Mercurial > hg
annotate hgext/largefiles/proto.py @ 29151:7996c3acc33b
wireproto: optimize handling of large batch responses
Now that batch can be used by remotefilelog, the quadratic string
copying this was doing was actually disastrous. In my local testing,
fetching a 56 meg file used to take 3 minutes, and now takes only a
few seconds.
author | Augie Fackler <augie@google.com> |
---|---|
date | Thu, 12 May 2016 09:39:14 -0400 |
parents | 032c4c2f802a |
children | 29139be0ccc7 |
rev | line source |
---|---|
15168 | 1 # Copyright 2011 Fog Creek Software |
2 # | |
3 # This software may be used and distributed according to the terms of the | |
4 # GNU General Public License version 2 or any later version. | |
5 | |
6 import os | |
19917
cff331cbb5ee
largefiles: make the protocol hack for replacing heads with lheads more precise
Mads Kiilerich <madski@unity3d.com>
parents:
19009
diff
changeset
|
7 import re |
15168 | 8 |
17192
1ac628cd7113
peer: introduce real peer classes
Peter Arrenbrecht <peter.arrenbrecht@gmail.com>
parents:
17127
diff
changeset
|
9 from mercurial import error, httppeer, util, wireproto |
15168 | 10 from mercurial.i18n import _ |
11 | |
28883
032c4c2f802a
pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents:
28576
diff
changeset
|
12 urlerr = util.urlerr |
032c4c2f802a
pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents:
28576
diff
changeset
|
13 urlreq = util.urlreq |
032c4c2f802a
pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents:
28576
diff
changeset
|
14 |
15168 | 15 import lfutil |
16 | |
15255
7ab05d752405
largefiles: cosmetics, whitespace, code style
Greg Ward <greg@gerg.ca>
parents:
15252
diff
changeset
|
17 LARGEFILES_REQUIRED_MSG = ('\nThis repository uses the largefiles extension.' |
7ab05d752405
largefiles: cosmetics, whitespace, code style
Greg Ward <greg@gerg.ca>
parents:
15252
diff
changeset
|
18 '\n\nPlease enable it in your Mercurial config ' |
7ab05d752405
largefiles: cosmetics, whitespace, code style
Greg Ward <greg@gerg.ca>
parents:
15252
diff
changeset
|
19 'file.\n') |
15168 | 20 |
18922
d2c4d37f7db5
largefiles: quiet (and document) undefined name errors (issue3886)
Bryan O'Sullivan <bryano@fb.com>
parents:
18488
diff
changeset
|
21 # these will all be replaced by largefiles.uisetup |
d2c4d37f7db5
largefiles: quiet (and document) undefined name errors (issue3886)
Bryan O'Sullivan <bryano@fb.com>
parents:
18488
diff
changeset
|
22 capabilitiesorig = None |
d2c4d37f7db5
largefiles: quiet (and document) undefined name errors (issue3886)
Bryan O'Sullivan <bryano@fb.com>
parents:
18488
diff
changeset
|
23 ssholdcallstream = None |
d2c4d37f7db5
largefiles: quiet (and document) undefined name errors (issue3886)
Bryan O'Sullivan <bryano@fb.com>
parents:
18488
diff
changeset
|
24 httpoldcallstream = None |
d2c4d37f7db5
largefiles: quiet (and document) undefined name errors (issue3886)
Bryan O'Sullivan <bryano@fb.com>
parents:
18488
diff
changeset
|
25 |
15168 | 26 def putlfile(repo, proto, sha): |
28576
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
27 '''Server command for putting a largefile into a repository's local store |
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
28 and into the user cache.''' |
15168 | 29 proto.redirect() |
15391
a5a6a9b7f3b9
largefiles: replace tempfile.NamedTemporaryFile with tempfile.mkstemp
Hao Lian <hao@fogcreek.com>
parents:
15317
diff
changeset
|
30 |
16594
5516fdf3fe24
largefiles: in putlfile, ensure tempfile's directory exists prior to creation
hlian
parents:
16247
diff
changeset
|
31 path = lfutil.storepath(repo, sha) |
5516fdf3fe24
largefiles: in putlfile, ensure tempfile's directory exists prior to creation
hlian
parents:
16247
diff
changeset
|
32 util.makedirs(os.path.dirname(path)) |
5516fdf3fe24
largefiles: in putlfile, ensure tempfile's directory exists prior to creation
hlian
parents:
16247
diff
changeset
|
33 tmpfp = util.atomictempfile(path, createmode=repo.store.createmode) |
5516fdf3fe24
largefiles: in putlfile, ensure tempfile's directory exists prior to creation
hlian
parents:
16247
diff
changeset
|
34 |
15168 | 35 try: |
25079
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
36 proto.getfile(tmpfp) |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
37 tmpfp._fp.seek(0) |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
38 if sha != lfutil.hexsha1(tmpfp._fp): |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
39 raise IOError(0, _('largefile contents do not match hash')) |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
40 tmpfp.close() |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
41 lfutil.linktousercache(repo, sha) |
25660
328739ea70c3
global: mass rewrite to use modern exception syntax
Gregory Szorc <gregory.szorc@gmail.com>
parents:
25079
diff
changeset
|
42 except IOError as e: |
25079
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
43 repo.ui.warn(_('largefiles: failed to put %s into store: %s\n') % |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
44 (sha, e.strerror)) |
bee00e0c2e45
largefiles: use try/except/finally
Matt Mackall <mpm@selenic.com>
parents:
21084
diff
changeset
|
45 return wireproto.pushres(1) |
15168 | 46 finally: |
16155
1b2b42e866be
largefiles: respect store.createmode and avoid extra file copy
Martin Geisler <mg@aragost.com>
parents:
15778
diff
changeset
|
47 tmpfp.discard() |
15168 | 48 |
49 return wireproto.pushres(0) | |
50 | |
51 def getlfile(repo, proto, sha): | |
28576
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
52 '''Server command for retrieving a largefile from the repository-local |
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
53 cache or user cache.''' |
15168 | 54 filename = lfutil.findfile(repo, sha) |
55 if not filename: | |
26587
56b2bcea2529
error: get Abort from 'error' instead of 'util'
Pierre-Yves David <pierre-yves.david@fb.com>
parents:
25660
diff
changeset
|
56 raise error.Abort(_('requested largefile %s not present in cache') |
56b2bcea2529
error: get Abort from 'error' instead of 'util'
Pierre-Yves David <pierre-yves.david@fb.com>
parents:
25660
diff
changeset
|
57 % sha) |
15168 | 58 f = open(filename, 'rb') |
59 length = os.fstat(f.fileno())[6] | |
15252
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
60 |
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
61 # Since we can't set an HTTP content-length header here, and |
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
62 # Mercurial core provides no way to give the length of a streamres |
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
63 # (and reading the entire file into RAM would be ill-advised), we |
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
64 # just send the length on the first line of the response, like the |
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
65 # ssh proto does for string responses. |
15168 | 66 def generator(): |
67 yield '%d\n' % length | |
19009
07e40d589b64
largefiles: use filechunkiter for iterating largefile when serving getlfile
Mads Kiilerich <madski@unity3d.com>
parents:
19006
diff
changeset
|
68 for chunk in util.filechunkiter(f): |
15168 | 69 yield chunk |
70 return wireproto.streamres(generator()) | |
71 | |
72 def statlfile(repo, proto, sha): | |
28576
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
73 '''Server command for checking if a largefile is present - returns '2\n' if |
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
74 the largefile is missing, '0\n' if it seems to be in good condition. |
18488
a977b42df8b3
largefiles: don't verify largefile hashes on servers when processing statlfile
Mads Kiilerich <madski@unity3d.com>
parents:
18298
diff
changeset
|
75 |
a977b42df8b3
largefiles: don't verify largefile hashes on servers when processing statlfile
Mads Kiilerich <madski@unity3d.com>
parents:
18298
diff
changeset
|
76 The value 1 is reserved for mismatched checksum, but that is too expensive |
a977b42df8b3
largefiles: don't verify largefile hashes on servers when processing statlfile
Mads Kiilerich <madski@unity3d.com>
parents:
18298
diff
changeset
|
77 to be verified on every stat and must be caught be running 'hg verify' |
a977b42df8b3
largefiles: don't verify largefile hashes on servers when processing statlfile
Mads Kiilerich <madski@unity3d.com>
parents:
18298
diff
changeset
|
78 server side.''' |
15168 | 79 filename = lfutil.findfile(repo, sha) |
80 if not filename: | |
81 return '2\n' | |
18488
a977b42df8b3
largefiles: don't verify largefile hashes on servers when processing statlfile
Mads Kiilerich <madski@unity3d.com>
parents:
18298
diff
changeset
|
82 return '0\n' |
15168 | 83 |
84 def wirereposetup(ui, repo): | |
85 class lfileswirerepository(repo.__class__): | |
86 def putlfile(self, sha, fd): | |
87 # unfortunately, httprepository._callpush tries to convert its | |
88 # input file-like into a bundle before sending it, so we can't use | |
89 # it ... | |
17192
1ac628cd7113
peer: introduce real peer classes
Peter Arrenbrecht <peter.arrenbrecht@gmail.com>
parents:
17127
diff
changeset
|
90 if issubclass(self.__class__, httppeer.httppeer): |
26825
78539633acf3
largefiles: don't mute and obfuscate http errors when putlfile fails
Mads Kiilerich <madski@unity3d.com>
parents:
26587
diff
changeset
|
91 res = self._call('putlfile', data=fd, sha=sha, |
78539633acf3
largefiles: don't mute and obfuscate http errors when putlfile fails
Mads Kiilerich <madski@unity3d.com>
parents:
26587
diff
changeset
|
92 headers={'content-type':'application/mercurial-0.1'}) |
15168 | 93 try: |
15778
f15c646bffc7
largefiles: display remote errors from putlfile (issue3123) (issue3149)
Kevin Gessner <kevin@fogcreek.com>
parents:
15391
diff
changeset
|
94 d, output = res.split('\n', 1) |
f15c646bffc7
largefiles: display remote errors from putlfile (issue3123) (issue3149)
Kevin Gessner <kevin@fogcreek.com>
parents:
15391
diff
changeset
|
95 for l in output.splitlines(True): |
19949
29f12a7a03ee
largefiles: don't add extra \n when displaying remote messages in putlfile
Mads Kiilerich <madski@unity3d.com>
parents:
19948
diff
changeset
|
96 self.ui.warn(_('remote: '), l) # assume l ends with \n |
15778
f15c646bffc7
largefiles: display remote errors from putlfile (issue3123) (issue3149)
Kevin Gessner <kevin@fogcreek.com>
parents:
15391
diff
changeset
|
97 return int(d) |
26825
78539633acf3
largefiles: don't mute and obfuscate http errors when putlfile fails
Mads Kiilerich <madski@unity3d.com>
parents:
26587
diff
changeset
|
98 except ValueError: |
19947
2a03faf8b5fe
largefiles: fix 'unexpected response' warning newlines
Mads Kiilerich <madski@unity3d.com>
parents:
19917
diff
changeset
|
99 self.ui.warn(_('unexpected putlfile response: %r\n') % res) |
15168 | 100 return 1 |
101 # ... but we can't use sshrepository._call because the data= | |
102 # argument won't get sent, and _callpush does exactly what we want | |
103 # in this case: send the data straight through | |
104 else: | |
105 try: | |
106 ret, output = self._callpush("putlfile", fd, sha=sha) | |
107 if ret == "": | |
108 raise error.ResponseError(_('putlfile failed:'), | |
109 output) | |
110 return int(ret) | |
111 except IOError: | |
112 return 1 | |
113 except ValueError: | |
114 raise error.ResponseError( | |
115 _('putlfile failed (unexpected response):'), ret) | |
116 | |
117 def getlfile(self, sha): | |
19004
6614e5e24e66
largefiles: move protocol conversion into getlfile and make it an iterable
Mads Kiilerich <madski@unity3d.com>
parents:
18922
diff
changeset
|
118 """returns an iterable with the chunks of the file with sha sha""" |
15168 | 119 stream = self._callstream("getlfile", sha=sha) |
120 length = stream.readline() | |
121 try: | |
122 length = int(length) | |
123 except ValueError: | |
15170
c1a4a3220711
largefiles: fix over-long lines
Matt Mackall <mpm@selenic.com>
parents:
15168
diff
changeset
|
124 self._abort(error.ResponseError(_("unexpected response:"), |
c1a4a3220711
largefiles: fix over-long lines
Matt Mackall <mpm@selenic.com>
parents:
15168
diff
changeset
|
125 length)) |
19004
6614e5e24e66
largefiles: move protocol conversion into getlfile and make it an iterable
Mads Kiilerich <madski@unity3d.com>
parents:
18922
diff
changeset
|
126 |
19005
1b84047e7d16
largefiles: drop limitreader, use filechunkiter limit
Mads Kiilerich <madski@unity3d.com>
parents:
19004
diff
changeset
|
127 # SSH streams will block if reading more than length |
1b84047e7d16
largefiles: drop limitreader, use filechunkiter limit
Mads Kiilerich <madski@unity3d.com>
parents:
19004
diff
changeset
|
128 for chunk in util.filechunkiter(stream, 128 * 1024, length): |
19004
6614e5e24e66
largefiles: move protocol conversion into getlfile and make it an iterable
Mads Kiilerich <madski@unity3d.com>
parents:
18922
diff
changeset
|
129 yield chunk |
19006
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
130 # HTTP streams must hit the end to process the last empty |
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
131 # chunk of Chunked-Encoding so the connection can be reused. |
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
132 if issubclass(self.__class__, httppeer.httppeer): |
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
133 chunk = stream.read(1) |
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
134 if chunk: |
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
135 self._abort(error.ResponseError(_("unexpected response:"), |
0b3b84222a2d
largefiles: getlfile must hit end of HTTP chunked streams to reuse connections
Mads Kiilerich <madski@unity3d.com>
parents:
19005
diff
changeset
|
136 chunk)) |
15168 | 137 |
21084
70252bdfd39c
largefiles: import whole modules instead of importing parts of them
Mads Kiilerich <madski@unity3d.com>
parents:
19949
diff
changeset
|
138 @wireproto.batchable |
15168 | 139 def statlfile(self, sha): |
21084
70252bdfd39c
largefiles: import whole modules instead of importing parts of them
Mads Kiilerich <madski@unity3d.com>
parents:
19949
diff
changeset
|
140 f = wireproto.future() |
17127
9e1616307c4c
largefiles: batch statlfile requests when pushing a largefiles repo (issue3386)
Na'Tosha Bard <natosha@unity3d.com>
parents:
16594
diff
changeset
|
141 result = {'sha': sha} |
9e1616307c4c
largefiles: batch statlfile requests when pushing a largefiles repo (issue3386)
Na'Tosha Bard <natosha@unity3d.com>
parents:
16594
diff
changeset
|
142 yield result, f |
15168 | 143 try: |
17127
9e1616307c4c
largefiles: batch statlfile requests when pushing a largefiles repo (issue3386)
Na'Tosha Bard <natosha@unity3d.com>
parents:
16594
diff
changeset
|
144 yield int(f.value) |
28883
032c4c2f802a
pycompat: switch to util.urlreq/util.urlerr for py3 compat
timeless <timeless@mozdev.org>
parents:
28576
diff
changeset
|
145 except (ValueError, urlerr.httperror): |
15252
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
146 # If the server returns anything but an integer followed by a |
15168 | 147 # newline, newline, it's not speaking our language; if we get |
148 # an HTTP error, we can't be sure the largefile is present; | |
15252
6e809bb4f969
largefiles: improve comments, internal docstrings
Greg Ward <greg@gerg.ca>
parents:
15224
diff
changeset
|
149 # either way, consider it missing. |
17127
9e1616307c4c
largefiles: batch statlfile requests when pushing a largefiles repo (issue3386)
Na'Tosha Bard <natosha@unity3d.com>
parents:
16594
diff
changeset
|
150 yield 2 |
15168 | 151 |
152 repo.__class__ = lfileswirerepository | |
153 | |
154 # advertise the largefiles=serve capability | |
155 def capabilities(repo, proto): | |
28576
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
156 '''Wrap server command to announce largefile server capability''' |
16247
d87d9d8a8e03
largefiles: remove use of underscores that breaks coding convention
Na'Tosha Bard <natosha@unity3d.com>
parents:
16155
diff
changeset
|
157 return capabilitiesorig(repo, proto) + ' largefiles=serve' |
15168 | 158 |
159 def heads(repo, proto): | |
28576
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
160 '''Wrap server command - largefile capable clients will know to call |
33bd95443e7f
largefiles: add some docstrings
Mads Kiilerich <madski@unity3d.com>
parents:
26825
diff
changeset
|
161 lheads instead''' |
15168 | 162 if lfutil.islfilesrepo(repo): |
15224
7c604d8c7e83
largefiles: remove pre-1.9 code from extension first bundled with 1.9
Na'Tosha Bard <natosha@unity3d.com>
parents:
15170
diff
changeset
|
163 return wireproto.ooberror(LARGEFILES_REQUIRED_MSG) |
15168 | 164 return wireproto.heads(repo, proto) |
165 | |
16247
d87d9d8a8e03
largefiles: remove use of underscores that breaks coding convention
Na'Tosha Bard <natosha@unity3d.com>
parents:
16155
diff
changeset
|
166 def sshrepocallstream(self, cmd, **args): |
15168 | 167 if cmd == 'heads' and self.capable('largefiles'): |
168 cmd = 'lheads' | |
169 if cmd == 'batch' and self.capable('largefiles'): | |
170 args['cmds'] = args['cmds'].replace('heads ', 'lheads ') | |
16247
d87d9d8a8e03
largefiles: remove use of underscores that breaks coding convention
Na'Tosha Bard <natosha@unity3d.com>
parents:
16155
diff
changeset
|
171 return ssholdcallstream(self, cmd, **args) |
15168 | 172 |
19917
cff331cbb5ee
largefiles: make the protocol hack for replacing heads with lheads more precise
Mads Kiilerich <madski@unity3d.com>
parents:
19009
diff
changeset
|
173 headsre = re.compile(r'(^|;)heads\b') |
cff331cbb5ee
largefiles: make the protocol hack for replacing heads with lheads more precise
Mads Kiilerich <madski@unity3d.com>
parents:
19009
diff
changeset
|
174 |
16247
d87d9d8a8e03
largefiles: remove use of underscores that breaks coding convention
Na'Tosha Bard <natosha@unity3d.com>
parents:
16155
diff
changeset
|
175 def httprepocallstream(self, cmd, **args): |
15168 | 176 if cmd == 'heads' and self.capable('largefiles'): |
177 cmd = 'lheads' | |
178 if cmd == 'batch' and self.capable('largefiles'): | |
19917
cff331cbb5ee
largefiles: make the protocol hack for replacing heads with lheads more precise
Mads Kiilerich <madski@unity3d.com>
parents:
19009
diff
changeset
|
179 args['cmds'] = headsre.sub('lheads', args['cmds']) |
16247
d87d9d8a8e03
largefiles: remove use of underscores that breaks coding convention
Na'Tosha Bard <natosha@unity3d.com>
parents:
16155
diff
changeset
|
180 return httpoldcallstream(self, cmd, **args) |