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

annotate mercurial/hgweb/request.py @ 36865:422be99519e5

hgweb: remove support for short query string based aliases (BC) Form data exposed by hgweb is post-processed to expand certain shortcuts. For example, URLs with "?cs=@" is essentially expanded to "?cmd=changeset&node=@". And the URL router treats this the same as "/changeset/@". These shortcuts were initially added in 2005 in 34cb3957d875 and 964baa35faf8. They have rarely been touched in the last decade (just moving code around a bit). We have almost no test coverage of this feature. AFAICT no templates reference URLs of this form. I even looked at the initial version of paper and coal from ~2008 and they use the "/command/params" URL form and not these shortcuts. Furthermore, I couldn't even get some shortcuts to work! For example, "?sl=@" attempts to do a revision search instead of showing shortlog starting at revision @. Maybe I'm just doing it wrong? Because this is ancient, mostly untested code, there is a migration path to something better, and because anyone passionate enough to preserve URLs can install URL redirects, let's nuke the feature. .. bc:: Query string shorts in hgweb like ``?cs=@`` have been removed. Use URLs of the form ``/:cmd`` instead. Differential Revision: https://phab.mercurial-scm.org/D2773

author	Gregory Szorc <gregory.szorc@gmail.com>
date	Fri, 09 Mar 2018 17:10:36 -0800
parents	01f6bba64424
children	a88d68dc3ee8

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 errno
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	12 import socket
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	13 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	14 #import wsgiref.validate
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	15
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	16 from .common import (
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	17 ErrorResponse,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	18 HTTP_NOT_MODIFIED,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	19 statusmessage,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	20 )
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	21
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	22 from ..thirdparty import (
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	23 attr,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	24 )
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	25 from .. import (
34514 528b21b853aa request: coerce content-type to native str Augie Fackler <augie@google.com> parents: 34513 diff changeset	26 pycompat,
27046 37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	27 util,
37fcfe52c68c hgweb: use absolute_import Yuya Nishihara <yuya@tcha.org> parents: 26846 diff changeset	28 )
138 c77a679e9cfa Revamped templated hgweb mpm@selenic.com parents: 137 diff changeset	29
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	30 @attr.s(frozen=True)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	31 class parsedrequest(object):
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	32 """Represents a parsed WSGI request.
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	33
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	34 Contains both parsed parameters as well as a handle on the input stream.
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	35 """
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	36
36854 16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	37 # Request method.
16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	38 method = attr.ib()
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	39 # Full URL for this request.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	40 url = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	41 # 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	42 baseurl = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	43 # 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	44 # 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	45 advertisedurl = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	46 advertisedbaseurl = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	47 # WSGI application path.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	48 apppath = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	49 # 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	50 dispatchparts = attr.ib()
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	51 # 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	52 dispatchpath = attr.ib()
36819 cfb9ef24968c hgweb: use parsed request to construct query parameters Gregory Szorc <gregory.szorc@gmail.com> parents: 36817 diff changeset	53 # 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	54 # 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	55 havepathinfo = attr.ib()
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	56 # 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	57 querystring = attr.ib()
36817 3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	58 # 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	59 querystringlist = attr.ib()
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	60 # 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	61 querystringdict = attr.ib()
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	62 # 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	63 # insensitive keys.
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	64 headers = attr.ib()
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	65 # Request body input stream.
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	66 bodyfh = attr.ib()
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	67
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	68 def parserequestfromenv(env, bodyfh):
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	69 """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	70
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	71 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	72 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	73 """
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	74 # 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	75
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	76 # 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	77 # 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	78 # 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	79 #wsgiref.validate.check_environ(env)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	80
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	81 # 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	82 # (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	83 # 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	84 # 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	85 if pycompat.ispy3:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	86 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	87 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	88 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	89
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	90 # 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	91 # the environment variables.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	92 # 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	93 # how URLs are reconstructed.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	94 fullurl = env['wsgi.url_scheme'] + '://'
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	95 advertisedfullurl = fullurl
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	96
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	97 def addport(s):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	98 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	99 if env['SERVER_PORT'] != '443':
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	100 s += ':' + env['SERVER_PORT']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	101 else:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	102 if env['SERVER_PORT'] != '80':
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	103 s += ':' + env['SERVER_PORT']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	104
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	105 return s
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	106
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	107 if env.get('HTTP_HOST'):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	108 fullurl += env['HTTP_HOST']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	109 else:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	110 fullurl += env['SERVER_NAME']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	111 fullurl = addport(fullurl)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	112
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	113 advertisedfullurl += env['SERVER_NAME']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	114 advertisedfullurl = addport(advertisedfullurl)
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	115
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	116 baseurl = fullurl
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	117 advertisedbaseurl = advertisedfullurl
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	118
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	119 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	120 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	121 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	122 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	123
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	124 if env.get('QUERY_STRING'):
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	125 fullurl += '?' + env['QUERY_STRING']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	126 advertisedfullurl += '?' + env['QUERY_STRING']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	127
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	128 # 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	129 # 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	130 # 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	131 # 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	132 # 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	133 # the dispatch path.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	134
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	135 apppath = env['SCRIPT_NAME']
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	136
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	137 if env.get('REPO_NAME'):
36816 0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	138 if not apppath.endswith('/'):
0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	139 apppath += '/'
0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	140
0031e972ded2 hgweb: use the parsed application path directly Gregory Szorc <gregory.szorc@gmail.com> parents: 36814 diff changeset	141 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	142
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	143 if 'PATH_INFO' in env:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	144 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	145
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	146 # Strip out repo parts.
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	147 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	148 if dispatchparts[:len(repoparts)] == repoparts:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	149 dispatchparts = dispatchparts[len(repoparts):]
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	150 else:
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	151 dispatchparts = []
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	152
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	153 dispatchpath = '/'.join(dispatchparts)
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 querystring = env.get('QUERY_STRING', '')
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	156
36817 3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	157 # 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	158 # a dict to facilitate fast lookup.
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	159 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	160
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	161 querystringdict = {}
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	162 for k, v in querystringlist:
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	163 if k in querystringdict:
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	164 querystringdict[k].append(v)
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	165 else:
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	166 querystringdict[k] = [v]
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	167
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	168 # 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	169 # 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	170 # 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	171 headers = []
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	172 for k, v in env.iteritems():
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	173 if k.startswith('HTTP_'):
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	174 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	175
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	176 headers = wsgiheaders.Headers(headers)
f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	177
36853 ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	178 # 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	179 # 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	180 # 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	181 # bytes are available to read.
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	182 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	183 headers['Content-Length'] = env['CONTENT_LENGTH']
ed0456fde625 hgweb: handle CONTENT_LENGTH Gregory Szorc <gregory.szorc@gmail.com> parents: 36822 diff changeset	184
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	185 # TODO do this once we remove wsgirequest.inp, otherwise we could have
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	186 # multiple readers from the underlying input stream.
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	187 #bodyfh = env['wsgi.input']
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	188 #if 'Content-Length' in headers:
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	189 # bodyfh = util.cappedreader(bodyfh, int(headers['Content-Length']))
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	190
36854 16292bbda39c hgweb: store and use request method on parsed request Gregory Szorc <gregory.szorc@gmail.com> parents: 36853 diff changeset	191 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	192 url=fullurl, baseurl=baseurl,
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	193 advertisedurl=advertisedfullurl,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	194 advertisedbaseurl=advertisedbaseurl,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	195 apppath=apppath,
69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	196 dispatchparts=dispatchparts, dispatchpath=dispatchpath,
36819 cfb9ef24968c hgweb: use parsed request to construct query parameters Gregory Szorc <gregory.szorc@gmail.com> parents: 36817 diff changeset	197 havepathinfo='PATH_INFO' in env,
36817 3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	198 querystring=querystring,
3c15b84ab66c hgweb: teach WSGI parser about query strings Gregory Szorc <gregory.szorc@gmail.com> parents: 36816 diff changeset	199 querystringlist=querystringlist,
36822 f9078c6caeb6 hgweb: parse and store HTTP request headers Gregory Szorc <gregory.szorc@gmail.com> parents: 36819 diff changeset	200 querystringdict=querystringdict,
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	201 headers=headers,
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	202 bodyfh=bodyfh)
36814 69b2d0900cd7 hgweb: parse WSGI request into a data structure Gregory Szorc <gregory.szorc@gmail.com> parents: 36736 diff changeset	203
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	204 class wsgirequest(object):
26132 9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	205 """Higher-level API for a WSGI request.
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	206
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	207 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	208 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	209 for obtaining request parameters, writing HTTP output, etc.
9df8c729e2e7 hgweb: add some documentation Gregory Szorc <gregory.szorc@gmail.com> parents: 25660 diff changeset	210 """
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	211 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	212 version = wsgienv[r'wsgi.version']
3673 eb0b4a2d70a9 white space and line break cleanups Thomas Arendsen Hein <thomas@intevation.de> parents: 2859 diff changeset	213 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	214 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	215 % version)
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	216
da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	217 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	218
290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	219 if r'HTTP_CONTENT_LENGTH' in wsgienv:
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	220 inp = util.cappedreader(inp, int(wsgienv[r'HTTP_CONTENT_LENGTH']))
36860 290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	221 elif r'CONTENT_LENGTH' in wsgienv:
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	222 inp = util.cappedreader(inp, int(wsgienv[r'CONTENT_LENGTH']))
36860 290fc4c3d1e0 hgweb: use a capped reader for WSGI input stream Gregory Szorc <gregory.szorc@gmail.com> parents: 36858 diff changeset	223
34512 482d6f6dba91 hgweb: when constructing or adding to a wsgi environ dict, use native strs Augie Fackler <augie@google.com> parents: 27046 diff changeset	224 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	225 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	226 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	227 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	228 self.env = wsgienv
36864 01f6bba64424 hgweb: remove support for POST form data (BC) Gregory Szorc <gregory.szorc@gmail.com> parents: 36863 diff changeset	229 self.req = parserequestfromenv(wsgienv, inp)
36865 422be99519e5 hgweb: remove support for short query string based aliases (BC) Gregory Szorc <gregory.szorc@gmail.com> parents: 36864 diff changeset	230 self.form = self.req.querystringdict
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	231 self._start_response = start_response
5993 948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	232 self.server_write = None
2506 d0db3462d568 This patch make several WSGI related alterations. Eric Hopper <hopper@omnifarious.org> parents: 2466 diff changeset	233 self.headers = []
d0db3462d568 This patch make several WSGI related alterations. Eric Hopper <hopper@omnifarious.org> parents: 2466 diff changeset	234
18352 e33b9b92a200 hgweb: pass the actual response body to request.response, not just the length Mads Kiilerich <mads@kiilerich.com> parents: 18351 diff changeset	235 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	236 if not isinstance(type, str):
528b21b853aa request: coerce content-type to native str Augie Fackler <augie@google.com> parents: 34513 diff changeset	237 type = pycompat.sysstr(type)
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	238 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	239 self.headers.append((r'Content-Type', type))
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	240 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	241 filename = (filename.rpartition('/')[-1]
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	242 .replace('\\', '\\\\').replace('"', '\\"'))
764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	243 self.headers.append(('Content-Disposition',
764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	244 '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	245 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	246 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	247
5926 15ef6b9c1f2f hgweb: be sure to send a valid content-type for raw files Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5922 diff changeset	248 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	249 if not isinstance(v, str):
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	250 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	251
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	252 if isinstance(status, ErrorResponse):
18348 764a758780b6 hgweb: simplify wsgirequest header handling Mads Kiilerich <mads@kiilerich.com> parents: 18347 diff changeset	253 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	254 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	255 # 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	256 # 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	257 # body
8dcd3203a261 hgweb: don't send a body or illegal headers during 304 response Augie Fackler <durin42@gmail.com> parents: 10263 diff changeset	258 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	259 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	260 'Cache-Control', 'Vary')]
36288 a0a004b29a51 hgweb: correctly bytes-ify status, not string-ify Augie Fackler <augie@google.com> parents: 34722 diff changeset	261 status = statusmessage(status.code, pycompat.bytestr(status))
5993 948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	262 elif status == 200:
948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	263 status = '200 Script output follows'
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	264 elif isinstance(status, int):
956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	265 status = statusmessage(status)
956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	266
36861 2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	267 # 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	268 # 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	269 # (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	270 # 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	271 #
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	272 # 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	273 # sending any response in some conditions.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	274 drain = False
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	275 close = False
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	276
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	277 # 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	278 # 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	279 # the request. (httplib doesn't do this.)
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	280 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	281 pass
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	282 # 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	283 # 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	284 # WSGI applications.
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	285 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	286 pass
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	287 else:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	288 # 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	289 # 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	290 # 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	291 # 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	292 # 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	293 # state.
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	294 if not isinstance(self.req.bodyfh, util.cappedreader):
36861 2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	295 close = True
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	296 else:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	297 # 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	298 # 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	299 # 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	300 # 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	301 # 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	302 drain = True
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	303
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	304 if close:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	305 self.headers.append((r'Connection', r'Close'))
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	306
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	307 if drain:
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	308 assert isinstance(self.req.bodyfh, util.cappedreader)
36861 2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	309 while True:
36863 da4e2f87167d hgweb: expose input stream on parsed WSGI request object Gregory Szorc <gregory.szorc@gmail.com> parents: 36862 diff changeset	310 chunk = self.req.bodyfh.read(32768)
36861 2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	311 if not chunk:
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	312 break
2cdf47e14c30 hgweb: refactor the request draining code Gregory Szorc <gregory.szorc@gmail.com> parents: 36860 diff changeset	313
36291 af0a19d8812b py3: get bytes-repr of network errors portably Augie Fackler <augie@google.com> parents: 36288 diff changeset	314 self.server_write = self._start_response(
af0a19d8812b py3: get bytes-repr of network errors portably Augie Fackler <augie@google.com> parents: 36288 diff changeset	315 pycompat.sysstr(status), self.headers)
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	316 self._start_response = None
956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	317 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	318 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	319 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	320 self.server_write = None
5888 956afc025c0f hgweb: separate out start_response() calling Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5887 diff changeset	321
5993 948a41e77902 hgweb: explicit response status Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5930 diff changeset	322 def write(self, thing):
18351 3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	323 if thing:
3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	324 try:
3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	325 self.server_write(thing)
25660 328739ea70c3 global: mass rewrite to use modern exception syntax Gregory Szorc <gregory.szorc@gmail.com> parents: 18352 diff changeset	326 except socket.error as inst:
18351 3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	327 if inst[0] != errno.ECONNRESET:
3fbdbeab38cc hgweb: don't pass empty response chunks on Mads Kiilerich <mads@kiilerich.com> parents: 18350 diff changeset	328 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	329
4246 cc81c512a531 avoid _wsgioutputfile <-> _wsgirequest circular reference Alexis S. L. Carvalho <alexis@cecm.usp.br> parents: 3673 diff changeset	330 def flush(self):
cc81c512a531 avoid _wsgioutputfile <-> _wsgirequest circular reference Alexis S. L. Carvalho <alexis@cecm.usp.br> parents: 3673 diff changeset	331 return None
cc81c512a531 avoid _wsgioutputfile <-> _wsgirequest circular reference Alexis S. L. Carvalho <alexis@cecm.usp.br> parents: 3673 diff changeset	332
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	333 def wsgiapplication(app_maker):
5887 41a3fce17625 hgweb: return iterable, add deprecation note Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5886 diff changeset	334 '''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	335 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	336 application = app_maker()
0145f9afb0e7 Removed tabs and trailing whitespace in python files Thomas Arendsen Hein <thomas@intevation.de> parents: 5566 diff changeset	337 def run_wsgi(env, respond):
5887 41a3fce17625 hgweb: return iterable, add deprecation note Dirkjan Ochtman <dirkjan@ochtman.nl> parents: 5886 diff changeset	338 return application(env, respond)
5760 0145f9afb0e7 Removed tabs and trailing whitespace in python files Thomas Arendsen Hein <thomas@intevation.de> parents: 5566 diff changeset	339 return run_wsgi

Mercurial > public > mercurial-scm > hg-stable

annotate mercurial/hgweb/request.py @ 36865:422be99519e5