mercurial-scm/hg-stable: mercurial/hgweb/request.py annotate

annotate mercurial/hgweb/request.py @ 36861:2cdf47e14c30

hgweb: refactor the request draining code The previous code for draining was only invoked in a few places in the wire protocol. Behavior wasn't consist. Furthermore, it was difficult to reason about. With us converting the input stream to a capped reader, it is now safe to always drain the input stream when its size is known because we can never overrun the input and read into the next HTTP request. The only question is "should we?" This commit changes the draining code so every request is examined. Draining now kicks in for a few requests where it wouldn't before. But I think the code is sufficiently restricted so the behavior is safe. Possibly the most dangerous part of this code is the issuing of Connection: close for POST and PUT requests that don't have a Content-Length. I don't think there are any such uses in our WSGI application, so this should be safe. In the near future, I plan to significantly refactor the WSGI response handling. I anticipate this code evolving a bit. So any minor regressions around draining or connection closing behavior might be fixed as a result of that work. All tests pass with this change. That scares me a bit because it means we are lacking low-level tests for the HTTP protocol. Differential Revision: https://phab.mercurial-scm.org/D2769

author	Gregory Szorc <gregory.szorc@gmail.com>
date	Sat, 10 Mar 2018 11:03:45 -0800
parents	290fc4c3d1e0
children	1f7d9024674c

rev	line source
2391 d351a3be3371 Fixing up comment headers for split up code. Eric Hopper <hopper@omnifarious.org> parents: 2355 diff changeset	1 # hgweb/request.py - An http request from either CGI or the standalone server.
131 c9d51742471c moving hgweb to mercurial subdir jake@edge2.net parents: diff changeset	2 #
238 3b92f8fe47ae hgweb.py: kill #! line, clean up copyright notice mpm@selenic.com parents: 222 diff changeset	3 # Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net>
2859 345bac2bc4ec update copyrights. Vadim Gelfer <vadim.gelfer@gmail.com> parents: 2535 diff changeset	4 # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com>
131 c9d51742471c moving hgweb to mercurial subdir jake@edge2.net parents: diff changeset	5 #
8225 46293a0c7e9f updated license to be explicit about GPL version 2 Martin Geisler <mg@lazybytes.net> parents: 7742 diff changeset	6 # 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: 10261 diff changeset	7 # GNU General Public License version 2 or any later version.
131 c9d51742471c moving hgweb to mercurial subdir jake@edge2.net parents: diff changeset	8
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	9 from __future__ import absolute_import
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	10
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	11 import cgi
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	12 import errno
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	13 import socket
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	14 import wsgiref.headers as wsgiheaders
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	15 #import wsgiref.validate
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	16
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	17 from .common import (
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	18 ErrorResponse,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	19 HTTP_NOT_MODIFIED,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	20 statusmessage,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	21 )
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	22
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	23 from ..thirdparty import (
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	24 attr,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	25 )
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	26 from .. import (
34514 528b21b853aa request: coerce content-type to native str Augie Fackler <augie@google.com> parents: 34513 diff changeset	27 pycompat,
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	28 util,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	29 )
138 c77a679e9cfa Revamped templated hgweb mpm@selenic.com parents: 137 diff changeset	30
6774 0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	31 shortcuts = {
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	32 'cl': [('cmd', ['changelog']), ('rev', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	33 'sl': [('cmd', ['shortlog']), ('rev', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	34 'cs': [('cmd', ['changeset']), ('node', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	35 'f': [('cmd', ['file']), ('filenode', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	36 'fl': [('cmd', ['filelog']), ('filenode', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	37 'fd': [('cmd', ['filediff']), ('node', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	38 'fa': [('cmd', ['annotate']), ('filenode', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	39 'mf': [('cmd', ['manifest']), ('manifest', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	40 'ca': [('cmd', ['archive']), ('node', None)],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	41 'tags': [('cmd', ['tags'])],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	42 'tip': [('cmd', ['changeset']), ('node', ['tip'])],
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	43 'static': [('cmd', ['static']), ('file', None)]
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	44 }
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	45
10261 5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	46 def normalize(form):
5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	47 # first expand the shortcuts
34513 34fcb0f66837 request: use trivial iterator over dictionary keys Augie Fackler <augie@google.com> parents: 34512 diff changeset	48 for k in shortcuts:
6774 0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	49 if k in form:
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	50 for name, value in shortcuts[k]:
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	51 if value is None:
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	52 value = form[k]
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	53 form[name] = value
0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	54 del form[k]
10261 5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	55 # And strip the values
36736 2442927cdd96 hgweb: convert req.form to bytes for all keys and values Augie Fackler <augie@google.com> parents: 36291 diff changeset	56 bytesform = {}
10261 5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	57 for k, v in form.iteritems():
36736 2442927cdd96 hgweb: convert req.form to bytes for all keys and values Augie Fackler <augie@google.com> parents: 36291 diff changeset	58 bytesform[pycompat.bytesurl(k)] = [
2442927cdd96 hgweb: convert req.form to bytes for all keys and values Augie Fackler <augie@google.com> parents: 36291 diff changeset	59 pycompat.bytesurl(i.strip()) for i in v]
2442927cdd96 hgweb: convert req.form to bytes for all keys and values Augie Fackler <augie@google.com> parents: 36291 diff changeset	60 return bytesform
6774 0dbb56e90a71 hgweb: move shortcut expansion to request instantiation Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 6212 diff changeset	61
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	62 @attr.s(frozen=True)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	63 class parsedrequest(object):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	64 """Represents a parsed WSGI request / static HTTP request parameters."""
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	65
36854 16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	66 # Request method.
16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	67 method = attr.ib()
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	68 # Full URL for this request.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	69 url = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	70 # URL without any path components. Just <proto>://<host><port>.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	71 baseurl = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	72 # Advertised URL. Like ``url`` and ``baseurl`` but uses SERVER_NAME instead
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	73 # of HTTP: Host header for hostname. This is likely what clients used.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	74 advertisedurl = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	75 advertisedbaseurl = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	76 # WSGI application path.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	77 apppath = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	78 # List of path parts to be used for dispatch.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	79 dispatchparts = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	80 # URL path component (no query string) used for dispatch.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	81 dispatchpath = attr.ib()
36819 cfb9ef24968c hgweb: use parsed request to construct query parameters Gregory Szorc <gregory.szorc@gmail.com> parents: 36817 diff changeset	82 # Whether there is a path component to this request. This can be true
cfb9ef24968c hgweb: use parsed request to construct query parameters Gregory Szorc <gregory.szorc@gmail.com> parents: 36817 diff changeset	83 # when ``dispatchpath`` is empty due to REPO_NAME muckery.
cfb9ef24968c hgweb: use parsed request to construct query parameters Gregory Szorc <gregory.szorc@gmail.com> parents: 36817 diff changeset	84 havepathinfo = attr.ib()
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	85 # Raw query string (part after "?" in URL).
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	86 querystring = attr.ib()
36817 3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	87 # List of 2-tuples of query string arguments.
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	88 querystringlist = attr.ib()
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	89 # Dict of query string arguments. Values are lists with at least 1 item.
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	90 querystringdict = attr.ib()
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	91 # wsgiref.headers.Headers instance. Operates like a dict with case
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	92 # insensitive keys.
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	93 headers = attr.ib()
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	94
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	95 def parserequestfromenv(env):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	96 """Parse URL components from environment variables.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	97
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	98 WSGI defines request attributes via environment variables. This function
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	99 parses the environment variables into a data structure.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	100 """
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	101 # PEP-0333 defines the WSGI spec and is a useful reference for this code.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	102
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	103 # We first validate that the incoming object conforms with the WSGI spec.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	104 # We only want to be dealing with spec-conforming WSGI implementations.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	105 # TODO enable this once we fix internal violations.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	106 #wsgiref.validate.check_environ(env)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	107
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	108 # PEP-0333 states that environment keys and values are native strings
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	109 # (bytes on Python 2 and str on Python 3). The code points for the Unicode
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	110 # strings on Python 3 must be between \00000-\000FF. We deal with bytes
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	111 # in Mercurial, so mass convert string keys and values to bytes.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	112 if pycompat.ispy3:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	113 env = {k.encode('latin-1'): v for k, v in env.iteritems()}
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	114 env = {k: v.encode('latin-1') if isinstance(v, str) else v
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	115 for k, v in env.iteritems()}
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	116
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	117 # https://www.python.org/dev/peps/pep-0333/#environ-variables defines
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	118 # the environment variables.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	119 # https://www.python.org/dev/peps/pep-0333/#url-reconstruction defines
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	120 # how URLs are reconstructed.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	121 fullurl = env['wsgi.url_scheme'] + '://'
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	122 advertisedfullurl = fullurl
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	123
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	124 def addport(s):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	125 if env['wsgi.url_scheme'] == 'https':
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	126 if env['SERVER_PORT'] != '443':
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	127 s += ':' + env['SERVER_PORT']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	128 else:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	129 if env['SERVER_PORT'] != '80':
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	130 s += ':' + env['SERVER_PORT']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	131
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	132 return s
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	133
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	134 if env.get('HTTP_HOST'):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	135 fullurl += env['HTTP_HOST']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	136 else:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	137 fullurl += env['SERVER_NAME']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	138 fullurl = addport(fullurl)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	139
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	140 advertisedfullurl += env['SERVER_NAME']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	141 advertisedfullurl = addport(advertisedfullurl)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	142
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	143 baseurl = fullurl
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	144 advertisedbaseurl = advertisedfullurl
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	145
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	146 fullurl += util.urlreq.quote(env.get('SCRIPT_NAME', ''))
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	147 advertisedfullurl += util.urlreq.quote(env.get('SCRIPT_NAME', ''))
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	148 fullurl += util.urlreq.quote(env.get('PATH_INFO', ''))
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	149 advertisedfullurl += util.urlreq.quote(env.get('PATH_INFO', ''))
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	150
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	151 if env.get('QUERY_STRING'):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	152 fullurl += '?' + env['QUERY_STRING']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	153 advertisedfullurl += '?' + env['QUERY_STRING']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	154
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	155 # When dispatching requests, we look at the URL components (PATH_INFO
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	156 # and QUERY_STRING) after the application root (SCRIPT_NAME). But hgwebdir
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	157 # has the concept of "virtual" repositories. This is defined via REPO_NAME.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	158 # If REPO_NAME is defined, we append it to SCRIPT_NAME to form a new app
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	159 # root. We also exclude its path components from PATH_INFO when resolving
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	160 # the dispatch path.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	161
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	162 apppath = env['SCRIPT_NAME']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	163
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	164 if env.get('REPO_NAME'):
36816 0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	165 if not apppath.endswith('/'):
0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	166 apppath += '/'
0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	167
0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	168 apppath += env.get('REPO_NAME')
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	169
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	170 if 'PATH_INFO' in env:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	171 dispatchparts = env['PATH_INFO'].strip('/').split('/')
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	172
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	173 # Strip out repo parts.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	174 repoparts = env.get('REPO_NAME', '').split('/')
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	175 if dispatchparts[:len(repoparts)] == repoparts:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	176 dispatchparts = dispatchparts[len(repoparts):]
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	177 else:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	178 dispatchparts = []
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	179
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	180 dispatchpath = '/'.join(dispatchparts)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	181
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	182 querystring = env.get('QUERY_STRING', '')
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	183
36817 3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	184 # We store as a list so we have ordering information. We also store as
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	185 # a dict to facilitate fast lookup.
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	186 querystringlist = util.urlreq.parseqsl(querystring, keep_blank_values=True)
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	187
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	188 querystringdict = {}
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	189 for k, v in querystringlist:
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	190 if k in querystringdict:
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	191 querystringdict[k].append(v)
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	192 else:
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	193 querystringdict[k] = [v]
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	194
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	195 # HTTP_* keys contain HTTP request headers. The Headers structure should
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	196 # perform case normalization for us. We just rewrite underscore to dash
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	197 # so keys match what likely went over the wire.
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	198 headers = []
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	199 for k, v in env.iteritems():
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	200 if k.startswith('HTTP_'):
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	201 headers.append((k[len('HTTP_'):].replace('_', '-'), v))
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	202
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	203 headers = wsgiheaders.Headers(headers)
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	204
36853 ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	205 # This is kind of a lie because the HTTP header wasn't explicitly
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	206 # sent. But for all intents and purposes it should be OK to lie about
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	207 # this, since a consumer will either either value to determine how many
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	208 # bytes are available to read.
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	209 if 'CONTENT_LENGTH' in env and 'HTTP_CONTENT_LENGTH' not in env:
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	210 headers['Content-Length'] = env['CONTENT_LENGTH']
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	211
36854 16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	212 return parsedrequest(method=env['REQUEST_METHOD'],
16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	213 url=fullurl, baseurl=baseurl,
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	214 advertisedurl=advertisedfullurl,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	215 advertisedbaseurl=advertisedbaseurl,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	216 apppath=apppath,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	217 dispatchparts=dispatchparts, dispatchpath=dispatchpath,
36819 cfb9ef24968c hgweb: use parsed request to construct query parameters Gregory Szorc <gregory.szorc@gmail.com> parents: 36817 diff changeset	218 havepathinfo='PATH_INFO' in env,
36817 3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	219 querystring=querystring,
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	220 querystringlist=querystringlist,
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	221 querystringdict=querystringdict,
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	222 headers=headers)
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	223
5566 d74fc8dec2b4 Less indirection in the WSGI web interface. This simplifies some code, and makes it more compliant with WSGI. Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5563 diff changeset	224 class wsgirequest(object):
26132 9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	225 """Higher-level API for a WSGI request.
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	226
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	227 WSGI applications are invoked with 2 arguments. They are used to
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	228 instantiate instances of this class, which provides higher-level APIs
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	229 for obtaining request parameters, writing HTTP output, etc.
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	230 """
5566 d74fc8dec2b4 Less indirection in the WSGI web interface. This simplifies some code, and makes it more compliant with WSGI. Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5563 diff changeset	231 def __init__(self, wsgienv, start_response):
34512 482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	232 version = wsgienv[r'wsgi.version']
3673 eb0b4a2d70a9 white space and line break cleanups Thomas Arendsen Hein <thomas@intevation.de> parents: 2859 diff changeset	233 if (version < (1, 0)) or (version >= (2, 0)):
4633 ff7253a0d1da Cleanup of whitespace, indentation and line continuation. Thomas Arendsen Hein <thomas@intevation.de> parents: 4250 diff changeset	234 raise RuntimeError("Unknown and unsupported WSGI version %d.%d"
2506 d0db3462d568 This patch make several WSGI related alterations. Eric Hopper <hopper@omnifarious.org> parents: 2466 diff changeset	235 % version)
34512 482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	236 self.inp = wsgienv[r'wsgi.input']
36860 290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	237
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	238 if r'HTTP_CONTENT_LENGTH' in wsgienv:
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	239 self.inp = util.cappedreader(self.inp,
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	240 int(wsgienv[r'HTTP_CONTENT_LENGTH']))
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	241 elif r'CONTENT_LENGTH' in wsgienv:
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	242 self.inp = util.cappedreader(self.inp,
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	243 int(wsgienv[r'CONTENT_LENGTH']))
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	244
34512 482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	245 self.err = wsgienv[r'wsgi.errors']
482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	246 self.threaded = wsgienv[r'wsgi.multithread']
482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	247 self.multiprocess = wsgienv[r'wsgi.multiprocess']
482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	248 self.run_once = wsgienv[r'wsgi.run_once']
2506 d0db3462d568 This patch make several WSGI related alterations. Eric Hopper <hopper@omnifarious.org> parents: 2466 diff changeset	249 self.env = wsgienv
10261 5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	250 self.form = normalize(cgi.parse(self.inp,
5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	251 self.env,
5eae671c0b57 hgweb: request: strip() form values Nicolas Dumazet <nicdumz.commits@gmail.com> parents: 9694 diff changeset	252 keep_blank_values=1))
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	253 self._start_response = start_response
5993 948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	254 self.server_write = None
2506 d0db3462d568 This patch make several WSGI related alterations. Eric Hopper <hopper@omnifarious.org> parents: 2466 diff changeset	255 self.headers = []
d0db3462d568 This patch make several WSGI related alterations. Eric Hopper <hopper@omnifarious.org> parents: 2466 diff changeset	256
18352 e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length Mads Kiilerich <mads@kiilerich.com> parents: 18351 diff changeset	257 def respond(self, status, type, filename=None, body=None):
34514 528b21b853aa request: coerce content-type to native str Augie Fackler <augie@google.com> parents: 34513 diff changeset	258 if not isinstance(type, str):
528b21b853aa request: coerce content-type to native str Augie Fackler <augie@google.com> parents: 34513 diff changeset	259 type = pycompat.sysstr(type)
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	260 if self._start_response is not None:
34722 95be8928d6b2 hgweb: fill in content-type and content-length as native strings Augie Fackler <augie@google.com> parents: 34514 diff changeset	261 self.headers.append((r'Content-Type', type))
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	262 if filename:
26846 7c1b4840c2cd hgweb: replace some str.split() calls by str.partition() or str.rpartition() Anton Shestakov <av6@dwimlabs.net> parents: 26200 diff changeset	263 filename = (filename.rpartition('/')[-1]
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	264 .replace('\\', '\\\\').replace('"', '\\"'))
764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	265 self.headers.append(('Content-Disposition',
764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	266 'inline; filename="%s"' % filename))
18352 e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length Mads Kiilerich <mads@kiilerich.com> parents: 18351 diff changeset	267 if body is not None:
34722 95be8928d6b2 hgweb: fill in content-type and content-length as native strings Augie Fackler <augie@google.com> parents: 34514 diff changeset	268 self.headers.append((r'Content-Length', str(len(body))))
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	269
5926 15ef6b9c1f2f hgweb: be sure to send a valid content-type for raw files Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5922 diff changeset	270 for k, v in self.headers:
15ef6b9c1f2f hgweb: be sure to send a valid content-type for raw files Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5922 diff changeset	271 if not isinstance(v, str):
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	272 raise TypeError('header value must be string: %r' % (v,))
5926 15ef6b9c1f2f hgweb: be sure to send a valid content-type for raw files Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5922 diff changeset	273
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	274 if isinstance(status, ErrorResponse):
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	275 self.headers.extend(status.headers)
12739 8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	276 if status.code == HTTP_NOT_MODIFIED:
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	277 # RFC 2616 Section 10.3.5: 304 Not Modified has cases where
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	278 # it MUST NOT include any headers other than these and no
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	279 # body
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	280 self.headers = [(k, v) for (k, v) in self.headers if
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	281 k in ('Date', 'ETag', 'Expires',
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	282 'Cache-Control', 'Vary')]
36288 a0a004b29a51 hgweb: correctly bytes-ify status, not string-ify Augie Fackler <augie@google.com> parents: 34722 diff changeset	283 status = statusmessage(status.code, pycompat.bytestr(status))
5993 948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	284 elif status == 200:
948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	285 status = '200 Script output follows'
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	286 elif isinstance(status, int):
956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	287 status = statusmessage(status)
956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	288
36861 2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	289 # Various HTTP clients (notably httplib) won't read the HTTP
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	290 # response until the HTTP request has been sent in full. If servers
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	291 # (us) send a response before the HTTP request has been fully sent,
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	292 # the connection may deadlock because neither end is reading.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	293 #
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	294 # We work around this by "draining" the request data before
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	295 # sending any response in some conditions.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	296 drain = False
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	297 close = False
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	298
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	299 # If the client sent Expect: 100-continue, we assume it is smart
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	300 # enough to deal with the server sending a response before reading
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	301 # the request. (httplib doesn't do this.)
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	302 if self.env.get(r'HTTP_EXPECT', r'').lower() == r'100-continue':
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	303 pass
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	304 # Only tend to request methods that have bodies. Strictly speaking,
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	305 # we should sniff for a body. But this is fine for our existing
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	306 # WSGI applications.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	307 elif self.env[r'REQUEST_METHOD'] not in (r'POST', r'PUT'):
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	308 pass
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	309 else:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	310 # If we don't know how much data to read, there's no guarantee
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	311 # that we can drain the request responsibly. The WSGI
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	312 # specification only says that servers should ensure the
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	313 # input stream doesn't overrun the actual request. So there's
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	314 # no guarantee that reading until EOF won't corrupt the stream
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	315 # state.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	316 if not isinstance(self.inp, util.cappedreader):
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	317 close = True
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	318 else:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	319 # We /could/ only drain certain HTTP response codes. But 200
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	320 # and non-200 wire protocol responses both require draining.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	321 # Since we have a capped reader in place for all situations
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	322 # where we drain, it is safe to read from that stream. We'll
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	323 # either do a drain or no-op if we're already at EOF.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	324 drain = True
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	325
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	326 if close:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	327 self.headers.append((r'Connection', r'Close'))
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	328
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	329 if drain:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	330 assert isinstance(self.inp, util.cappedreader)
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	331 while True:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	332 chunk = self.inp.read(32768)
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	333 if not chunk:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	334 break
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	335
36291 af0a19d8812b py3: get bytes-repr of network errors portably Augie Fackler <augie@google.com> parents: 36288 diff changeset	336 self.server_write = self._start_response(
af0a19d8812b py3: get bytes-repr of network errors portably Augie Fackler <augie@google.com> parents: 36288 diff changeset	337 pycompat.sysstr(status), self.headers)
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	338 self._start_response = None
956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	339 self.headers = []
18352 e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length Mads Kiilerich <mads@kiilerich.com> parents: 18351 diff changeset	340 if body is not None:
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length Mads Kiilerich <mads@kiilerich.com> parents: 18351 diff changeset	341 self.write(body)
e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length Mads Kiilerich <mads@kiilerich.com> parents: 18351 diff changeset	342 self.server_write = None
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	343
5993 948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	344 def write(self, thing):
18351 3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	345 if thing:
3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	346 try:
3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	347 self.server_write(thing)
25660 328739ea70c3 global: mass rewrite to use modern exception syntax Gregory Szorc <gregory.szorc@gmail.com> parents: 18352 diff changeset	348 except socket.error as inst:
18351 3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	349 if inst[0] != errno.ECONNRESET:
3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	350 raise
1159 b6f5a947e62e Change use of global sys.stdout, sys.stdin os.environ to a hgrequest object. Vincent Wagelaar <vincent@ricardis.tudelft.nl> parents: 1143 diff changeset	351
4246 cc81c512a531 avoid _wsgioutputfile <-> _wsgirequest circular reference Alexis S. L. Carvalho <alexis@cecm.usp.br> parents: 3673 diff changeset	352 def flush(self):
cc81c512a531 avoid _wsgioutputfile <-> _wsgirequest circular reference Alexis S. L. Carvalho <alexis@cecm.usp.br> parents: 3673 diff changeset	353 return None
cc81c512a531 avoid _wsgioutputfile <-> _wsgirequest circular reference Alexis S. L. Carvalho <alexis@cecm.usp.br> parents: 3673 diff changeset	354
5566 d74fc8dec2b4 Less indirection in the WSGI web interface. This simplifies some code, and makes it more compliant with WSGI. Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5563 diff changeset	355 def wsgiapplication(app_maker):
5887 41a3fce17625 hgweb: return iterable, add deprecation note Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5886 diff changeset	356 '''For compatibility with old CGI scripts. A plain hgweb() or hgwebdir()
41a3fce17625 hgweb: return iterable, add deprecation note Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5886 diff changeset	357 can and should now be used as a WSGI application.'''
5760 0145f9afb0e7 Removed tabs and trailing whitespace in python files Thomas Arendsen Hein <thomas@intevation.de> parents: 5566 diff changeset	358 application = app_maker()
0145f9afb0e7 Removed tabs and trailing whitespace in python files Thomas Arendsen Hein <thomas@intevation.de> parents: 5566 diff changeset	359 def run_wsgi(env, respond):
5887 41a3fce17625 hgweb: return iterable, add deprecation note Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5886 diff changeset	360 return application(env, respond)
5760 0145f9afb0e7 Removed tabs and trailing whitespace in python files Thomas Arendsen Hein <thomas@intevation.de> parents: 5566 diff changeset	361 return run_wsgi

Mercurial > public > mercurial-scm > hg-stable

annotate mercurial/hgweb/request.py @ 36861:2cdf47e14c30