 28861
diff changeset
    87
            urlreq.quote(self.req.env.get('REMOTE_HOST', '')),
032c4c2f802a pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents: 28861
diff changeset
    88
            urlreq.quote(self.req.env.get('REMOTE_USER', '')))
11595
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    89
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    90
def iscmd(cmd):
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    91
    return cmd in wireproto.commands
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    92
368cd5325348 protocol: move hgweb protocol support back into protocol.py
Matt Mackall <mpm@selenic.com>
parents: 11594
diff changeset
    93
def call(repo, req, cmd):
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
    94
    p = webproto(req, repo.ui)
11625
cdeb861335d5 protocol: wrap non-string protocol responses in classes
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11623
diff changeset
    95
    rsp = wireproto.dispatch(repo, p, cmd)
11626
2f8adc60e013 protocol: use generators instead of req.write() for hgweb stream responses
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11625
diff changeset
    96
    if isinstance(rsp, str):
18352
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
    97
        req.respond(HTTP_OK, HGTYPE, body=rsp)
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
    98
        return []
11626
2f8adc60e013 protocol: use generators instead of req.write() for hgweb stream responses
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11625
diff changeset
    99
    elif isinstance(rsp, wireproto.streamres):
2f8adc60e013 protocol: use generators instead of req.write() for hgweb stream responses
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11625
diff changeset
   100
        req.respond(HTTP_OK, HGTYPE)
2f8adc60e013 protocol: use generators instead of req.write() for hgweb stream responses
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11625
diff changeset
   101
        return rsp.gen
2f8adc60e013 protocol: use generators instead of req.write() for hgweb stream responses
Dirkjan Ochtman <dirkjan@ochtman.nl>
parents: 11625
diff changeset
   102
    elif isinstance(rsp, wireproto.pushres):
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
   103
        val = p.restore()
18346
6c2563b2c1c6 hgweb: use Content-Length for pushres
Mads Kiilerich <mads@kiilerich.com>
parents: 15017
diff changeset
   104
        rsp = '%d\n%s' % (rsp.res, val)
18352
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
   105
        req.respond(HTTP_OK, HGTYPE, body=rsp)
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
   106
        return []
12703
40bb5853fc4b wireproto: introduce pusherr() to deal with "unsynced changes" error
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 11626
diff changeset
   107
    elif isinstance(rsp, wireproto.pusherr):
12704
ca6e2adc3e4d wireproto/http: drain the incoming bundle in case of errors
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 12703
diff changeset
   108
        # drain the incoming bundle
ca6e2adc3e4d wireproto/http: drain the incoming bundle in case of errors
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 12703
diff changeset
   109
        req.drain()
14614
afccc64eea73 ui: use I/O descriptors internally
Idan Kamara <idankk86@gmail.com>
parents: 14494
diff changeset
   110
        p.restore()
12703
40bb5853fc4b wireproto: introduce pusherr() to deal with "unsynced changes" error
Benoit Boissinot <benoit.boissinot@ens-lyon.org>
parents: 11626
diff changeset
   111
        rsp = '0\n%s\n' % rsp.res
18352
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
   112
        req.respond(HTTP_OK, HGTYPE, body=rsp)
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
   113
        return []
15017
f4522df38c65 wireproto: add out-of-band error class to allow remote repo to report errors
Andrew Pritchard <andrewp@fogcreek.com>
parents: 14614
diff changeset
   114
    elif isinstance(rsp, wireproto.ooberror):
f4522df38c65 wireproto: add out-of-band error class to allow remote repo to report errors
Andrew Pritchard <andrewp@fogcreek.com>
parents: 14614
diff changeset
   115
        rsp = rsp.message
18352
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
   116
        req.respond(HTTP_OK, HGERRTYPE, body=rsp)
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length
Mads Kiilerich <mads@kiilerich.com>
parents: 18346
diff changeset
   117
        return []