Show More
@@ -1,566 +1,529 | |||||
1 | # hgweb/hgwebdir_mod.py - Web interface for a directory of repositories. |
|
1 | # hgweb/hgwebdir_mod.py - Web interface for a directory of repositories. | |
2 | # |
|
2 | # | |
3 | # Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net> |
|
3 | # Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net> | |
4 | # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com> |
|
4 | # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com> | |
5 | # |
|
5 | # | |
6 | # This software may be used and distributed according to the terms of the |
|
6 | # This software may be used and distributed according to the terms of the | |
7 | # GNU General Public License version 2 or any later version. |
|
7 | # GNU General Public License version 2 or any later version. | |
8 |
|
8 | |||
9 | from __future__ import absolute_import |
|
9 | from __future__ import absolute_import | |
10 |
|
10 | |||
11 | import os |
|
11 | import os | |
12 | import re |
|
12 | import re | |
13 | import time |
|
13 | import time | |
14 |
|
14 | |||
15 | from ..i18n import _ |
|
15 | from ..i18n import _ | |
16 |
|
16 | |||
17 | from .common import ( |
|
17 | from .common import ( | |
18 | ErrorResponse, |
|
18 | ErrorResponse, | |
19 | HTTP_NOT_FOUND, |
|
19 | HTTP_NOT_FOUND, | |
20 | HTTP_OK, |
|
20 | HTTP_OK, | |
21 | HTTP_SERVER_ERROR, |
|
21 | HTTP_SERVER_ERROR, | |
22 | cspvalues, |
|
22 | cspvalues, | |
23 | get_contact, |
|
23 | get_contact, | |
24 | get_mtime, |
|
24 | get_mtime, | |
25 | ismember, |
|
25 | ismember, | |
26 | paritygen, |
|
26 | paritygen, | |
27 | staticfile, |
|
27 | staticfile, | |
28 | ) |
|
28 | ) | |
29 |
|
29 | |||
30 | from .. import ( |
|
30 | from .. import ( | |
31 | configitems, |
|
31 | configitems, | |
32 | encoding, |
|
32 | encoding, | |
33 | error, |
|
33 | error, | |
34 | hg, |
|
34 | hg, | |
35 | profiling, |
|
35 | profiling, | |
36 | pycompat, |
|
|||
37 | scmutil, |
|
36 | scmutil, | |
38 | templater, |
|
37 | templater, | |
39 | ui as uimod, |
|
38 | ui as uimod, | |
40 | util, |
|
39 | util, | |
41 | ) |
|
40 | ) | |
42 |
|
41 | |||
43 | from . import ( |
|
42 | from . import ( | |
44 | hgweb_mod, |
|
43 | hgweb_mod, | |
45 | request as requestmod, |
|
44 | request as requestmod, | |
46 | webutil, |
|
45 | webutil, | |
47 | wsgicgi, |
|
46 | wsgicgi, | |
48 | ) |
|
47 | ) | |
49 | from ..utils import dateutil |
|
48 | from ..utils import dateutil | |
50 |
|
49 | |||
51 | def cleannames(items): |
|
50 | def cleannames(items): | |
52 | return [(util.pconvert(name).strip('/'), path) for name, path in items] |
|
51 | return [(util.pconvert(name).strip('/'), path) for name, path in items] | |
53 |
|
52 | |||
54 | def findrepos(paths): |
|
53 | def findrepos(paths): | |
55 | repos = [] |
|
54 | repos = [] | |
56 | for prefix, root in cleannames(paths): |
|
55 | for prefix, root in cleannames(paths): | |
57 | roothead, roottail = os.path.split(root) |
|
56 | roothead, roottail = os.path.split(root) | |
58 | # "foo = /bar/*" or "foo = /bar/**" lets every repo /bar/N in or below |
|
57 | # "foo = /bar/*" or "foo = /bar/**" lets every repo /bar/N in or below | |
59 | # /bar/ be served as as foo/N . |
|
58 | # /bar/ be served as as foo/N . | |
60 | # '*' will not search inside dirs with .hg (except .hg/patches), |
|
59 | # '*' will not search inside dirs with .hg (except .hg/patches), | |
61 | # '**' will search inside dirs with .hg (and thus also find subrepos). |
|
60 | # '**' will search inside dirs with .hg (and thus also find subrepos). | |
62 | try: |
|
61 | try: | |
63 | recurse = {'*': False, '**': True}[roottail] |
|
62 | recurse = {'*': False, '**': True}[roottail] | |
64 | except KeyError: |
|
63 | except KeyError: | |
65 | repos.append((prefix, root)) |
|
64 | repos.append((prefix, root)) | |
66 | continue |
|
65 | continue | |
67 | roothead = os.path.normpath(os.path.abspath(roothead)) |
|
66 | roothead = os.path.normpath(os.path.abspath(roothead)) | |
68 | paths = scmutil.walkrepos(roothead, followsym=True, recurse=recurse) |
|
67 | paths = scmutil.walkrepos(roothead, followsym=True, recurse=recurse) | |
69 | repos.extend(urlrepos(prefix, roothead, paths)) |
|
68 | repos.extend(urlrepos(prefix, roothead, paths)) | |
70 | return repos |
|
69 | return repos | |
71 |
|
70 | |||
72 | def urlrepos(prefix, roothead, paths): |
|
71 | def urlrepos(prefix, roothead, paths): | |
73 | """yield url paths and filesystem paths from a list of repo paths |
|
72 | """yield url paths and filesystem paths from a list of repo paths | |
74 |
|
73 | |||
75 | >>> conv = lambda seq: [(v, util.pconvert(p)) for v,p in seq] |
|
74 | >>> conv = lambda seq: [(v, util.pconvert(p)) for v,p in seq] | |
76 | >>> conv(urlrepos(b'hg', b'/opt', [b'/opt/r', b'/opt/r/r', b'/opt'])) |
|
75 | >>> conv(urlrepos(b'hg', b'/opt', [b'/opt/r', b'/opt/r/r', b'/opt'])) | |
77 | [('hg/r', '/opt/r'), ('hg/r/r', '/opt/r/r'), ('hg', '/opt')] |
|
76 | [('hg/r', '/opt/r'), ('hg/r/r', '/opt/r/r'), ('hg', '/opt')] | |
78 | >>> conv(urlrepos(b'', b'/opt', [b'/opt/r', b'/opt/r/r', b'/opt'])) |
|
77 | >>> conv(urlrepos(b'', b'/opt', [b'/opt/r', b'/opt/r/r', b'/opt'])) | |
79 | [('r', '/opt/r'), ('r/r', '/opt/r/r'), ('', '/opt')] |
|
78 | [('r', '/opt/r'), ('r/r', '/opt/r/r'), ('', '/opt')] | |
80 | """ |
|
79 | """ | |
81 | for path in paths: |
|
80 | for path in paths: | |
82 | path = os.path.normpath(path) |
|
81 | path = os.path.normpath(path) | |
83 | yield (prefix + '/' + |
|
82 | yield (prefix + '/' + | |
84 | util.pconvert(path[len(roothead):]).lstrip('/')).strip('/'), path |
|
83 | util.pconvert(path[len(roothead):]).lstrip('/')).strip('/'), path | |
85 |
|
84 | |||
86 | def geturlcgivars(baseurl, port): |
|
|||
87 | """ |
|
|||
88 | Extract CGI variables from baseurl |
|
|||
89 |
|
||||
90 | >>> geturlcgivars(b"http://host.org/base", b"80") |
|
|||
91 | ('host.org', '80', '/base') |
|
|||
92 | >>> geturlcgivars(b"http://host.org:8000/base", b"80") |
|
|||
93 | ('host.org', '8000', '/base') |
|
|||
94 | >>> geturlcgivars(b'/base', 8000) |
|
|||
95 | ('', '8000', '/base') |
|
|||
96 | >>> geturlcgivars(b"base", b'8000') |
|
|||
97 | ('', '8000', '/base') |
|
|||
98 | >>> geturlcgivars(b"http://host", b'8000') |
|
|||
99 | ('host', '8000', '/') |
|
|||
100 | >>> geturlcgivars(b"http://host/", b'8000') |
|
|||
101 | ('host', '8000', '/') |
|
|||
102 | """ |
|
|||
103 | u = util.url(baseurl) |
|
|||
104 | name = u.host or '' |
|
|||
105 | if u.port: |
|
|||
106 | port = u.port |
|
|||
107 | path = u.path or "" |
|
|||
108 | if not path.startswith('/'): |
|
|||
109 | path = '/' + path |
|
|||
110 |
|
||||
111 | return name, pycompat.bytestr(port), path |
|
|||
112 |
|
||||
113 | def readallowed(ui, req): |
|
85 | def readallowed(ui, req): | |
114 | """Check allow_read and deny_read config options of a repo's ui object |
|
86 | """Check allow_read and deny_read config options of a repo's ui object | |
115 | to determine user permissions. By default, with neither option set (or |
|
87 | to determine user permissions. By default, with neither option set (or | |
116 | both empty), allow all users to read the repo. There are two ways a |
|
88 | both empty), allow all users to read the repo. There are two ways a | |
117 | user can be denied read access: (1) deny_read is not empty, and the |
|
89 | user can be denied read access: (1) deny_read is not empty, and the | |
118 | user is unauthenticated or deny_read contains user (or *), and (2) |
|
90 | user is unauthenticated or deny_read contains user (or *), and (2) | |
119 | allow_read is not empty and the user is not in allow_read. Return True |
|
91 | allow_read is not empty and the user is not in allow_read. Return True | |
120 | if user is allowed to read the repo, else return False.""" |
|
92 | if user is allowed to read the repo, else return False.""" | |
121 |
|
93 | |||
122 | user = req.remoteuser |
|
94 | user = req.remoteuser | |
123 |
|
95 | |||
124 | deny_read = ui.configlist('web', 'deny_read', untrusted=True) |
|
96 | deny_read = ui.configlist('web', 'deny_read', untrusted=True) | |
125 | if deny_read and (not user or ismember(ui, user, deny_read)): |
|
97 | if deny_read and (not user or ismember(ui, user, deny_read)): | |
126 | return False |
|
98 | return False | |
127 |
|
99 | |||
128 | allow_read = ui.configlist('web', 'allow_read', untrusted=True) |
|
100 | allow_read = ui.configlist('web', 'allow_read', untrusted=True) | |
129 | # by default, allow reading if no allow_read option has been set |
|
101 | # by default, allow reading if no allow_read option has been set | |
130 | if not allow_read or ismember(ui, user, allow_read): |
|
102 | if not allow_read or ismember(ui, user, allow_read): | |
131 | return True |
|
103 | return True | |
132 |
|
104 | |||
133 | return False |
|
105 | return False | |
134 |
|
106 | |||
135 | def archivelist(ui, nodeid, url): |
|
107 | def archivelist(ui, nodeid, url): | |
136 | allowed = ui.configlist('web', 'allow_archive', untrusted=True) |
|
108 | allowed = ui.configlist('web', 'allow_archive', untrusted=True) | |
137 | archives = [] |
|
109 | archives = [] | |
138 |
|
110 | |||
139 | for typ, spec in hgweb_mod.archivespecs.iteritems(): |
|
111 | for typ, spec in hgweb_mod.archivespecs.iteritems(): | |
140 | if typ in allowed or ui.configbool('web', 'allow' + typ, |
|
112 | if typ in allowed or ui.configbool('web', 'allow' + typ, | |
141 | untrusted=True): |
|
113 | untrusted=True): | |
142 | archives.append({ |
|
114 | archives.append({ | |
143 | 'type': typ, |
|
115 | 'type': typ, | |
144 | 'extension': spec[2], |
|
116 | 'extension': spec[2], | |
145 | 'node': nodeid, |
|
117 | 'node': nodeid, | |
146 | 'url': url, |
|
118 | 'url': url, | |
147 | }) |
|
119 | }) | |
148 |
|
120 | |||
149 | return archives |
|
121 | return archives | |
150 |
|
122 | |||
151 | def rawindexentries(ui, repos, wsgireq, req, subdir=''): |
|
123 | def rawindexentries(ui, repos, wsgireq, req, subdir=''): | |
152 | descend = ui.configbool('web', 'descend') |
|
124 | descend = ui.configbool('web', 'descend') | |
153 | collapse = ui.configbool('web', 'collapse') |
|
125 | collapse = ui.configbool('web', 'collapse') | |
154 | seenrepos = set() |
|
126 | seenrepos = set() | |
155 | seendirs = set() |
|
127 | seendirs = set() | |
156 | for name, path in repos: |
|
128 | for name, path in repos: | |
157 |
|
129 | |||
158 | if not name.startswith(subdir): |
|
130 | if not name.startswith(subdir): | |
159 | continue |
|
131 | continue | |
160 | name = name[len(subdir):] |
|
132 | name = name[len(subdir):] | |
161 | directory = False |
|
133 | directory = False | |
162 |
|
134 | |||
163 | if '/' in name: |
|
135 | if '/' in name: | |
164 | if not descend: |
|
136 | if not descend: | |
165 | continue |
|
137 | continue | |
166 |
|
138 | |||
167 | nameparts = name.split('/') |
|
139 | nameparts = name.split('/') | |
168 | rootname = nameparts[0] |
|
140 | rootname = nameparts[0] | |
169 |
|
141 | |||
170 | if not collapse: |
|
142 | if not collapse: | |
171 | pass |
|
143 | pass | |
172 | elif rootname in seendirs: |
|
144 | elif rootname in seendirs: | |
173 | continue |
|
145 | continue | |
174 | elif rootname in seenrepos: |
|
146 | elif rootname in seenrepos: | |
175 | pass |
|
147 | pass | |
176 | else: |
|
148 | else: | |
177 | directory = True |
|
149 | directory = True | |
178 | name = rootname |
|
150 | name = rootname | |
179 |
|
151 | |||
180 | # redefine the path to refer to the directory |
|
152 | # redefine the path to refer to the directory | |
181 | discarded = '/'.join(nameparts[1:]) |
|
153 | discarded = '/'.join(nameparts[1:]) | |
182 |
|
154 | |||
183 | # remove name parts plus accompanying slash |
|
155 | # remove name parts plus accompanying slash | |
184 | path = path[:-len(discarded) - 1] |
|
156 | path = path[:-len(discarded) - 1] | |
185 |
|
157 | |||
186 | try: |
|
158 | try: | |
187 | r = hg.repository(ui, path) |
|
159 | r = hg.repository(ui, path) | |
188 | directory = False |
|
160 | directory = False | |
189 | except (IOError, error.RepoError): |
|
161 | except (IOError, error.RepoError): | |
190 | pass |
|
162 | pass | |
191 |
|
163 | |||
192 | parts = [name] |
|
164 | parts = [name] | |
193 | parts.insert(0, '/' + subdir.rstrip('/')) |
|
165 | parts.insert(0, '/' + subdir.rstrip('/')) | |
194 | if wsgireq.env['SCRIPT_NAME']: |
|
166 | if wsgireq.env['SCRIPT_NAME']: | |
195 | parts.insert(0, wsgireq.env['SCRIPT_NAME']) |
|
167 | parts.insert(0, wsgireq.env['SCRIPT_NAME']) | |
196 | url = re.sub(r'/+', '/', '/'.join(parts) + '/') |
|
168 | url = re.sub(r'/+', '/', '/'.join(parts) + '/') | |
197 |
|
169 | |||
198 | # show either a directory entry or a repository |
|
170 | # show either a directory entry or a repository | |
199 | if directory: |
|
171 | if directory: | |
200 | # get the directory's time information |
|
172 | # get the directory's time information | |
201 | try: |
|
173 | try: | |
202 | d = (get_mtime(path), dateutil.makedate()[1]) |
|
174 | d = (get_mtime(path), dateutil.makedate()[1]) | |
203 | except OSError: |
|
175 | except OSError: | |
204 | continue |
|
176 | continue | |
205 |
|
177 | |||
206 | # add '/' to the name to make it obvious that |
|
178 | # add '/' to the name to make it obvious that | |
207 | # the entry is a directory, not a regular repository |
|
179 | # the entry is a directory, not a regular repository | |
208 | row = {'contact': "", |
|
180 | row = {'contact': "", | |
209 | 'contact_sort': "", |
|
181 | 'contact_sort': "", | |
210 | 'name': name + '/', |
|
182 | 'name': name + '/', | |
211 | 'name_sort': name, |
|
183 | 'name_sort': name, | |
212 | 'url': url, |
|
184 | 'url': url, | |
213 | 'description': "", |
|
185 | 'description': "", | |
214 | 'description_sort': "", |
|
186 | 'description_sort': "", | |
215 | 'lastchange': d, |
|
187 | 'lastchange': d, | |
216 | 'lastchange_sort': d[1] - d[0], |
|
188 | 'lastchange_sort': d[1] - d[0], | |
217 | 'archives': [], |
|
189 | 'archives': [], | |
218 | 'isdirectory': True, |
|
190 | 'isdirectory': True, | |
219 | 'labels': [], |
|
191 | 'labels': [], | |
220 | } |
|
192 | } | |
221 |
|
193 | |||
222 | seendirs.add(name) |
|
194 | seendirs.add(name) | |
223 | yield row |
|
195 | yield row | |
224 | continue |
|
196 | continue | |
225 |
|
197 | |||
226 | u = ui.copy() |
|
198 | u = ui.copy() | |
227 | try: |
|
199 | try: | |
228 | u.readconfig(os.path.join(path, '.hg', 'hgrc')) |
|
200 | u.readconfig(os.path.join(path, '.hg', 'hgrc')) | |
229 | except Exception as e: |
|
201 | except Exception as e: | |
230 | u.warn(_('error reading %s/.hg/hgrc: %s\n') % (path, e)) |
|
202 | u.warn(_('error reading %s/.hg/hgrc: %s\n') % (path, e)) | |
231 | continue |
|
203 | continue | |
232 |
|
204 | |||
233 | def get(section, name, default=uimod._unset): |
|
205 | def get(section, name, default=uimod._unset): | |
234 | return u.config(section, name, default, untrusted=True) |
|
206 | return u.config(section, name, default, untrusted=True) | |
235 |
|
207 | |||
236 | if u.configbool("web", "hidden", untrusted=True): |
|
208 | if u.configbool("web", "hidden", untrusted=True): | |
237 | continue |
|
209 | continue | |
238 |
|
210 | |||
239 | if not readallowed(u, req): |
|
211 | if not readallowed(u, req): | |
240 | continue |
|
212 | continue | |
241 |
|
213 | |||
242 | # update time with local timezone |
|
214 | # update time with local timezone | |
243 | try: |
|
215 | try: | |
244 | r = hg.repository(ui, path) |
|
216 | r = hg.repository(ui, path) | |
245 | except IOError: |
|
217 | except IOError: | |
246 | u.warn(_('error accessing repository at %s\n') % path) |
|
218 | u.warn(_('error accessing repository at %s\n') % path) | |
247 | continue |
|
219 | continue | |
248 | except error.RepoError: |
|
220 | except error.RepoError: | |
249 | u.warn(_('error accessing repository at %s\n') % path) |
|
221 | u.warn(_('error accessing repository at %s\n') % path) | |
250 | continue |
|
222 | continue | |
251 | try: |
|
223 | try: | |
252 | d = (get_mtime(r.spath), dateutil.makedate()[1]) |
|
224 | d = (get_mtime(r.spath), dateutil.makedate()[1]) | |
253 | except OSError: |
|
225 | except OSError: | |
254 | continue |
|
226 | continue | |
255 |
|
227 | |||
256 | contact = get_contact(get) |
|
228 | contact = get_contact(get) | |
257 | description = get("web", "description") |
|
229 | description = get("web", "description") | |
258 | seenrepos.add(name) |
|
230 | seenrepos.add(name) | |
259 | name = get("web", "name", name) |
|
231 | name = get("web", "name", name) | |
260 | row = {'contact': contact or "unknown", |
|
232 | row = {'contact': contact or "unknown", | |
261 | 'contact_sort': contact.upper() or "unknown", |
|
233 | 'contact_sort': contact.upper() or "unknown", | |
262 | 'name': name, |
|
234 | 'name': name, | |
263 | 'name_sort': name, |
|
235 | 'name_sort': name, | |
264 | 'url': url, |
|
236 | 'url': url, | |
265 | 'description': description or "unknown", |
|
237 | 'description': description or "unknown", | |
266 | 'description_sort': description.upper() or "unknown", |
|
238 | 'description_sort': description.upper() or "unknown", | |
267 | 'lastchange': d, |
|
239 | 'lastchange': d, | |
268 | 'lastchange_sort': d[1] - d[0], |
|
240 | 'lastchange_sort': d[1] - d[0], | |
269 | 'archives': archivelist(u, "tip", url), |
|
241 | 'archives': archivelist(u, "tip", url), | |
270 | 'isdirectory': None, |
|
242 | 'isdirectory': None, | |
271 | 'labels': u.configlist('web', 'labels', untrusted=True), |
|
243 | 'labels': u.configlist('web', 'labels', untrusted=True), | |
272 | } |
|
244 | } | |
273 |
|
245 | |||
274 | yield row |
|
246 | yield row | |
275 |
|
247 | |||
276 | def indexentries(ui, repos, wsgireq, req, stripecount, sortcolumn='', |
|
248 | def indexentries(ui, repos, wsgireq, req, stripecount, sortcolumn='', | |
277 | descending=False, subdir=''): |
|
249 | descending=False, subdir=''): | |
278 |
|
250 | |||
279 | rows = rawindexentries(ui, repos, wsgireq, req, subdir=subdir) |
|
251 | rows = rawindexentries(ui, repos, wsgireq, req, subdir=subdir) | |
280 |
|
252 | |||
281 | sortdefault = None, False |
|
253 | sortdefault = None, False | |
282 |
|
254 | |||
283 | if sortcolumn and sortdefault != (sortcolumn, descending): |
|
255 | if sortcolumn and sortdefault != (sortcolumn, descending): | |
284 | sortkey = '%s_sort' % sortcolumn |
|
256 | sortkey = '%s_sort' % sortcolumn | |
285 | rows = sorted(rows, key=lambda x: x[sortkey], |
|
257 | rows = sorted(rows, key=lambda x: x[sortkey], | |
286 | reverse=descending) |
|
258 | reverse=descending) | |
287 |
|
259 | |||
288 | for row, parity in zip(rows, paritygen(stripecount)): |
|
260 | for row, parity in zip(rows, paritygen(stripecount)): | |
289 | row['parity'] = parity |
|
261 | row['parity'] = parity | |
290 | yield row |
|
262 | yield row | |
291 |
|
263 | |||
292 | class hgwebdir(object): |
|
264 | class hgwebdir(object): | |
293 | """HTTP server for multiple repositories. |
|
265 | """HTTP server for multiple repositories. | |
294 |
|
266 | |||
295 | Given a configuration, different repositories will be served depending |
|
267 | Given a configuration, different repositories will be served depending | |
296 | on the request path. |
|
268 | on the request path. | |
297 |
|
269 | |||
298 | Instances are typically used as WSGI applications. |
|
270 | Instances are typically used as WSGI applications. | |
299 | """ |
|
271 | """ | |
300 | def __init__(self, conf, baseui=None): |
|
272 | def __init__(self, conf, baseui=None): | |
301 | self.conf = conf |
|
273 | self.conf = conf | |
302 | self.baseui = baseui |
|
274 | self.baseui = baseui | |
303 | self.ui = None |
|
275 | self.ui = None | |
304 | self.lastrefresh = 0 |
|
276 | self.lastrefresh = 0 | |
305 | self.motd = None |
|
277 | self.motd = None | |
306 | self.refresh() |
|
278 | self.refresh() | |
307 |
|
279 | |||
308 | def refresh(self): |
|
280 | def refresh(self): | |
309 | if self.ui: |
|
281 | if self.ui: | |
310 | refreshinterval = self.ui.configint('web', 'refreshinterval') |
|
282 | refreshinterval = self.ui.configint('web', 'refreshinterval') | |
311 | else: |
|
283 | else: | |
312 | item = configitems.coreitems['web']['refreshinterval'] |
|
284 | item = configitems.coreitems['web']['refreshinterval'] | |
313 | refreshinterval = item.default |
|
285 | refreshinterval = item.default | |
314 |
|
286 | |||
315 | # refreshinterval <= 0 means to always refresh. |
|
287 | # refreshinterval <= 0 means to always refresh. | |
316 | if (refreshinterval > 0 and |
|
288 | if (refreshinterval > 0 and | |
317 | self.lastrefresh + refreshinterval > time.time()): |
|
289 | self.lastrefresh + refreshinterval > time.time()): | |
318 | return |
|
290 | return | |
319 |
|
291 | |||
320 | if self.baseui: |
|
292 | if self.baseui: | |
321 | u = self.baseui.copy() |
|
293 | u = self.baseui.copy() | |
322 | else: |
|
294 | else: | |
323 | u = uimod.ui.load() |
|
295 | u = uimod.ui.load() | |
324 | u.setconfig('ui', 'report_untrusted', 'off', 'hgwebdir') |
|
296 | u.setconfig('ui', 'report_untrusted', 'off', 'hgwebdir') | |
325 | u.setconfig('ui', 'nontty', 'true', 'hgwebdir') |
|
297 | u.setconfig('ui', 'nontty', 'true', 'hgwebdir') | |
326 | # displaying bundling progress bar while serving feels wrong and may |
|
298 | # displaying bundling progress bar while serving feels wrong and may | |
327 | # break some wsgi implementations. |
|
299 | # break some wsgi implementations. | |
328 | u.setconfig('progress', 'disable', 'true', 'hgweb') |
|
300 | u.setconfig('progress', 'disable', 'true', 'hgweb') | |
329 |
|
301 | |||
330 | if not isinstance(self.conf, (dict, list, tuple)): |
|
302 | if not isinstance(self.conf, (dict, list, tuple)): | |
331 | map = {'paths': 'hgweb-paths'} |
|
303 | map = {'paths': 'hgweb-paths'} | |
332 | if not os.path.exists(self.conf): |
|
304 | if not os.path.exists(self.conf): | |
333 | raise error.Abort(_('config file %s not found!') % self.conf) |
|
305 | raise error.Abort(_('config file %s not found!') % self.conf) | |
334 | u.readconfig(self.conf, remap=map, trust=True) |
|
306 | u.readconfig(self.conf, remap=map, trust=True) | |
335 | paths = [] |
|
307 | paths = [] | |
336 | for name, ignored in u.configitems('hgweb-paths'): |
|
308 | for name, ignored in u.configitems('hgweb-paths'): | |
337 | for path in u.configlist('hgweb-paths', name): |
|
309 | for path in u.configlist('hgweb-paths', name): | |
338 | paths.append((name, path)) |
|
310 | paths.append((name, path)) | |
339 | elif isinstance(self.conf, (list, tuple)): |
|
311 | elif isinstance(self.conf, (list, tuple)): | |
340 | paths = self.conf |
|
312 | paths = self.conf | |
341 | elif isinstance(self.conf, dict): |
|
313 | elif isinstance(self.conf, dict): | |
342 | paths = self.conf.items() |
|
314 | paths = self.conf.items() | |
343 |
|
315 | |||
344 | repos = findrepos(paths) |
|
316 | repos = findrepos(paths) | |
345 | for prefix, root in u.configitems('collections'): |
|
317 | for prefix, root in u.configitems('collections'): | |
346 | prefix = util.pconvert(prefix) |
|
318 | prefix = util.pconvert(prefix) | |
347 | for path in scmutil.walkrepos(root, followsym=True): |
|
319 | for path in scmutil.walkrepos(root, followsym=True): | |
348 | repo = os.path.normpath(path) |
|
320 | repo = os.path.normpath(path) | |
349 | name = util.pconvert(repo) |
|
321 | name = util.pconvert(repo) | |
350 | if name.startswith(prefix): |
|
322 | if name.startswith(prefix): | |
351 | name = name[len(prefix):] |
|
323 | name = name[len(prefix):] | |
352 | repos.append((name.lstrip('/'), repo)) |
|
324 | repos.append((name.lstrip('/'), repo)) | |
353 |
|
325 | |||
354 | self.repos = repos |
|
326 | self.repos = repos | |
355 | self.ui = u |
|
327 | self.ui = u | |
356 | encoding.encoding = self.ui.config('web', 'encoding') |
|
328 | encoding.encoding = self.ui.config('web', 'encoding') | |
357 | self.style = self.ui.config('web', 'style') |
|
329 | self.style = self.ui.config('web', 'style') | |
358 | self.templatepath = self.ui.config('web', 'templates', untrusted=False) |
|
330 | self.templatepath = self.ui.config('web', 'templates', untrusted=False) | |
359 | self.stripecount = self.ui.config('web', 'stripes') |
|
331 | self.stripecount = self.ui.config('web', 'stripes') | |
360 | if self.stripecount: |
|
332 | if self.stripecount: | |
361 | self.stripecount = int(self.stripecount) |
|
333 | self.stripecount = int(self.stripecount) | |
362 | self._baseurl = self.ui.config('web', 'baseurl') |
|
|||
363 | prefix = self.ui.config('web', 'prefix') |
|
334 | prefix = self.ui.config('web', 'prefix') | |
364 | if prefix.startswith('/'): |
|
335 | if prefix.startswith('/'): | |
365 | prefix = prefix[1:] |
|
336 | prefix = prefix[1:] | |
366 | if prefix.endswith('/'): |
|
337 | if prefix.endswith('/'): | |
367 | prefix = prefix[:-1] |
|
338 | prefix = prefix[:-1] | |
368 | self.prefix = prefix |
|
339 | self.prefix = prefix | |
369 | self.lastrefresh = time.time() |
|
340 | self.lastrefresh = time.time() | |
370 |
|
341 | |||
371 | def run(self): |
|
342 | def run(self): | |
372 | if not encoding.environ.get('GATEWAY_INTERFACE', |
|
343 | if not encoding.environ.get('GATEWAY_INTERFACE', | |
373 | '').startswith("CGI/1."): |
|
344 | '').startswith("CGI/1."): | |
374 | raise RuntimeError("This function is only intended to be " |
|
345 | raise RuntimeError("This function is only intended to be " | |
375 | "called while running as a CGI script.") |
|
346 | "called while running as a CGI script.") | |
376 | wsgicgi.launch(self) |
|
347 | wsgicgi.launch(self) | |
377 |
|
348 | |||
378 | def __call__(self, env, respond): |
|
349 | def __call__(self, env, respond): | |
379 | wsgireq = requestmod.wsgirequest(env, respond) |
|
350 | baseurl = self.ui.config('web', 'baseurl') | |
|
351 | wsgireq = requestmod.wsgirequest(env, respond, altbaseurl=baseurl) | |||
380 | return self.run_wsgi(wsgireq) |
|
352 | return self.run_wsgi(wsgireq) | |
381 |
|
353 | |||
382 | def run_wsgi(self, wsgireq): |
|
354 | def run_wsgi(self, wsgireq): | |
383 | profile = self.ui.configbool('profiling', 'enabled') |
|
355 | profile = self.ui.configbool('profiling', 'enabled') | |
384 | with profiling.profile(self.ui, enabled=profile): |
|
356 | with profiling.profile(self.ui, enabled=profile): | |
385 | for r in self._runwsgi(wsgireq): |
|
357 | for r in self._runwsgi(wsgireq): | |
386 | yield r |
|
358 | yield r | |
387 |
|
359 | |||
388 | def _runwsgi(self, wsgireq): |
|
360 | def _runwsgi(self, wsgireq): | |
389 | req = wsgireq.req |
|
361 | req = wsgireq.req | |
390 | res = wsgireq.res |
|
362 | res = wsgireq.res | |
391 |
|
363 | |||
392 | try: |
|
364 | try: | |
393 | self.refresh() |
|
365 | self.refresh() | |
394 |
|
366 | |||
395 | csp, nonce = cspvalues(self.ui) |
|
367 | csp, nonce = cspvalues(self.ui) | |
396 | if csp: |
|
368 | if csp: | |
397 | res.headers['Content-Security-Policy'] = csp |
|
369 | res.headers['Content-Security-Policy'] = csp | |
398 | wsgireq.headers.append(('Content-Security-Policy', csp)) |
|
370 | wsgireq.headers.append(('Content-Security-Policy', csp)) | |
399 |
|
371 | |||
400 | virtual = wsgireq.env.get("PATH_INFO", "").strip('/') |
|
372 | virtual = wsgireq.env.get("PATH_INFO", "").strip('/') | |
401 | tmpl = self.templater(wsgireq, nonce) |
|
373 | tmpl = self.templater(wsgireq, nonce) | |
402 | ctype = tmpl('mimetype', encoding=encoding.encoding) |
|
374 | ctype = tmpl('mimetype', encoding=encoding.encoding) | |
403 | ctype = templater.stringify(ctype) |
|
375 | ctype = templater.stringify(ctype) | |
404 |
|
376 | |||
405 | # Global defaults. These can be overridden by any handler. |
|
377 | # Global defaults. These can be overridden by any handler. | |
406 | res.status = '200 Script output follows' |
|
378 | res.status = '200 Script output follows' | |
407 | res.headers['Content-Type'] = ctype |
|
379 | res.headers['Content-Type'] = ctype | |
408 |
|
380 | |||
409 | # a static file |
|
381 | # a static file | |
410 | if virtual.startswith('static/') or 'static' in req.qsparams: |
|
382 | if virtual.startswith('static/') or 'static' in req.qsparams: | |
411 | if virtual.startswith('static/'): |
|
383 | if virtual.startswith('static/'): | |
412 | fname = virtual[7:] |
|
384 | fname = virtual[7:] | |
413 | else: |
|
385 | else: | |
414 | fname = req.qsparams['static'] |
|
386 | fname = req.qsparams['static'] | |
415 | static = self.ui.config("web", "static", None, |
|
387 | static = self.ui.config("web", "static", None, | |
416 | untrusted=False) |
|
388 | untrusted=False) | |
417 | if not static: |
|
389 | if not static: | |
418 | tp = self.templatepath or templater.templatepaths() |
|
390 | tp = self.templatepath or templater.templatepaths() | |
419 | if isinstance(tp, str): |
|
391 | if isinstance(tp, str): | |
420 | tp = [tp] |
|
392 | tp = [tp] | |
421 | static = [os.path.join(p, 'static') for p in tp] |
|
393 | static = [os.path.join(p, 'static') for p in tp] | |
422 |
|
394 | |||
423 | staticfile(static, fname, res) |
|
395 | staticfile(static, fname, res) | |
424 | return res.sendresponse() |
|
396 | return res.sendresponse() | |
425 |
|
397 | |||
426 | # top-level index |
|
398 | # top-level index | |
427 |
|
399 | |||
428 | repos = dict(self.repos) |
|
400 | repos = dict(self.repos) | |
429 |
|
401 | |||
430 | if (not virtual or virtual == 'index') and virtual not in repos: |
|
402 | if (not virtual or virtual == 'index') and virtual not in repos: | |
431 | wsgireq.respond(HTTP_OK, ctype) |
|
403 | wsgireq.respond(HTTP_OK, ctype) | |
432 | return self.makeindex(wsgireq, tmpl) |
|
404 | return self.makeindex(wsgireq, tmpl) | |
433 |
|
405 | |||
434 | # nested indexes and hgwebs |
|
406 | # nested indexes and hgwebs | |
435 |
|
407 | |||
436 | if virtual.endswith('/index') and virtual not in repos: |
|
408 | if virtual.endswith('/index') and virtual not in repos: | |
437 | subdir = virtual[:-len('index')] |
|
409 | subdir = virtual[:-len('index')] | |
438 | if any(r.startswith(subdir) for r in repos): |
|
410 | if any(r.startswith(subdir) for r in repos): | |
439 | wsgireq.respond(HTTP_OK, ctype) |
|
411 | wsgireq.respond(HTTP_OK, ctype) | |
440 | return self.makeindex(wsgireq, tmpl, subdir) |
|
412 | return self.makeindex(wsgireq, tmpl, subdir) | |
441 |
|
413 | |||
442 | def _virtualdirs(): |
|
414 | def _virtualdirs(): | |
443 | # Check the full virtual path, each parent, and the root ('') |
|
415 | # Check the full virtual path, each parent, and the root ('') | |
444 | if virtual != '': |
|
416 | if virtual != '': | |
445 | yield virtual |
|
417 | yield virtual | |
446 |
|
418 | |||
447 | for p in util.finddirs(virtual): |
|
419 | for p in util.finddirs(virtual): | |
448 | yield p |
|
420 | yield p | |
449 |
|
421 | |||
450 | yield '' |
|
422 | yield '' | |
451 |
|
423 | |||
452 | for virtualrepo in _virtualdirs(): |
|
424 | for virtualrepo in _virtualdirs(): | |
453 | real = repos.get(virtualrepo) |
|
425 | real = repos.get(virtualrepo) | |
454 | if real: |
|
426 | if real: | |
455 | # Re-parse the WSGI environment to take into account our |
|
427 | # Re-parse the WSGI environment to take into account our | |
456 | # repository path component. |
|
428 | # repository path component. | |
457 | wsgireq.req = requestmod.parserequestfromenv( |
|
429 | wsgireq.req = requestmod.parserequestfromenv( | |
458 |
wsgireq.env, wsgireq.req.bodyfh, reponame=virtualrepo |
|
430 | wsgireq.env, wsgireq.req.bodyfh, reponame=virtualrepo, | |
|
431 | altbaseurl=self.ui.config('web', 'baseurl')) | |||
459 | try: |
|
432 | try: | |
460 | # ensure caller gets private copy of ui |
|
433 | # ensure caller gets private copy of ui | |
461 | repo = hg.repository(self.ui.copy(), real) |
|
434 | repo = hg.repository(self.ui.copy(), real) | |
462 | return hgweb_mod.hgweb(repo).run_wsgi(wsgireq) |
|
435 | return hgweb_mod.hgweb(repo).run_wsgi(wsgireq) | |
463 | except IOError as inst: |
|
436 | except IOError as inst: | |
464 | msg = encoding.strtolocal(inst.strerror) |
|
437 | msg = encoding.strtolocal(inst.strerror) | |
465 | raise ErrorResponse(HTTP_SERVER_ERROR, msg) |
|
438 | raise ErrorResponse(HTTP_SERVER_ERROR, msg) | |
466 | except error.RepoError as inst: |
|
439 | except error.RepoError as inst: | |
467 | raise ErrorResponse(HTTP_SERVER_ERROR, bytes(inst)) |
|
440 | raise ErrorResponse(HTTP_SERVER_ERROR, bytes(inst)) | |
468 |
|
441 | |||
469 | # browse subdirectories |
|
442 | # browse subdirectories | |
470 | subdir = virtual + '/' |
|
443 | subdir = virtual + '/' | |
471 | if [r for r in repos if r.startswith(subdir)]: |
|
444 | if [r for r in repos if r.startswith(subdir)]: | |
472 | wsgireq.respond(HTTP_OK, ctype) |
|
445 | wsgireq.respond(HTTP_OK, ctype) | |
473 | return self.makeindex(wsgireq, tmpl, subdir) |
|
446 | return self.makeindex(wsgireq, tmpl, subdir) | |
474 |
|
447 | |||
475 | # prefixes not found |
|
448 | # prefixes not found | |
476 | wsgireq.respond(HTTP_NOT_FOUND, ctype) |
|
449 | wsgireq.respond(HTTP_NOT_FOUND, ctype) | |
477 | return tmpl("notfound", repo=virtual) |
|
450 | return tmpl("notfound", repo=virtual) | |
478 |
|
451 | |||
479 | except ErrorResponse as err: |
|
452 | except ErrorResponse as err: | |
480 | wsgireq.respond(err, ctype) |
|
453 | wsgireq.respond(err, ctype) | |
481 | return tmpl('error', error=err.message or '') |
|
454 | return tmpl('error', error=err.message or '') | |
482 | finally: |
|
455 | finally: | |
483 | tmpl = None |
|
456 | tmpl = None | |
484 |
|
457 | |||
485 | def makeindex(self, wsgireq, tmpl, subdir=""): |
|
458 | def makeindex(self, wsgireq, tmpl, subdir=""): | |
486 | req = wsgireq.req |
|
459 | req = wsgireq.req | |
487 |
|
460 | |||
488 | self.refresh() |
|
461 | self.refresh() | |
489 | sortable = ["name", "description", "contact", "lastchange"] |
|
462 | sortable = ["name", "description", "contact", "lastchange"] | |
490 | sortcolumn, descending = None, False |
|
463 | sortcolumn, descending = None, False | |
491 | if 'sort' in req.qsparams: |
|
464 | if 'sort' in req.qsparams: | |
492 | sortcolumn = req.qsparams['sort'] |
|
465 | sortcolumn = req.qsparams['sort'] | |
493 | descending = sortcolumn.startswith('-') |
|
466 | descending = sortcolumn.startswith('-') | |
494 | if descending: |
|
467 | if descending: | |
495 | sortcolumn = sortcolumn[1:] |
|
468 | sortcolumn = sortcolumn[1:] | |
496 | if sortcolumn not in sortable: |
|
469 | if sortcolumn not in sortable: | |
497 | sortcolumn = "" |
|
470 | sortcolumn = "" | |
498 |
|
471 | |||
499 | sort = [("sort_%s" % column, |
|
472 | sort = [("sort_%s" % column, | |
500 | "%s%s" % ((not descending and column == sortcolumn) |
|
473 | "%s%s" % ((not descending and column == sortcolumn) | |
501 | and "-" or "", column)) |
|
474 | and "-" or "", column)) | |
502 | for column in sortable] |
|
475 | for column in sortable] | |
503 |
|
476 | |||
504 | self.refresh() |
|
477 | self.refresh() | |
505 | self.updatereqenv(wsgireq.env) |
|
|||
506 |
|
478 | |||
507 | entries = indexentries(self.ui, self.repos, wsgireq, req, |
|
479 | entries = indexentries(self.ui, self.repos, wsgireq, req, | |
508 | self.stripecount, sortcolumn=sortcolumn, |
|
480 | self.stripecount, sortcolumn=sortcolumn, | |
509 | descending=descending, subdir=subdir) |
|
481 | descending=descending, subdir=subdir) | |
510 |
|
482 | |||
511 | return tmpl("index", entries=entries, subdir=subdir, |
|
483 | return tmpl("index", entries=entries, subdir=subdir, | |
512 | pathdef=hgweb_mod.makebreadcrumb('/' + subdir, self.prefix), |
|
484 | pathdef=hgweb_mod.makebreadcrumb('/' + subdir, self.prefix), | |
513 | sortcolumn=sortcolumn, descending=descending, |
|
485 | sortcolumn=sortcolumn, descending=descending, | |
514 | **dict(sort)) |
|
486 | **dict(sort)) | |
515 |
|
487 | |||
516 | def templater(self, wsgireq, nonce): |
|
488 | def templater(self, wsgireq, nonce): | |
517 |
|
489 | |||
518 | def motd(**map): |
|
490 | def motd(**map): | |
519 | if self.motd is not None: |
|
491 | if self.motd is not None: | |
520 | yield self.motd |
|
492 | yield self.motd | |
521 | else: |
|
493 | else: | |
522 | yield config('web', 'motd') |
|
494 | yield config('web', 'motd') | |
523 |
|
495 | |||
524 | def config(section, name, default=uimod._unset, untrusted=True): |
|
496 | def config(section, name, default=uimod._unset, untrusted=True): | |
525 | return self.ui.config(section, name, default, untrusted) |
|
497 | return self.ui.config(section, name, default, untrusted) | |
526 |
|
498 | |||
527 | self.updatereqenv(wsgireq.env) |
|
|||
528 |
|
||||
529 | url = wsgireq.env.get('SCRIPT_NAME', '') |
|
499 | url = wsgireq.env.get('SCRIPT_NAME', '') | |
530 | if not url.endswith('/'): |
|
500 | if not url.endswith('/'): | |
531 | url += '/' |
|
501 | url += '/' | |
532 |
|
502 | |||
533 | vars = {} |
|
503 | vars = {} | |
534 | styles, (style, mapfile) = hgweb_mod.getstyle(wsgireq.req, config, |
|
504 | styles, (style, mapfile) = hgweb_mod.getstyle(wsgireq.req, config, | |
535 | self.templatepath) |
|
505 | self.templatepath) | |
536 | if style == styles[0]: |
|
506 | if style == styles[0]: | |
537 | vars['style'] = style |
|
507 | vars['style'] = style | |
538 |
|
508 | |||
539 | sessionvars = webutil.sessionvars(vars, r'?') |
|
509 | sessionvars = webutil.sessionvars(vars, r'?') | |
540 | logourl = config('web', 'logourl') |
|
510 | logourl = config('web', 'logourl') | |
541 | logoimg = config('web', 'logoimg') |
|
511 | logoimg = config('web', 'logoimg') | |
542 | staticurl = (config('web', 'staticurl') |
|
512 | staticurl = (config('web', 'staticurl') | |
543 | or wsgireq.req.apppath + '/static/') |
|
513 | or wsgireq.req.apppath + '/static/') | |
544 | if not staticurl.endswith('/'): |
|
514 | if not staticurl.endswith('/'): | |
545 | staticurl += '/' |
|
515 | staticurl += '/' | |
546 |
|
516 | |||
547 | defaults = { |
|
517 | defaults = { | |
548 | "encoding": encoding.encoding, |
|
518 | "encoding": encoding.encoding, | |
549 | "motd": motd, |
|
519 | "motd": motd, | |
550 | "url": url, |
|
520 | "url": url, | |
551 | "logourl": logourl, |
|
521 | "logourl": logourl, | |
552 | "logoimg": logoimg, |
|
522 | "logoimg": logoimg, | |
553 | "staticurl": staticurl, |
|
523 | "staticurl": staticurl, | |
554 | "sessionvars": sessionvars, |
|
524 | "sessionvars": sessionvars, | |
555 | "style": style, |
|
525 | "style": style, | |
556 | "nonce": nonce, |
|
526 | "nonce": nonce, | |
557 | } |
|
527 | } | |
558 | tmpl = templater.templater.frommapfile(mapfile, defaults=defaults) |
|
528 | tmpl = templater.templater.frommapfile(mapfile, defaults=defaults) | |
559 | return tmpl |
|
529 | return tmpl | |
560 |
|
||||
561 | def updatereqenv(self, env): |
|
|||
562 | if self._baseurl is not None: |
|
|||
563 | name, port, path = geturlcgivars(self._baseurl, env['SERVER_PORT']) |
|
|||
564 | env['SERVER_NAME'] = name |
|
|||
565 | env['SERVER_PORT'] = port |
|
|||
566 | env['SCRIPT_NAME'] = path |
|
@@ -1,669 +1,713 | |||||
1 | # hgweb/request.py - An http request from either CGI or the standalone server. |
|
1 | # hgweb/request.py - An http request from either CGI or the standalone server. | |
2 | # |
|
2 | # | |
3 | # Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net> |
|
3 | # Copyright 21 May 2005 - (c) 2005 Jake Edge <jake@edge2.net> | |
4 | # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com> |
|
4 | # Copyright 2005, 2006 Matt Mackall <mpm@selenic.com> | |
5 | # |
|
5 | # | |
6 | # This software may be used and distributed according to the terms of the |
|
6 | # This software may be used and distributed according to the terms of the | |
7 | # GNU General Public License version 2 or any later version. |
|
7 | # GNU General Public License version 2 or any later version. | |
8 |
|
8 | |||
9 | from __future__ import absolute_import |
|
9 | from __future__ import absolute_import | |
10 |
|
10 | |||
11 | import errno |
|
11 | import errno | |
12 | import socket |
|
12 | import socket | |
13 | import wsgiref.headers as wsgiheaders |
|
13 | import wsgiref.headers as wsgiheaders | |
14 | #import wsgiref.validate |
|
14 | #import wsgiref.validate | |
15 |
|
15 | |||
16 | from .common import ( |
|
16 | from .common import ( | |
17 | ErrorResponse, |
|
17 | ErrorResponse, | |
18 | statusmessage, |
|
18 | statusmessage, | |
19 | ) |
|
19 | ) | |
20 |
|
20 | |||
21 | from ..thirdparty import ( |
|
21 | from ..thirdparty import ( | |
22 | attr, |
|
22 | attr, | |
23 | ) |
|
23 | ) | |
24 | from .. import ( |
|
24 | from .. import ( | |
25 | error, |
|
25 | error, | |
26 | pycompat, |
|
26 | pycompat, | |
27 | util, |
|
27 | util, | |
28 | ) |
|
28 | ) | |
29 |
|
29 | |||
30 | class multidict(object): |
|
30 | class multidict(object): | |
31 | """A dict like object that can store multiple values for a key. |
|
31 | """A dict like object that can store multiple values for a key. | |
32 |
|
32 | |||
33 | Used to store parsed request parameters. |
|
33 | Used to store parsed request parameters. | |
34 |
|
34 | |||
35 | This is inspired by WebOb's class of the same name. |
|
35 | This is inspired by WebOb's class of the same name. | |
36 | """ |
|
36 | """ | |
37 | def __init__(self): |
|
37 | def __init__(self): | |
38 | # Stores (key, value) 2-tuples. This isn't the most efficient. But we |
|
38 | # Stores (key, value) 2-tuples. This isn't the most efficient. But we | |
39 | # don't rely on parameters that much, so it shouldn't be a perf issue. |
|
39 | # don't rely on parameters that much, so it shouldn't be a perf issue. | |
40 | # we can always add dict for fast lookups. |
|
40 | # we can always add dict for fast lookups. | |
41 | self._items = [] |
|
41 | self._items = [] | |
42 |
|
42 | |||
43 | def __getitem__(self, key): |
|
43 | def __getitem__(self, key): | |
44 | """Returns the last set value for a key.""" |
|
44 | """Returns the last set value for a key.""" | |
45 | for k, v in reversed(self._items): |
|
45 | for k, v in reversed(self._items): | |
46 | if k == key: |
|
46 | if k == key: | |
47 | return v |
|
47 | return v | |
48 |
|
48 | |||
49 | raise KeyError(key) |
|
49 | raise KeyError(key) | |
50 |
|
50 | |||
51 | def __setitem__(self, key, value): |
|
51 | def __setitem__(self, key, value): | |
52 | """Replace a values for a key with a new value.""" |
|
52 | """Replace a values for a key with a new value.""" | |
53 | try: |
|
53 | try: | |
54 | del self[key] |
|
54 | del self[key] | |
55 | except KeyError: |
|
55 | except KeyError: | |
56 | pass |
|
56 | pass | |
57 |
|
57 | |||
58 | self._items.append((key, value)) |
|
58 | self._items.append((key, value)) | |
59 |
|
59 | |||
60 | def __delitem__(self, key): |
|
60 | def __delitem__(self, key): | |
61 | """Delete all values for a key.""" |
|
61 | """Delete all values for a key.""" | |
62 | oldlen = len(self._items) |
|
62 | oldlen = len(self._items) | |
63 |
|
63 | |||
64 | self._items[:] = [(k, v) for k, v in self._items if k != key] |
|
64 | self._items[:] = [(k, v) for k, v in self._items if k != key] | |
65 |
|
65 | |||
66 | if oldlen == len(self._items): |
|
66 | if oldlen == len(self._items): | |
67 | raise KeyError(key) |
|
67 | raise KeyError(key) | |
68 |
|
68 | |||
69 | def __contains__(self, key): |
|
69 | def __contains__(self, key): | |
70 | return any(k == key for k, v in self._items) |
|
70 | return any(k == key for k, v in self._items) | |
71 |
|
71 | |||
72 | def __len__(self): |
|
72 | def __len__(self): | |
73 | return len(self._items) |
|
73 | return len(self._items) | |
74 |
|
74 | |||
75 | def get(self, key, default=None): |
|
75 | def get(self, key, default=None): | |
76 | try: |
|
76 | try: | |
77 | return self.__getitem__(key) |
|
77 | return self.__getitem__(key) | |
78 | except KeyError: |
|
78 | except KeyError: | |
79 | return default |
|
79 | return default | |
80 |
|
80 | |||
81 | def add(self, key, value): |
|
81 | def add(self, key, value): | |
82 | """Add a new value for a key. Does not replace existing values.""" |
|
82 | """Add a new value for a key. Does not replace existing values.""" | |
83 | self._items.append((key, value)) |
|
83 | self._items.append((key, value)) | |
84 |
|
84 | |||
85 | def getall(self, key): |
|
85 | def getall(self, key): | |
86 | """Obtains all values for a key.""" |
|
86 | """Obtains all values for a key.""" | |
87 | return [v for k, v in self._items if k == key] |
|
87 | return [v for k, v in self._items if k == key] | |
88 |
|
88 | |||
89 | def getone(self, key): |
|
89 | def getone(self, key): | |
90 | """Obtain a single value for a key. |
|
90 | """Obtain a single value for a key. | |
91 |
|
91 | |||
92 | Raises KeyError if key not defined or it has multiple values set. |
|
92 | Raises KeyError if key not defined or it has multiple values set. | |
93 | """ |
|
93 | """ | |
94 | vals = self.getall(key) |
|
94 | vals = self.getall(key) | |
95 |
|
95 | |||
96 | if not vals: |
|
96 | if not vals: | |
97 | raise KeyError(key) |
|
97 | raise KeyError(key) | |
98 |
|
98 | |||
99 | if len(vals) > 1: |
|
99 | if len(vals) > 1: | |
100 | raise KeyError('multiple values for %r' % key) |
|
100 | raise KeyError('multiple values for %r' % key) | |
101 |
|
101 | |||
102 | return vals[0] |
|
102 | return vals[0] | |
103 |
|
103 | |||
104 | def asdictoflists(self): |
|
104 | def asdictoflists(self): | |
105 | d = {} |
|
105 | d = {} | |
106 | for k, v in self._items: |
|
106 | for k, v in self._items: | |
107 | if k in d: |
|
107 | if k in d: | |
108 | d[k].append(v) |
|
108 | d[k].append(v) | |
109 | else: |
|
109 | else: | |
110 | d[k] = [v] |
|
110 | d[k] = [v] | |
111 |
|
111 | |||
112 | return d |
|
112 | return d | |
113 |
|
113 | |||
114 | @attr.s(frozen=True) |
|
114 | @attr.s(frozen=True) | |
115 | class parsedrequest(object): |
|
115 | class parsedrequest(object): | |
116 | """Represents a parsed WSGI request. |
|
116 | """Represents a parsed WSGI request. | |
117 |
|
117 | |||
118 | Contains both parsed parameters as well as a handle on the input stream. |
|
118 | Contains both parsed parameters as well as a handle on the input stream. | |
119 | """ |
|
119 | """ | |
120 |
|
120 | |||
121 | # Request method. |
|
121 | # Request method. | |
122 | method = attr.ib() |
|
122 | method = attr.ib() | |
123 | # Full URL for this request. |
|
123 | # Full URL for this request. | |
124 | url = attr.ib() |
|
124 | url = attr.ib() | |
125 | # URL without any path components. Just <proto>://<host><port>. |
|
125 | # URL without any path components. Just <proto>://<host><port>. | |
126 | baseurl = attr.ib() |
|
126 | baseurl = attr.ib() | |
127 | # Advertised URL. Like ``url`` and ``baseurl`` but uses SERVER_NAME instead |
|
127 | # Advertised URL. Like ``url`` and ``baseurl`` but uses SERVER_NAME instead | |
128 | # of HTTP: Host header for hostname. This is likely what clients used. |
|
128 | # of HTTP: Host header for hostname. This is likely what clients used. | |
129 | advertisedurl = attr.ib() |
|
129 | advertisedurl = attr.ib() | |
130 | advertisedbaseurl = attr.ib() |
|
130 | advertisedbaseurl = attr.ib() | |
131 | # URL scheme (part before ``://``). e.g. ``http`` or ``https``. |
|
131 | # URL scheme (part before ``://``). e.g. ``http`` or ``https``. | |
132 | urlscheme = attr.ib() |
|
132 | urlscheme = attr.ib() | |
133 | # Value of REMOTE_USER, if set, or None. |
|
133 | # Value of REMOTE_USER, if set, or None. | |
134 | remoteuser = attr.ib() |
|
134 | remoteuser = attr.ib() | |
135 | # Value of REMOTE_HOST, if set, or None. |
|
135 | # Value of REMOTE_HOST, if set, or None. | |
136 | remotehost = attr.ib() |
|
136 | remotehost = attr.ib() | |
137 | # Relative WSGI application path. If defined, will begin with a |
|
137 | # Relative WSGI application path. If defined, will begin with a | |
138 | # ``/``. |
|
138 | # ``/``. | |
139 | apppath = attr.ib() |
|
139 | apppath = attr.ib() | |
140 | # List of path parts to be used for dispatch. |
|
140 | # List of path parts to be used for dispatch. | |
141 | dispatchparts = attr.ib() |
|
141 | dispatchparts = attr.ib() | |
142 | # URL path component (no query string) used for dispatch. Can be |
|
142 | # URL path component (no query string) used for dispatch. Can be | |
143 | # ``None`` to signal no path component given to the request, an |
|
143 | # ``None`` to signal no path component given to the request, an | |
144 | # empty string to signal a request to the application's root URL, |
|
144 | # empty string to signal a request to the application's root URL, | |
145 | # or a string not beginning with ``/`` containing the requested |
|
145 | # or a string not beginning with ``/`` containing the requested | |
146 | # path under the application. |
|
146 | # path under the application. | |
147 | dispatchpath = attr.ib() |
|
147 | dispatchpath = attr.ib() | |
148 | # The name of the repository being accessed. |
|
148 | # The name of the repository being accessed. | |
149 | reponame = attr.ib() |
|
149 | reponame = attr.ib() | |
150 | # Raw query string (part after "?" in URL). |
|
150 | # Raw query string (part after "?" in URL). | |
151 | querystring = attr.ib() |
|
151 | querystring = attr.ib() | |
152 | # multidict of query string parameters. |
|
152 | # multidict of query string parameters. | |
153 | qsparams = attr.ib() |
|
153 | qsparams = attr.ib() | |
154 | # wsgiref.headers.Headers instance. Operates like a dict with case |
|
154 | # wsgiref.headers.Headers instance. Operates like a dict with case | |
155 | # insensitive keys. |
|
155 | # insensitive keys. | |
156 | headers = attr.ib() |
|
156 | headers = attr.ib() | |
157 | # Request body input stream. |
|
157 | # Request body input stream. | |
158 | bodyfh = attr.ib() |
|
158 | bodyfh = attr.ib() | |
159 |
|
159 | |||
160 | def parserequestfromenv(env, bodyfh, reponame=None): |
|
160 | def parserequestfromenv(env, bodyfh, reponame=None, altbaseurl=None): | |
161 | """Parse URL components from environment variables. |
|
161 | """Parse URL components from environment variables. | |
162 |
|
162 | |||
163 | WSGI defines request attributes via environment variables. This function |
|
163 | WSGI defines request attributes via environment variables. This function | |
164 | parses the environment variables into a data structure. |
|
164 | parses the environment variables into a data structure. | |
165 |
|
165 | |||
166 | If ``reponame`` is defined, the leading path components matching that |
|
166 | If ``reponame`` is defined, the leading path components matching that | |
167 | string are effectively shifted from ``PATH_INFO`` to ``SCRIPT_NAME``. |
|
167 | string are effectively shifted from ``PATH_INFO`` to ``SCRIPT_NAME``. | |
168 | This simulates the world view of a WSGI application that processes |
|
168 | This simulates the world view of a WSGI application that processes | |
169 | requests from the base URL of a repo. |
|
169 | requests from the base URL of a repo. | |
|
170 | ||||
|
171 | If ``altbaseurl`` (typically comes from ``web.baseurl`` config option) | |||
|
172 | is defined, it is used - instead of the WSGI environment variables - for | |||
|
173 | constructing URL components up to and including the WSGI application path. | |||
|
174 | For example, if the current WSGI application is at ``/repo`` and a request | |||
|
175 | is made to ``/rev/@`` with this argument set to | |||
|
176 | ``http://myserver:9000/prefix``, the URL and path components will resolve as | |||
|
177 | if the request were to ``http://myserver:9000/prefix/rev/@``. In other | |||
|
178 | words, ``wsgi.url_scheme``, ``SERVER_NAME``, ``SERVER_PORT``, and | |||
|
179 | ``SCRIPT_NAME`` are all effectively replaced by components from this URL. | |||
170 | """ |
|
180 | """ | |
171 |
# PEP |
|
181 | # PEP 3333 defines the WSGI spec and is a useful reference for this code. | |
172 |
|
182 | |||
173 | # We first validate that the incoming object conforms with the WSGI spec. |
|
183 | # We first validate that the incoming object conforms with the WSGI spec. | |
174 | # We only want to be dealing with spec-conforming WSGI implementations. |
|
184 | # We only want to be dealing with spec-conforming WSGI implementations. | |
175 | # TODO enable this once we fix internal violations. |
|
185 | # TODO enable this once we fix internal violations. | |
176 | #wsgiref.validate.check_environ(env) |
|
186 | #wsgiref.validate.check_environ(env) | |
177 |
|
187 | |||
178 | # PEP-0333 states that environment keys and values are native strings |
|
188 | # PEP-0333 states that environment keys and values are native strings | |
179 | # (bytes on Python 2 and str on Python 3). The code points for the Unicode |
|
189 | # (bytes on Python 2 and str on Python 3). The code points for the Unicode | |
180 | # strings on Python 3 must be between \00000-\000FF. We deal with bytes |
|
190 | # strings on Python 3 must be between \00000-\000FF. We deal with bytes | |
181 | # in Mercurial, so mass convert string keys and values to bytes. |
|
191 | # in Mercurial, so mass convert string keys and values to bytes. | |
182 | if pycompat.ispy3: |
|
192 | if pycompat.ispy3: | |
183 | env = {k.encode('latin-1'): v for k, v in env.iteritems()} |
|
193 | env = {k.encode('latin-1'): v for k, v in env.iteritems()} | |
184 | env = {k: v.encode('latin-1') if isinstance(v, str) else v |
|
194 | env = {k: v.encode('latin-1') if isinstance(v, str) else v | |
185 | for k, v in env.iteritems()} |
|
195 | for k, v in env.iteritems()} | |
186 |
|
196 | |||
|
197 | if altbaseurl: | |||
|
198 | altbaseurl = util.url(altbaseurl) | |||
|
199 | ||||
187 | # https://www.python.org/dev/peps/pep-0333/#environ-variables defines |
|
200 | # https://www.python.org/dev/peps/pep-0333/#environ-variables defines | |
188 | # the environment variables. |
|
201 | # the environment variables. | |
189 | # https://www.python.org/dev/peps/pep-0333/#url-reconstruction defines |
|
202 | # https://www.python.org/dev/peps/pep-0333/#url-reconstruction defines | |
190 | # how URLs are reconstructed. |
|
203 | # how URLs are reconstructed. | |
191 | fullurl = env['wsgi.url_scheme'] + '://' |
|
204 | fullurl = env['wsgi.url_scheme'] + '://' | |
192 | advertisedfullurl = fullurl |
|
205 | ||
|
206 | if altbaseurl and altbaseurl.scheme: | |||
|
207 | advertisedfullurl = altbaseurl.scheme + '://' | |||
|
208 | else: | |||
|
209 | advertisedfullurl = fullurl | |||
193 |
|
210 | |||
194 | def addport(s): |
|
211 | def addport(s, port): | |
195 | if env['wsgi.url_scheme'] == 'https': |
|
212 | if s.startswith('https://'): | |
196 |
if |
|
213 | if port != '443': | |
197 |
s += ':' + |
|
214 | s += ':' + port | |
198 | else: |
|
215 | else: | |
199 |
if |
|
216 | if port != '80': | |
200 |
s += ':' + |
|
217 | s += ':' + port | |
201 |
|
218 | |||
202 | return s |
|
219 | return s | |
203 |
|
220 | |||
204 | if env.get('HTTP_HOST'): |
|
221 | if env.get('HTTP_HOST'): | |
205 | fullurl += env['HTTP_HOST'] |
|
222 | fullurl += env['HTTP_HOST'] | |
206 | else: |
|
223 | else: | |
207 | fullurl += env['SERVER_NAME'] |
|
224 | fullurl += env['SERVER_NAME'] | |
208 | fullurl = addport(fullurl) |
|
225 | fullurl = addport(fullurl, env['SERVER_PORT']) | |
|
226 | ||||
|
227 | if altbaseurl and altbaseurl.host: | |||
|
228 | advertisedfullurl += altbaseurl.host | |||
209 |
|
229 | |||
210 | advertisedfullurl += env['SERVER_NAME'] |
|
230 | if altbaseurl.port: | |
211 | advertisedfullurl = addport(advertisedfullurl) |
|
231 | port = altbaseurl.port | |
|
232 | elif altbaseurl.scheme == 'http' and not altbaseurl.port: | |||
|
233 | port = '80' | |||
|
234 | elif altbaseurl.scheme == 'https' and not altbaseurl.port: | |||
|
235 | port = '443' | |||
|
236 | else: | |||
|
237 | port = env['SERVER_PORT'] | |||
|
238 | ||||
|
239 | advertisedfullurl = addport(advertisedfullurl, port) | |||
|
240 | else: | |||
|
241 | advertisedfullurl += env['SERVER_NAME'] | |||
|
242 | advertisedfullurl = addport(advertisedfullurl, env['SERVER_PORT']) | |||
212 |
|
243 | |||
213 | baseurl = fullurl |
|
244 | baseurl = fullurl | |
214 | advertisedbaseurl = advertisedfullurl |
|
245 | advertisedbaseurl = advertisedfullurl | |
215 |
|
246 | |||
216 | fullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) |
|
247 | fullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) | |
217 | advertisedfullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) |
|
|||
218 | fullurl += util.urlreq.quote(env.get('PATH_INFO', '')) |
|
248 | fullurl += util.urlreq.quote(env.get('PATH_INFO', '')) | |
|
249 | ||||
|
250 | if altbaseurl: | |||
|
251 | path = altbaseurl.path or '' | |||
|
252 | if path and not path.startswith('/'): | |||
|
253 | path = '/' + path | |||
|
254 | advertisedfullurl += util.urlreq.quote(path) | |||
|
255 | else: | |||
|
256 | advertisedfullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) | |||
|
257 | ||||
219 | advertisedfullurl += util.urlreq.quote(env.get('PATH_INFO', '')) |
|
258 | advertisedfullurl += util.urlreq.quote(env.get('PATH_INFO', '')) | |
220 |
|
259 | |||
221 | if env.get('QUERY_STRING'): |
|
260 | if env.get('QUERY_STRING'): | |
222 | fullurl += '?' + env['QUERY_STRING'] |
|
261 | fullurl += '?' + env['QUERY_STRING'] | |
223 | advertisedfullurl += '?' + env['QUERY_STRING'] |
|
262 | advertisedfullurl += '?' + env['QUERY_STRING'] | |
224 |
|
263 | |||
225 | # If ``reponame`` is defined, that must be a prefix on PATH_INFO |
|
264 | # If ``reponame`` is defined, that must be a prefix on PATH_INFO | |
226 | # that represents the repository being dispatched to. When computing |
|
265 | # that represents the repository being dispatched to. When computing | |
227 | # the dispatch info, we ignore these leading path components. |
|
266 | # the dispatch info, we ignore these leading path components. | |
228 |
|
267 | |||
229 | apppath = env.get('SCRIPT_NAME', '') |
|
268 | if altbaseurl: | |
|
269 | apppath = altbaseurl.path or '' | |||
|
270 | if apppath and not apppath.startswith('/'): | |||
|
271 | apppath = '/' + apppath | |||
|
272 | else: | |||
|
273 | apppath = env.get('SCRIPT_NAME', '') | |||
230 |
|
274 | |||
231 | if reponame: |
|
275 | if reponame: | |
232 | repoprefix = '/' + reponame.strip('/') |
|
276 | repoprefix = '/' + reponame.strip('/') | |
233 |
|
277 | |||
234 | if not env.get('PATH_INFO'): |
|
278 | if not env.get('PATH_INFO'): | |
235 | raise error.ProgrammingError('reponame requires PATH_INFO') |
|
279 | raise error.ProgrammingError('reponame requires PATH_INFO') | |
236 |
|
280 | |||
237 | if not env['PATH_INFO'].startswith(repoprefix): |
|
281 | if not env['PATH_INFO'].startswith(repoprefix): | |
238 | raise error.ProgrammingError('PATH_INFO does not begin with repo ' |
|
282 | raise error.ProgrammingError('PATH_INFO does not begin with repo ' | |
239 | 'name: %s (%s)' % (env['PATH_INFO'], |
|
283 | 'name: %s (%s)' % (env['PATH_INFO'], | |
240 | reponame)) |
|
284 | reponame)) | |
241 |
|
285 | |||
242 | dispatchpath = env['PATH_INFO'][len(repoprefix):] |
|
286 | dispatchpath = env['PATH_INFO'][len(repoprefix):] | |
243 |
|
287 | |||
244 | if dispatchpath and not dispatchpath.startswith('/'): |
|
288 | if dispatchpath and not dispatchpath.startswith('/'): | |
245 | raise error.ProgrammingError('reponame prefix of PATH_INFO does ' |
|
289 | raise error.ProgrammingError('reponame prefix of PATH_INFO does ' | |
246 | 'not end at path delimiter: %s (%s)' % |
|
290 | 'not end at path delimiter: %s (%s)' % | |
247 | (env['PATH_INFO'], reponame)) |
|
291 | (env['PATH_INFO'], reponame)) | |
248 |
|
292 | |||
249 | apppath = apppath.rstrip('/') + repoprefix |
|
293 | apppath = apppath.rstrip('/') + repoprefix | |
250 | dispatchparts = dispatchpath.strip('/').split('/') |
|
294 | dispatchparts = dispatchpath.strip('/').split('/') | |
251 | dispatchpath = '/'.join(dispatchparts) |
|
295 | dispatchpath = '/'.join(dispatchparts) | |
252 |
|
296 | |||
253 | elif 'PATH_INFO' in env: |
|
297 | elif 'PATH_INFO' in env: | |
254 | if env['PATH_INFO'].strip('/'): |
|
298 | if env['PATH_INFO'].strip('/'): | |
255 | dispatchparts = env['PATH_INFO'].strip('/').split('/') |
|
299 | dispatchparts = env['PATH_INFO'].strip('/').split('/') | |
256 | dispatchpath = '/'.join(dispatchparts) |
|
300 | dispatchpath = '/'.join(dispatchparts) | |
257 | else: |
|
301 | else: | |
258 | dispatchparts = [] |
|
302 | dispatchparts = [] | |
259 | dispatchpath = '' |
|
303 | dispatchpath = '' | |
260 | else: |
|
304 | else: | |
261 | dispatchparts = [] |
|
305 | dispatchparts = [] | |
262 | dispatchpath = None |
|
306 | dispatchpath = None | |
263 |
|
307 | |||
264 | querystring = env.get('QUERY_STRING', '') |
|
308 | querystring = env.get('QUERY_STRING', '') | |
265 |
|
309 | |||
266 | # We store as a list so we have ordering information. We also store as |
|
310 | # We store as a list so we have ordering information. We also store as | |
267 | # a dict to facilitate fast lookup. |
|
311 | # a dict to facilitate fast lookup. | |
268 | qsparams = multidict() |
|
312 | qsparams = multidict() | |
269 | for k, v in util.urlreq.parseqsl(querystring, keep_blank_values=True): |
|
313 | for k, v in util.urlreq.parseqsl(querystring, keep_blank_values=True): | |
270 | qsparams.add(k, v) |
|
314 | qsparams.add(k, v) | |
271 |
|
315 | |||
272 | # HTTP_* keys contain HTTP request headers. The Headers structure should |
|
316 | # HTTP_* keys contain HTTP request headers. The Headers structure should | |
273 | # perform case normalization for us. We just rewrite underscore to dash |
|
317 | # perform case normalization for us. We just rewrite underscore to dash | |
274 | # so keys match what likely went over the wire. |
|
318 | # so keys match what likely went over the wire. | |
275 | headers = [] |
|
319 | headers = [] | |
276 | for k, v in env.iteritems(): |
|
320 | for k, v in env.iteritems(): | |
277 | if k.startswith('HTTP_'): |
|
321 | if k.startswith('HTTP_'): | |
278 | headers.append((k[len('HTTP_'):].replace('_', '-'), v)) |
|
322 | headers.append((k[len('HTTP_'):].replace('_', '-'), v)) | |
279 |
|
323 | |||
280 | headers = wsgiheaders.Headers(headers) |
|
324 | headers = wsgiheaders.Headers(headers) | |
281 |
|
325 | |||
282 | # This is kind of a lie because the HTTP header wasn't explicitly |
|
326 | # This is kind of a lie because the HTTP header wasn't explicitly | |
283 | # sent. But for all intents and purposes it should be OK to lie about |
|
327 | # sent. But for all intents and purposes it should be OK to lie about | |
284 | # this, since a consumer will either either value to determine how many |
|
328 | # this, since a consumer will either either value to determine how many | |
285 | # bytes are available to read. |
|
329 | # bytes are available to read. | |
286 | if 'CONTENT_LENGTH' in env and 'HTTP_CONTENT_LENGTH' not in env: |
|
330 | if 'CONTENT_LENGTH' in env and 'HTTP_CONTENT_LENGTH' not in env: | |
287 | headers['Content-Length'] = env['CONTENT_LENGTH'] |
|
331 | headers['Content-Length'] = env['CONTENT_LENGTH'] | |
288 |
|
332 | |||
289 | # TODO do this once we remove wsgirequest.inp, otherwise we could have |
|
333 | # TODO do this once we remove wsgirequest.inp, otherwise we could have | |
290 | # multiple readers from the underlying input stream. |
|
334 | # multiple readers from the underlying input stream. | |
291 | #bodyfh = env['wsgi.input'] |
|
335 | #bodyfh = env['wsgi.input'] | |
292 | #if 'Content-Length' in headers: |
|
336 | #if 'Content-Length' in headers: | |
293 | # bodyfh = util.cappedreader(bodyfh, int(headers['Content-Length'])) |
|
337 | # bodyfh = util.cappedreader(bodyfh, int(headers['Content-Length'])) | |
294 |
|
338 | |||
295 | return parsedrequest(method=env['REQUEST_METHOD'], |
|
339 | return parsedrequest(method=env['REQUEST_METHOD'], | |
296 | url=fullurl, baseurl=baseurl, |
|
340 | url=fullurl, baseurl=baseurl, | |
297 | advertisedurl=advertisedfullurl, |
|
341 | advertisedurl=advertisedfullurl, | |
298 | advertisedbaseurl=advertisedbaseurl, |
|
342 | advertisedbaseurl=advertisedbaseurl, | |
299 | urlscheme=env['wsgi.url_scheme'], |
|
343 | urlscheme=env['wsgi.url_scheme'], | |
300 | remoteuser=env.get('REMOTE_USER'), |
|
344 | remoteuser=env.get('REMOTE_USER'), | |
301 | remotehost=env.get('REMOTE_HOST'), |
|
345 | remotehost=env.get('REMOTE_HOST'), | |
302 | apppath=apppath, |
|
346 | apppath=apppath, | |
303 | dispatchparts=dispatchparts, dispatchpath=dispatchpath, |
|
347 | dispatchparts=dispatchparts, dispatchpath=dispatchpath, | |
304 | reponame=reponame, |
|
348 | reponame=reponame, | |
305 | querystring=querystring, |
|
349 | querystring=querystring, | |
306 | qsparams=qsparams, |
|
350 | qsparams=qsparams, | |
307 | headers=headers, |
|
351 | headers=headers, | |
308 | bodyfh=bodyfh) |
|
352 | bodyfh=bodyfh) | |
309 |
|
353 | |||
310 | class offsettrackingwriter(object): |
|
354 | class offsettrackingwriter(object): | |
311 | """A file object like object that is append only and tracks write count. |
|
355 | """A file object like object that is append only and tracks write count. | |
312 |
|
356 | |||
313 | Instances are bound to a callable. This callable is called with data |
|
357 | Instances are bound to a callable. This callable is called with data | |
314 | whenever a ``write()`` is attempted. |
|
358 | whenever a ``write()`` is attempted. | |
315 |
|
359 | |||
316 | Instances track the amount of written data so they can answer ``tell()`` |
|
360 | Instances track the amount of written data so they can answer ``tell()`` | |
317 | requests. |
|
361 | requests. | |
318 |
|
362 | |||
319 | The intent of this class is to wrap the ``write()`` function returned by |
|
363 | The intent of this class is to wrap the ``write()`` function returned by | |
320 | a WSGI ``start_response()`` function. Since ``write()`` is a callable and |
|
364 | a WSGI ``start_response()`` function. Since ``write()`` is a callable and | |
321 | not a file object, it doesn't implement other file object methods. |
|
365 | not a file object, it doesn't implement other file object methods. | |
322 | """ |
|
366 | """ | |
323 | def __init__(self, writefn): |
|
367 | def __init__(self, writefn): | |
324 | self._write = writefn |
|
368 | self._write = writefn | |
325 | self._offset = 0 |
|
369 | self._offset = 0 | |
326 |
|
370 | |||
327 | def write(self, s): |
|
371 | def write(self, s): | |
328 | res = self._write(s) |
|
372 | res = self._write(s) | |
329 | # Some Python objects don't report the number of bytes written. |
|
373 | # Some Python objects don't report the number of bytes written. | |
330 | if res is None: |
|
374 | if res is None: | |
331 | self._offset += len(s) |
|
375 | self._offset += len(s) | |
332 | else: |
|
376 | else: | |
333 | self._offset += res |
|
377 | self._offset += res | |
334 |
|
378 | |||
335 | def flush(self): |
|
379 | def flush(self): | |
336 | pass |
|
380 | pass | |
337 |
|
381 | |||
338 | def tell(self): |
|
382 | def tell(self): | |
339 | return self._offset |
|
383 | return self._offset | |
340 |
|
384 | |||
341 | class wsgiresponse(object): |
|
385 | class wsgiresponse(object): | |
342 | """Represents a response to a WSGI request. |
|
386 | """Represents a response to a WSGI request. | |
343 |
|
387 | |||
344 | A response consists of a status line, headers, and a body. |
|
388 | A response consists of a status line, headers, and a body. | |
345 |
|
389 | |||
346 | Consumers must populate the ``status`` and ``headers`` fields and |
|
390 | Consumers must populate the ``status`` and ``headers`` fields and | |
347 | make a call to a ``setbody*()`` method before the response can be |
|
391 | make a call to a ``setbody*()`` method before the response can be | |
348 | issued. |
|
392 | issued. | |
349 |
|
393 | |||
350 | When it is time to start sending the response over the wire, |
|
394 | When it is time to start sending the response over the wire, | |
351 | ``sendresponse()`` is called. It handles emitting the header portion |
|
395 | ``sendresponse()`` is called. It handles emitting the header portion | |
352 | of the response message. It then yields chunks of body data to be |
|
396 | of the response message. It then yields chunks of body data to be | |
353 | written to the peer. Typically, the WSGI application itself calls |
|
397 | written to the peer. Typically, the WSGI application itself calls | |
354 | and returns the value from ``sendresponse()``. |
|
398 | and returns the value from ``sendresponse()``. | |
355 | """ |
|
399 | """ | |
356 |
|
400 | |||
357 | def __init__(self, req, startresponse): |
|
401 | def __init__(self, req, startresponse): | |
358 | """Create an empty response tied to a specific request. |
|
402 | """Create an empty response tied to a specific request. | |
359 |
|
403 | |||
360 | ``req`` is a ``parsedrequest``. ``startresponse`` is the |
|
404 | ``req`` is a ``parsedrequest``. ``startresponse`` is the | |
361 | ``start_response`` function passed to the WSGI application. |
|
405 | ``start_response`` function passed to the WSGI application. | |
362 | """ |
|
406 | """ | |
363 | self._req = req |
|
407 | self._req = req | |
364 | self._startresponse = startresponse |
|
408 | self._startresponse = startresponse | |
365 |
|
409 | |||
366 | self.status = None |
|
410 | self.status = None | |
367 | self.headers = wsgiheaders.Headers([]) |
|
411 | self.headers = wsgiheaders.Headers([]) | |
368 |
|
412 | |||
369 | self._bodybytes = None |
|
413 | self._bodybytes = None | |
370 | self._bodygen = None |
|
414 | self._bodygen = None | |
371 | self._bodywillwrite = False |
|
415 | self._bodywillwrite = False | |
372 | self._started = False |
|
416 | self._started = False | |
373 | self._bodywritefn = None |
|
417 | self._bodywritefn = None | |
374 |
|
418 | |||
375 | def _verifybody(self): |
|
419 | def _verifybody(self): | |
376 | if (self._bodybytes is not None or self._bodygen is not None |
|
420 | if (self._bodybytes is not None or self._bodygen is not None | |
377 | or self._bodywillwrite): |
|
421 | or self._bodywillwrite): | |
378 | raise error.ProgrammingError('cannot define body multiple times') |
|
422 | raise error.ProgrammingError('cannot define body multiple times') | |
379 |
|
423 | |||
380 | def setbodybytes(self, b): |
|
424 | def setbodybytes(self, b): | |
381 | """Define the response body as static bytes. |
|
425 | """Define the response body as static bytes. | |
382 |
|
426 | |||
383 | The empty string signals that there is no response body. |
|
427 | The empty string signals that there is no response body. | |
384 | """ |
|
428 | """ | |
385 | self._verifybody() |
|
429 | self._verifybody() | |
386 | self._bodybytes = b |
|
430 | self._bodybytes = b | |
387 | self.headers['Content-Length'] = '%d' % len(b) |
|
431 | self.headers['Content-Length'] = '%d' % len(b) | |
388 |
|
432 | |||
389 | def setbodygen(self, gen): |
|
433 | def setbodygen(self, gen): | |
390 | """Define the response body as a generator of bytes.""" |
|
434 | """Define the response body as a generator of bytes.""" | |
391 | self._verifybody() |
|
435 | self._verifybody() | |
392 | self._bodygen = gen |
|
436 | self._bodygen = gen | |
393 |
|
437 | |||
394 | def setbodywillwrite(self): |
|
438 | def setbodywillwrite(self): | |
395 | """Signal an intent to use write() to emit the response body. |
|
439 | """Signal an intent to use write() to emit the response body. | |
396 |
|
440 | |||
397 | **This is the least preferred way to send a body.** |
|
441 | **This is the least preferred way to send a body.** | |
398 |
|
442 | |||
399 | It is preferred for WSGI applications to emit a generator of chunks |
|
443 | It is preferred for WSGI applications to emit a generator of chunks | |
400 | constituting the response body. However, some consumers can't emit |
|
444 | constituting the response body. However, some consumers can't emit | |
401 | data this way. So, WSGI provides a way to obtain a ``write(data)`` |
|
445 | data this way. So, WSGI provides a way to obtain a ``write(data)`` | |
402 | function that can be used to synchronously perform an unbuffered |
|
446 | function that can be used to synchronously perform an unbuffered | |
403 | write. |
|
447 | write. | |
404 |
|
448 | |||
405 | Calling this function signals an intent to produce the body in this |
|
449 | Calling this function signals an intent to produce the body in this | |
406 | manner. |
|
450 | manner. | |
407 | """ |
|
451 | """ | |
408 | self._verifybody() |
|
452 | self._verifybody() | |
409 | self._bodywillwrite = True |
|
453 | self._bodywillwrite = True | |
410 |
|
454 | |||
411 | def sendresponse(self): |
|
455 | def sendresponse(self): | |
412 | """Send the generated response to the client. |
|
456 | """Send the generated response to the client. | |
413 |
|
457 | |||
414 | Before this is called, ``status`` must be set and one of |
|
458 | Before this is called, ``status`` must be set and one of | |
415 | ``setbodybytes()`` or ``setbodygen()`` must be called. |
|
459 | ``setbodybytes()`` or ``setbodygen()`` must be called. | |
416 |
|
460 | |||
417 | Calling this method multiple times is not allowed. |
|
461 | Calling this method multiple times is not allowed. | |
418 | """ |
|
462 | """ | |
419 | if self._started: |
|
463 | if self._started: | |
420 | raise error.ProgrammingError('sendresponse() called multiple times') |
|
464 | raise error.ProgrammingError('sendresponse() called multiple times') | |
421 |
|
465 | |||
422 | self._started = True |
|
466 | self._started = True | |
423 |
|
467 | |||
424 | if not self.status: |
|
468 | if not self.status: | |
425 | raise error.ProgrammingError('status line not defined') |
|
469 | raise error.ProgrammingError('status line not defined') | |
426 |
|
470 | |||
427 | if (self._bodybytes is None and self._bodygen is None |
|
471 | if (self._bodybytes is None and self._bodygen is None | |
428 | and not self._bodywillwrite): |
|
472 | and not self._bodywillwrite): | |
429 | raise error.ProgrammingError('response body not defined') |
|
473 | raise error.ProgrammingError('response body not defined') | |
430 |
|
474 | |||
431 | # RFC 7232 Section 4.1 states that a 304 MUST generate one of |
|
475 | # RFC 7232 Section 4.1 states that a 304 MUST generate one of | |
432 | # {Cache-Control, Content-Location, Date, ETag, Expires, Vary} |
|
476 | # {Cache-Control, Content-Location, Date, ETag, Expires, Vary} | |
433 | # and SHOULD NOT generate other headers unless they could be used |
|
477 | # and SHOULD NOT generate other headers unless they could be used | |
434 | # to guide cache updates. Furthermore, RFC 7230 Section 3.3.2 |
|
478 | # to guide cache updates. Furthermore, RFC 7230 Section 3.3.2 | |
435 | # states that no response body can be issued. Content-Length can |
|
479 | # states that no response body can be issued. Content-Length can | |
436 | # be sent. But if it is present, it should be the size of the response |
|
480 | # be sent. But if it is present, it should be the size of the response | |
437 | # that wasn't transferred. |
|
481 | # that wasn't transferred. | |
438 | if self.status.startswith('304 '): |
|
482 | if self.status.startswith('304 '): | |
439 | # setbodybytes('') will set C-L to 0. This doesn't conform with the |
|
483 | # setbodybytes('') will set C-L to 0. This doesn't conform with the | |
440 | # spec. So remove it. |
|
484 | # spec. So remove it. | |
441 | if self.headers.get('Content-Length') == '0': |
|
485 | if self.headers.get('Content-Length') == '0': | |
442 | del self.headers['Content-Length'] |
|
486 | del self.headers['Content-Length'] | |
443 |
|
487 | |||
444 | # Strictly speaking, this is too strict. But until it causes |
|
488 | # Strictly speaking, this is too strict. But until it causes | |
445 | # problems, let's be strict. |
|
489 | # problems, let's be strict. | |
446 | badheaders = {k for k in self.headers.keys() |
|
490 | badheaders = {k for k in self.headers.keys() | |
447 | if k.lower() not in ('date', 'etag', 'expires', |
|
491 | if k.lower() not in ('date', 'etag', 'expires', | |
448 | 'cache-control', |
|
492 | 'cache-control', | |
449 | 'content-location', |
|
493 | 'content-location', | |
450 | 'vary')} |
|
494 | 'vary')} | |
451 | if badheaders: |
|
495 | if badheaders: | |
452 | raise error.ProgrammingError( |
|
496 | raise error.ProgrammingError( | |
453 | 'illegal header on 304 response: %s' % |
|
497 | 'illegal header on 304 response: %s' % | |
454 | ', '.join(sorted(badheaders))) |
|
498 | ', '.join(sorted(badheaders))) | |
455 |
|
499 | |||
456 | if self._bodygen is not None or self._bodywillwrite: |
|
500 | if self._bodygen is not None or self._bodywillwrite: | |
457 | raise error.ProgrammingError("must use setbodybytes('') with " |
|
501 | raise error.ProgrammingError("must use setbodybytes('') with " | |
458 | "304 responses") |
|
502 | "304 responses") | |
459 |
|
503 | |||
460 | # Various HTTP clients (notably httplib) won't read the HTTP response |
|
504 | # Various HTTP clients (notably httplib) won't read the HTTP response | |
461 | # until the HTTP request has been sent in full. If servers (us) send a |
|
505 | # until the HTTP request has been sent in full. If servers (us) send a | |
462 | # response before the HTTP request has been fully sent, the connection |
|
506 | # response before the HTTP request has been fully sent, the connection | |
463 | # may deadlock because neither end is reading. |
|
507 | # may deadlock because neither end is reading. | |
464 | # |
|
508 | # | |
465 | # We work around this by "draining" the request data before |
|
509 | # We work around this by "draining" the request data before | |
466 | # sending any response in some conditions. |
|
510 | # sending any response in some conditions. | |
467 | drain = False |
|
511 | drain = False | |
468 | close = False |
|
512 | close = False | |
469 |
|
513 | |||
470 | # If the client sent Expect: 100-continue, we assume it is smart enough |
|
514 | # If the client sent Expect: 100-continue, we assume it is smart enough | |
471 | # to deal with the server sending a response before reading the request. |
|
515 | # to deal with the server sending a response before reading the request. | |
472 | # (httplib doesn't do this.) |
|
516 | # (httplib doesn't do this.) | |
473 | if self._req.headers.get('Expect', '').lower() == '100-continue': |
|
517 | if self._req.headers.get('Expect', '').lower() == '100-continue': | |
474 | pass |
|
518 | pass | |
475 | # Only tend to request methods that have bodies. Strictly speaking, |
|
519 | # Only tend to request methods that have bodies. Strictly speaking, | |
476 | # we should sniff for a body. But this is fine for our existing |
|
520 | # we should sniff for a body. But this is fine for our existing | |
477 | # WSGI applications. |
|
521 | # WSGI applications. | |
478 | elif self._req.method not in ('POST', 'PUT'): |
|
522 | elif self._req.method not in ('POST', 'PUT'): | |
479 | pass |
|
523 | pass | |
480 | else: |
|
524 | else: | |
481 | # If we don't know how much data to read, there's no guarantee |
|
525 | # If we don't know how much data to read, there's no guarantee | |
482 | # that we can drain the request responsibly. The WSGI |
|
526 | # that we can drain the request responsibly. The WSGI | |
483 | # specification only says that servers *should* ensure the |
|
527 | # specification only says that servers *should* ensure the | |
484 | # input stream doesn't overrun the actual request. So there's |
|
528 | # input stream doesn't overrun the actual request. So there's | |
485 | # no guarantee that reading until EOF won't corrupt the stream |
|
529 | # no guarantee that reading until EOF won't corrupt the stream | |
486 | # state. |
|
530 | # state. | |
487 | if not isinstance(self._req.bodyfh, util.cappedreader): |
|
531 | if not isinstance(self._req.bodyfh, util.cappedreader): | |
488 | close = True |
|
532 | close = True | |
489 | else: |
|
533 | else: | |
490 | # We /could/ only drain certain HTTP response codes. But 200 and |
|
534 | # We /could/ only drain certain HTTP response codes. But 200 and | |
491 | # non-200 wire protocol responses both require draining. Since |
|
535 | # non-200 wire protocol responses both require draining. Since | |
492 | # we have a capped reader in place for all situations where we |
|
536 | # we have a capped reader in place for all situations where we | |
493 | # drain, it is safe to read from that stream. We'll either do |
|
537 | # drain, it is safe to read from that stream. We'll either do | |
494 | # a drain or no-op if we're already at EOF. |
|
538 | # a drain or no-op if we're already at EOF. | |
495 | drain = True |
|
539 | drain = True | |
496 |
|
540 | |||
497 | if close: |
|
541 | if close: | |
498 | self.headers['Connection'] = 'Close' |
|
542 | self.headers['Connection'] = 'Close' | |
499 |
|
543 | |||
500 | if drain: |
|
544 | if drain: | |
501 | assert isinstance(self._req.bodyfh, util.cappedreader) |
|
545 | assert isinstance(self._req.bodyfh, util.cappedreader) | |
502 | while True: |
|
546 | while True: | |
503 | chunk = self._req.bodyfh.read(32768) |
|
547 | chunk = self._req.bodyfh.read(32768) | |
504 | if not chunk: |
|
548 | if not chunk: | |
505 | break |
|
549 | break | |
506 |
|
550 | |||
507 | write = self._startresponse(pycompat.sysstr(self.status), |
|
551 | write = self._startresponse(pycompat.sysstr(self.status), | |
508 | self.headers.items()) |
|
552 | self.headers.items()) | |
509 |
|
553 | |||
510 | if self._bodybytes: |
|
554 | if self._bodybytes: | |
511 | yield self._bodybytes |
|
555 | yield self._bodybytes | |
512 | elif self._bodygen: |
|
556 | elif self._bodygen: | |
513 | for chunk in self._bodygen: |
|
557 | for chunk in self._bodygen: | |
514 | yield chunk |
|
558 | yield chunk | |
515 | elif self._bodywillwrite: |
|
559 | elif self._bodywillwrite: | |
516 | self._bodywritefn = write |
|
560 | self._bodywritefn = write | |
517 | else: |
|
561 | else: | |
518 | error.ProgrammingError('do not know how to send body') |
|
562 | error.ProgrammingError('do not know how to send body') | |
519 |
|
563 | |||
520 | def getbodyfile(self): |
|
564 | def getbodyfile(self): | |
521 | """Obtain a file object like object representing the response body. |
|
565 | """Obtain a file object like object representing the response body. | |
522 |
|
566 | |||
523 | For this to work, you must call ``setbodywillwrite()`` and then |
|
567 | For this to work, you must call ``setbodywillwrite()`` and then | |
524 | ``sendresponse()`` first. ``sendresponse()`` is a generator and the |
|
568 | ``sendresponse()`` first. ``sendresponse()`` is a generator and the | |
525 | function won't run to completion unless the generator is advanced. The |
|
569 | function won't run to completion unless the generator is advanced. The | |
526 | generator yields not items. The easiest way to consume it is with |
|
570 | generator yields not items. The easiest way to consume it is with | |
527 | ``list(res.sendresponse())``, which should resolve to an empty list - |
|
571 | ``list(res.sendresponse())``, which should resolve to an empty list - | |
528 | ``[]``. |
|
572 | ``[]``. | |
529 | """ |
|
573 | """ | |
530 | if not self._bodywillwrite: |
|
574 | if not self._bodywillwrite: | |
531 | raise error.ProgrammingError('must call setbodywillwrite() first') |
|
575 | raise error.ProgrammingError('must call setbodywillwrite() first') | |
532 |
|
576 | |||
533 | if not self._started: |
|
577 | if not self._started: | |
534 | raise error.ProgrammingError('must call sendresponse() first; did ' |
|
578 | raise error.ProgrammingError('must call sendresponse() first; did ' | |
535 | 'you remember to consume it since it ' |
|
579 | 'you remember to consume it since it ' | |
536 | 'is a generator?') |
|
580 | 'is a generator?') | |
537 |
|
581 | |||
538 | assert self._bodywritefn |
|
582 | assert self._bodywritefn | |
539 | return offsettrackingwriter(self._bodywritefn) |
|
583 | return offsettrackingwriter(self._bodywritefn) | |
540 |
|
584 | |||
541 | class wsgirequest(object): |
|
585 | class wsgirequest(object): | |
542 | """Higher-level API for a WSGI request. |
|
586 | """Higher-level API for a WSGI request. | |
543 |
|
587 | |||
544 | WSGI applications are invoked with 2 arguments. They are used to |
|
588 | WSGI applications are invoked with 2 arguments. They are used to | |
545 | instantiate instances of this class, which provides higher-level APIs |
|
589 | instantiate instances of this class, which provides higher-level APIs | |
546 | for obtaining request parameters, writing HTTP output, etc. |
|
590 | for obtaining request parameters, writing HTTP output, etc. | |
547 | """ |
|
591 | """ | |
548 | def __init__(self, wsgienv, start_response): |
|
592 | def __init__(self, wsgienv, start_response, altbaseurl=None): | |
549 | version = wsgienv[r'wsgi.version'] |
|
593 | version = wsgienv[r'wsgi.version'] | |
550 | if (version < (1, 0)) or (version >= (2, 0)): |
|
594 | if (version < (1, 0)) or (version >= (2, 0)): | |
551 | raise RuntimeError("Unknown and unsupported WSGI version %d.%d" |
|
595 | raise RuntimeError("Unknown and unsupported WSGI version %d.%d" | |
552 | % version) |
|
596 | % version) | |
553 |
|
597 | |||
554 | inp = wsgienv[r'wsgi.input'] |
|
598 | inp = wsgienv[r'wsgi.input'] | |
555 |
|
599 | |||
556 | if r'HTTP_CONTENT_LENGTH' in wsgienv: |
|
600 | if r'HTTP_CONTENT_LENGTH' in wsgienv: | |
557 | inp = util.cappedreader(inp, int(wsgienv[r'HTTP_CONTENT_LENGTH'])) |
|
601 | inp = util.cappedreader(inp, int(wsgienv[r'HTTP_CONTENT_LENGTH'])) | |
558 | elif r'CONTENT_LENGTH' in wsgienv: |
|
602 | elif r'CONTENT_LENGTH' in wsgienv: | |
559 | inp = util.cappedreader(inp, int(wsgienv[r'CONTENT_LENGTH'])) |
|
603 | inp = util.cappedreader(inp, int(wsgienv[r'CONTENT_LENGTH'])) | |
560 |
|
604 | |||
561 | self.err = wsgienv[r'wsgi.errors'] |
|
605 | self.err = wsgienv[r'wsgi.errors'] | |
562 | self.threaded = wsgienv[r'wsgi.multithread'] |
|
606 | self.threaded = wsgienv[r'wsgi.multithread'] | |
563 | self.multiprocess = wsgienv[r'wsgi.multiprocess'] |
|
607 | self.multiprocess = wsgienv[r'wsgi.multiprocess'] | |
564 | self.run_once = wsgienv[r'wsgi.run_once'] |
|
608 | self.run_once = wsgienv[r'wsgi.run_once'] | |
565 | self.env = wsgienv |
|
609 | self.env = wsgienv | |
566 | self.req = parserequestfromenv(wsgienv, inp) |
|
610 | self.req = parserequestfromenv(wsgienv, inp, altbaseurl=altbaseurl) | |
567 | self.res = wsgiresponse(self.req, start_response) |
|
611 | self.res = wsgiresponse(self.req, start_response) | |
568 | self._start_response = start_response |
|
612 | self._start_response = start_response | |
569 | self.server_write = None |
|
613 | self.server_write = None | |
570 | self.headers = [] |
|
614 | self.headers = [] | |
571 |
|
615 | |||
572 | def respond(self, status, type, filename=None, body=None): |
|
616 | def respond(self, status, type, filename=None, body=None): | |
573 | if not isinstance(type, str): |
|
617 | if not isinstance(type, str): | |
574 | type = pycompat.sysstr(type) |
|
618 | type = pycompat.sysstr(type) | |
575 | if self._start_response is not None: |
|
619 | if self._start_response is not None: | |
576 | self.headers.append((r'Content-Type', type)) |
|
620 | self.headers.append((r'Content-Type', type)) | |
577 | if filename: |
|
621 | if filename: | |
578 | filename = (filename.rpartition('/')[-1] |
|
622 | filename = (filename.rpartition('/')[-1] | |
579 | .replace('\\', '\\\\').replace('"', '\\"')) |
|
623 | .replace('\\', '\\\\').replace('"', '\\"')) | |
580 | self.headers.append(('Content-Disposition', |
|
624 | self.headers.append(('Content-Disposition', | |
581 | 'inline; filename="%s"' % filename)) |
|
625 | 'inline; filename="%s"' % filename)) | |
582 | if body is not None: |
|
626 | if body is not None: | |
583 | self.headers.append((r'Content-Length', str(len(body)))) |
|
627 | self.headers.append((r'Content-Length', str(len(body)))) | |
584 |
|
628 | |||
585 | for k, v in self.headers: |
|
629 | for k, v in self.headers: | |
586 | if not isinstance(v, str): |
|
630 | if not isinstance(v, str): | |
587 | raise TypeError('header value must be string: %r' % (v,)) |
|
631 | raise TypeError('header value must be string: %r' % (v,)) | |
588 |
|
632 | |||
589 | if isinstance(status, ErrorResponse): |
|
633 | if isinstance(status, ErrorResponse): | |
590 | self.headers.extend(status.headers) |
|
634 | self.headers.extend(status.headers) | |
591 | status = statusmessage(status.code, pycompat.bytestr(status)) |
|
635 | status = statusmessage(status.code, pycompat.bytestr(status)) | |
592 | elif status == 200: |
|
636 | elif status == 200: | |
593 | status = '200 Script output follows' |
|
637 | status = '200 Script output follows' | |
594 | elif isinstance(status, int): |
|
638 | elif isinstance(status, int): | |
595 | status = statusmessage(status) |
|
639 | status = statusmessage(status) | |
596 |
|
640 | |||
597 | # Various HTTP clients (notably httplib) won't read the HTTP |
|
641 | # Various HTTP clients (notably httplib) won't read the HTTP | |
598 | # response until the HTTP request has been sent in full. If servers |
|
642 | # response until the HTTP request has been sent in full. If servers | |
599 | # (us) send a response before the HTTP request has been fully sent, |
|
643 | # (us) send a response before the HTTP request has been fully sent, | |
600 | # the connection may deadlock because neither end is reading. |
|
644 | # the connection may deadlock because neither end is reading. | |
601 | # |
|
645 | # | |
602 | # We work around this by "draining" the request data before |
|
646 | # We work around this by "draining" the request data before | |
603 | # sending any response in some conditions. |
|
647 | # sending any response in some conditions. | |
604 | drain = False |
|
648 | drain = False | |
605 | close = False |
|
649 | close = False | |
606 |
|
650 | |||
607 | # If the client sent Expect: 100-continue, we assume it is smart |
|
651 | # If the client sent Expect: 100-continue, we assume it is smart | |
608 | # enough to deal with the server sending a response before reading |
|
652 | # enough to deal with the server sending a response before reading | |
609 | # the request. (httplib doesn't do this.) |
|
653 | # the request. (httplib doesn't do this.) | |
610 | if self.env.get(r'HTTP_EXPECT', r'').lower() == r'100-continue': |
|
654 | if self.env.get(r'HTTP_EXPECT', r'').lower() == r'100-continue': | |
611 | pass |
|
655 | pass | |
612 | # Only tend to request methods that have bodies. Strictly speaking, |
|
656 | # Only tend to request methods that have bodies. Strictly speaking, | |
613 | # we should sniff for a body. But this is fine for our existing |
|
657 | # we should sniff for a body. But this is fine for our existing | |
614 | # WSGI applications. |
|
658 | # WSGI applications. | |
615 | elif self.env[r'REQUEST_METHOD'] not in (r'POST', r'PUT'): |
|
659 | elif self.env[r'REQUEST_METHOD'] not in (r'POST', r'PUT'): | |
616 | pass |
|
660 | pass | |
617 | else: |
|
661 | else: | |
618 | # If we don't know how much data to read, there's no guarantee |
|
662 | # If we don't know how much data to read, there's no guarantee | |
619 | # that we can drain the request responsibly. The WSGI |
|
663 | # that we can drain the request responsibly. The WSGI | |
620 | # specification only says that servers *should* ensure the |
|
664 | # specification only says that servers *should* ensure the | |
621 | # input stream doesn't overrun the actual request. So there's |
|
665 | # input stream doesn't overrun the actual request. So there's | |
622 | # no guarantee that reading until EOF won't corrupt the stream |
|
666 | # no guarantee that reading until EOF won't corrupt the stream | |
623 | # state. |
|
667 | # state. | |
624 | if not isinstance(self.req.bodyfh, util.cappedreader): |
|
668 | if not isinstance(self.req.bodyfh, util.cappedreader): | |
625 | close = True |
|
669 | close = True | |
626 | else: |
|
670 | else: | |
627 | # We /could/ only drain certain HTTP response codes. But 200 |
|
671 | # We /could/ only drain certain HTTP response codes. But 200 | |
628 | # and non-200 wire protocol responses both require draining. |
|
672 | # and non-200 wire protocol responses both require draining. | |
629 | # Since we have a capped reader in place for all situations |
|
673 | # Since we have a capped reader in place for all situations | |
630 | # where we drain, it is safe to read from that stream. We'll |
|
674 | # where we drain, it is safe to read from that stream. We'll | |
631 | # either do a drain or no-op if we're already at EOF. |
|
675 | # either do a drain or no-op if we're already at EOF. | |
632 | drain = True |
|
676 | drain = True | |
633 |
|
677 | |||
634 | if close: |
|
678 | if close: | |
635 | self.headers.append((r'Connection', r'Close')) |
|
679 | self.headers.append((r'Connection', r'Close')) | |
636 |
|
680 | |||
637 | if drain: |
|
681 | if drain: | |
638 | assert isinstance(self.req.bodyfh, util.cappedreader) |
|
682 | assert isinstance(self.req.bodyfh, util.cappedreader) | |
639 | while True: |
|
683 | while True: | |
640 | chunk = self.req.bodyfh.read(32768) |
|
684 | chunk = self.req.bodyfh.read(32768) | |
641 | if not chunk: |
|
685 | if not chunk: | |
642 | break |
|
686 | break | |
643 |
|
687 | |||
644 | self.server_write = self._start_response( |
|
688 | self.server_write = self._start_response( | |
645 | pycompat.sysstr(status), self.headers) |
|
689 | pycompat.sysstr(status), self.headers) | |
646 | self._start_response = None |
|
690 | self._start_response = None | |
647 | self.headers = [] |
|
691 | self.headers = [] | |
648 | if body is not None: |
|
692 | if body is not None: | |
649 | self.write(body) |
|
693 | self.write(body) | |
650 | self.server_write = None |
|
694 | self.server_write = None | |
651 |
|
695 | |||
652 | def write(self, thing): |
|
696 | def write(self, thing): | |
653 | if thing: |
|
697 | if thing: | |
654 | try: |
|
698 | try: | |
655 | self.server_write(thing) |
|
699 | self.server_write(thing) | |
656 | except socket.error as inst: |
|
700 | except socket.error as inst: | |
657 | if inst[0] != errno.ECONNRESET: |
|
701 | if inst[0] != errno.ECONNRESET: | |
658 | raise |
|
702 | raise | |
659 |
|
703 | |||
660 | def flush(self): |
|
704 | def flush(self): | |
661 | return None |
|
705 | return None | |
662 |
|
706 | |||
663 | def wsgiapplication(app_maker): |
|
707 | def wsgiapplication(app_maker): | |
664 | '''For compatibility with old CGI scripts. A plain hgweb() or hgwebdir() |
|
708 | '''For compatibility with old CGI scripts. A plain hgweb() or hgwebdir() | |
665 | can and should now be used as a WSGI application.''' |
|
709 | can and should now be used as a WSGI application.''' | |
666 | application = app_maker() |
|
710 | application = app_maker() | |
667 | def run_wsgi(env, respond): |
|
711 | def run_wsgi(env, respond): | |
668 | return application(env, respond) |
|
712 | return application(env, respond) | |
669 | return run_wsgi |
|
713 | return run_wsgi |
@@ -1,247 +1,416 | |||||
1 | from __future__ import absolute_import, print_function |
|
1 | from __future__ import absolute_import, print_function | |
2 |
|
2 | |||
3 | import unittest |
|
3 | import unittest | |
4 |
|
4 | |||
5 | from mercurial.hgweb import ( |
|
5 | from mercurial.hgweb import ( | |
6 | request as requestmod, |
|
6 | request as requestmod, | |
7 | ) |
|
7 | ) | |
8 | from mercurial import ( |
|
8 | from mercurial import ( | |
9 | error, |
|
9 | error, | |
10 | ) |
|
10 | ) | |
11 |
|
11 | |||
12 | DEFAULT_ENV = { |
|
12 | DEFAULT_ENV = { | |
13 | r'REQUEST_METHOD': r'GET', |
|
13 | r'REQUEST_METHOD': r'GET', | |
14 | r'SERVER_NAME': r'testserver', |
|
14 | r'SERVER_NAME': r'testserver', | |
15 | r'SERVER_PORT': r'80', |
|
15 | r'SERVER_PORT': r'80', | |
16 | r'SERVER_PROTOCOL': r'http', |
|
16 | r'SERVER_PROTOCOL': r'http', | |
17 | r'wsgi.version': (1, 0), |
|
17 | r'wsgi.version': (1, 0), | |
18 | r'wsgi.url_scheme': r'http', |
|
18 | r'wsgi.url_scheme': r'http', | |
19 | r'wsgi.input': None, |
|
19 | r'wsgi.input': None, | |
20 | r'wsgi.errors': None, |
|
20 | r'wsgi.errors': None, | |
21 | r'wsgi.multithread': False, |
|
21 | r'wsgi.multithread': False, | |
22 | r'wsgi.multiprocess': True, |
|
22 | r'wsgi.multiprocess': True, | |
23 | r'wsgi.run_once': False, |
|
23 | r'wsgi.run_once': False, | |
24 | } |
|
24 | } | |
25 |
|
25 | |||
26 | def parse(env, bodyfh=None, reponame=None, extra=None): |
|
26 | def parse(env, bodyfh=None, reponame=None, altbaseurl=None, extra=None): | |
27 | env = dict(env) |
|
27 | env = dict(env) | |
28 | env.update(extra or {}) |
|
28 | env.update(extra or {}) | |
29 |
|
29 | |||
30 |
return requestmod.parserequestfromenv(env, bodyfh, reponame=reponame |
|
30 | return requestmod.parserequestfromenv(env, bodyfh, reponame=reponame, | |
|
31 | altbaseurl=altbaseurl) | |||
31 |
|
32 | |||
32 | class ParseRequestTests(unittest.TestCase): |
|
33 | class ParseRequestTests(unittest.TestCase): | |
33 | def testdefault(self): |
|
34 | def testdefault(self): | |
34 | r = parse(DEFAULT_ENV) |
|
35 | r = parse(DEFAULT_ENV) | |
35 | self.assertEqual(r.url, b'http://testserver') |
|
36 | self.assertEqual(r.url, b'http://testserver') | |
36 | self.assertEqual(r.baseurl, b'http://testserver') |
|
37 | self.assertEqual(r.baseurl, b'http://testserver') | |
37 | self.assertEqual(r.advertisedurl, b'http://testserver') |
|
38 | self.assertEqual(r.advertisedurl, b'http://testserver') | |
38 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
39 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
39 | self.assertEqual(r.urlscheme, b'http') |
|
40 | self.assertEqual(r.urlscheme, b'http') | |
40 | self.assertEqual(r.method, b'GET') |
|
41 | self.assertEqual(r.method, b'GET') | |
41 | self.assertIsNone(r.remoteuser) |
|
42 | self.assertIsNone(r.remoteuser) | |
42 | self.assertIsNone(r.remotehost) |
|
43 | self.assertIsNone(r.remotehost) | |
43 | self.assertEqual(r.apppath, b'') |
|
44 | self.assertEqual(r.apppath, b'') | |
44 | self.assertEqual(r.dispatchparts, []) |
|
45 | self.assertEqual(r.dispatchparts, []) | |
45 | self.assertIsNone(r.dispatchpath) |
|
46 | self.assertIsNone(r.dispatchpath) | |
46 | self.assertIsNone(r.reponame) |
|
47 | self.assertIsNone(r.reponame) | |
47 | self.assertEqual(r.querystring, b'') |
|
48 | self.assertEqual(r.querystring, b'') | |
48 | self.assertEqual(len(r.qsparams), 0) |
|
49 | self.assertEqual(len(r.qsparams), 0) | |
49 | self.assertEqual(len(r.headers), 0) |
|
50 | self.assertEqual(len(r.headers), 0) | |
50 |
|
51 | |||
51 | def testcustomport(self): |
|
52 | def testcustomport(self): | |
52 | r = parse(DEFAULT_ENV, extra={ |
|
53 | r = parse(DEFAULT_ENV, extra={ | |
53 | r'SERVER_PORT': r'8000', |
|
54 | r'SERVER_PORT': r'8000', | |
54 | }) |
|
55 | }) | |
55 |
|
56 | |||
56 | self.assertEqual(r.url, b'http://testserver:8000') |
|
57 | self.assertEqual(r.url, b'http://testserver:8000') | |
57 | self.assertEqual(r.baseurl, b'http://testserver:8000') |
|
58 | self.assertEqual(r.baseurl, b'http://testserver:8000') | |
58 | self.assertEqual(r.advertisedurl, b'http://testserver:8000') |
|
59 | self.assertEqual(r.advertisedurl, b'http://testserver:8000') | |
59 | self.assertEqual(r.advertisedbaseurl, b'http://testserver:8000') |
|
60 | self.assertEqual(r.advertisedbaseurl, b'http://testserver:8000') | |
60 |
|
61 | |||
61 | r = parse(DEFAULT_ENV, extra={ |
|
62 | r = parse(DEFAULT_ENV, extra={ | |
62 | r'SERVER_PORT': r'4000', |
|
63 | r'SERVER_PORT': r'4000', | |
63 | r'wsgi.url_scheme': r'https', |
|
64 | r'wsgi.url_scheme': r'https', | |
64 | }) |
|
65 | }) | |
65 |
|
66 | |||
66 | self.assertEqual(r.url, b'https://testserver:4000') |
|
67 | self.assertEqual(r.url, b'https://testserver:4000') | |
67 | self.assertEqual(r.baseurl, b'https://testserver:4000') |
|
68 | self.assertEqual(r.baseurl, b'https://testserver:4000') | |
68 | self.assertEqual(r.advertisedurl, b'https://testserver:4000') |
|
69 | self.assertEqual(r.advertisedurl, b'https://testserver:4000') | |
69 | self.assertEqual(r.advertisedbaseurl, b'https://testserver:4000') |
|
70 | self.assertEqual(r.advertisedbaseurl, b'https://testserver:4000') | |
70 |
|
71 | |||
71 | def testhttphost(self): |
|
72 | def testhttphost(self): | |
72 | r = parse(DEFAULT_ENV, extra={ |
|
73 | r = parse(DEFAULT_ENV, extra={ | |
73 | r'HTTP_HOST': r'altserver', |
|
74 | r'HTTP_HOST': r'altserver', | |
74 | }) |
|
75 | }) | |
75 |
|
76 | |||
76 | self.assertEqual(r.url, b'http://altserver') |
|
77 | self.assertEqual(r.url, b'http://altserver') | |
77 | self.assertEqual(r.baseurl, b'http://altserver') |
|
78 | self.assertEqual(r.baseurl, b'http://altserver') | |
78 | self.assertEqual(r.advertisedurl, b'http://testserver') |
|
79 | self.assertEqual(r.advertisedurl, b'http://testserver') | |
79 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
80 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
80 |
|
81 | |||
81 | def testscriptname(self): |
|
82 | def testscriptname(self): | |
82 | r = parse(DEFAULT_ENV, extra={ |
|
83 | r = parse(DEFAULT_ENV, extra={ | |
83 | r'SCRIPT_NAME': r'', |
|
84 | r'SCRIPT_NAME': r'', | |
84 | }) |
|
85 | }) | |
85 |
|
86 | |||
86 | self.assertEqual(r.url, b'http://testserver') |
|
87 | self.assertEqual(r.url, b'http://testserver') | |
87 | self.assertEqual(r.baseurl, b'http://testserver') |
|
88 | self.assertEqual(r.baseurl, b'http://testserver') | |
88 | self.assertEqual(r.advertisedurl, b'http://testserver') |
|
89 | self.assertEqual(r.advertisedurl, b'http://testserver') | |
89 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
90 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
90 | self.assertEqual(r.apppath, b'') |
|
91 | self.assertEqual(r.apppath, b'') | |
91 | self.assertEqual(r.dispatchparts, []) |
|
92 | self.assertEqual(r.dispatchparts, []) | |
92 | self.assertIsNone(r.dispatchpath) |
|
93 | self.assertIsNone(r.dispatchpath) | |
93 |
|
94 | |||
94 | r = parse(DEFAULT_ENV, extra={ |
|
95 | r = parse(DEFAULT_ENV, extra={ | |
95 | r'SCRIPT_NAME': r'/script', |
|
96 | r'SCRIPT_NAME': r'/script', | |
96 | }) |
|
97 | }) | |
97 |
|
98 | |||
98 | self.assertEqual(r.url, b'http://testserver/script') |
|
99 | self.assertEqual(r.url, b'http://testserver/script') | |
99 | self.assertEqual(r.baseurl, b'http://testserver') |
|
100 | self.assertEqual(r.baseurl, b'http://testserver') | |
100 | self.assertEqual(r.advertisedurl, b'http://testserver/script') |
|
101 | self.assertEqual(r.advertisedurl, b'http://testserver/script') | |
101 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
102 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
102 | self.assertEqual(r.apppath, b'/script') |
|
103 | self.assertEqual(r.apppath, b'/script') | |
103 | self.assertEqual(r.dispatchparts, []) |
|
104 | self.assertEqual(r.dispatchparts, []) | |
104 | self.assertIsNone(r.dispatchpath) |
|
105 | self.assertIsNone(r.dispatchpath) | |
105 |
|
106 | |||
106 | r = parse(DEFAULT_ENV, extra={ |
|
107 | r = parse(DEFAULT_ENV, extra={ | |
107 | r'SCRIPT_NAME': r'/multiple words', |
|
108 | r'SCRIPT_NAME': r'/multiple words', | |
108 | }) |
|
109 | }) | |
109 |
|
110 | |||
110 | self.assertEqual(r.url, b'http://testserver/multiple%20words') |
|
111 | self.assertEqual(r.url, b'http://testserver/multiple%20words') | |
111 | self.assertEqual(r.baseurl, b'http://testserver') |
|
112 | self.assertEqual(r.baseurl, b'http://testserver') | |
112 | self.assertEqual(r.advertisedurl, b'http://testserver/multiple%20words') |
|
113 | self.assertEqual(r.advertisedurl, b'http://testserver/multiple%20words') | |
113 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
114 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
114 | self.assertEqual(r.apppath, b'/multiple words') |
|
115 | self.assertEqual(r.apppath, b'/multiple words') | |
115 | self.assertEqual(r.dispatchparts, []) |
|
116 | self.assertEqual(r.dispatchparts, []) | |
116 | self.assertIsNone(r.dispatchpath) |
|
117 | self.assertIsNone(r.dispatchpath) | |
117 |
|
118 | |||
118 | def testpathinfo(self): |
|
119 | def testpathinfo(self): | |
119 | r = parse(DEFAULT_ENV, extra={ |
|
120 | r = parse(DEFAULT_ENV, extra={ | |
120 | r'PATH_INFO': r'', |
|
121 | r'PATH_INFO': r'', | |
121 | }) |
|
122 | }) | |
122 |
|
123 | |||
123 | self.assertEqual(r.url, b'http://testserver') |
|
124 | self.assertEqual(r.url, b'http://testserver') | |
124 | self.assertEqual(r.baseurl, b'http://testserver') |
|
125 | self.assertEqual(r.baseurl, b'http://testserver') | |
125 | self.assertEqual(r.advertisedurl, b'http://testserver') |
|
126 | self.assertEqual(r.advertisedurl, b'http://testserver') | |
126 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
127 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
127 | self.assertEqual(r.apppath, b'') |
|
128 | self.assertEqual(r.apppath, b'') | |
128 | self.assertEqual(r.dispatchparts, []) |
|
129 | self.assertEqual(r.dispatchparts, []) | |
129 | self.assertEqual(r.dispatchpath, b'') |
|
130 | self.assertEqual(r.dispatchpath, b'') | |
130 |
|
131 | |||
131 | r = parse(DEFAULT_ENV, extra={ |
|
132 | r = parse(DEFAULT_ENV, extra={ | |
132 | r'PATH_INFO': r'/pathinfo', |
|
133 | r'PATH_INFO': r'/pathinfo', | |
133 | }) |
|
134 | }) | |
134 |
|
135 | |||
135 | self.assertEqual(r.url, b'http://testserver/pathinfo') |
|
136 | self.assertEqual(r.url, b'http://testserver/pathinfo') | |
136 | self.assertEqual(r.baseurl, b'http://testserver') |
|
137 | self.assertEqual(r.baseurl, b'http://testserver') | |
137 | self.assertEqual(r.advertisedurl, b'http://testserver/pathinfo') |
|
138 | self.assertEqual(r.advertisedurl, b'http://testserver/pathinfo') | |
138 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
139 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
139 | self.assertEqual(r.apppath, b'') |
|
140 | self.assertEqual(r.apppath, b'') | |
140 | self.assertEqual(r.dispatchparts, [b'pathinfo']) |
|
141 | self.assertEqual(r.dispatchparts, [b'pathinfo']) | |
141 | self.assertEqual(r.dispatchpath, b'pathinfo') |
|
142 | self.assertEqual(r.dispatchpath, b'pathinfo') | |
142 |
|
143 | |||
143 | r = parse(DEFAULT_ENV, extra={ |
|
144 | r = parse(DEFAULT_ENV, extra={ | |
144 | r'PATH_INFO': r'/one/two/', |
|
145 | r'PATH_INFO': r'/one/two/', | |
145 | }) |
|
146 | }) | |
146 |
|
147 | |||
147 | self.assertEqual(r.url, b'http://testserver/one/two/') |
|
148 | self.assertEqual(r.url, b'http://testserver/one/two/') | |
148 | self.assertEqual(r.baseurl, b'http://testserver') |
|
149 | self.assertEqual(r.baseurl, b'http://testserver') | |
149 | self.assertEqual(r.advertisedurl, b'http://testserver/one/two/') |
|
150 | self.assertEqual(r.advertisedurl, b'http://testserver/one/two/') | |
150 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
151 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
151 | self.assertEqual(r.apppath, b'') |
|
152 | self.assertEqual(r.apppath, b'') | |
152 | self.assertEqual(r.dispatchparts, [b'one', b'two']) |
|
153 | self.assertEqual(r.dispatchparts, [b'one', b'two']) | |
153 | self.assertEqual(r.dispatchpath, b'one/two') |
|
154 | self.assertEqual(r.dispatchpath, b'one/two') | |
154 |
|
155 | |||
155 | def testscriptandpathinfo(self): |
|
156 | def testscriptandpathinfo(self): | |
156 | r = parse(DEFAULT_ENV, extra={ |
|
157 | r = parse(DEFAULT_ENV, extra={ | |
157 | r'SCRIPT_NAME': r'/script', |
|
158 | r'SCRIPT_NAME': r'/script', | |
158 | r'PATH_INFO': r'/pathinfo', |
|
159 | r'PATH_INFO': r'/pathinfo', | |
159 | }) |
|
160 | }) | |
160 |
|
161 | |||
161 | self.assertEqual(r.url, b'http://testserver/script/pathinfo') |
|
162 | self.assertEqual(r.url, b'http://testserver/script/pathinfo') | |
162 | self.assertEqual(r.baseurl, b'http://testserver') |
|
163 | self.assertEqual(r.baseurl, b'http://testserver') | |
163 | self.assertEqual(r.advertisedurl, b'http://testserver/script/pathinfo') |
|
164 | self.assertEqual(r.advertisedurl, b'http://testserver/script/pathinfo') | |
164 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
165 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
165 | self.assertEqual(r.apppath, b'/script') |
|
166 | self.assertEqual(r.apppath, b'/script') | |
166 | self.assertEqual(r.dispatchparts, [b'pathinfo']) |
|
167 | self.assertEqual(r.dispatchparts, [b'pathinfo']) | |
167 | self.assertEqual(r.dispatchpath, b'pathinfo') |
|
168 | self.assertEqual(r.dispatchpath, b'pathinfo') | |
168 |
|
169 | |||
169 | r = parse(DEFAULT_ENV, extra={ |
|
170 | r = parse(DEFAULT_ENV, extra={ | |
170 | r'SCRIPT_NAME': r'/script1/script2', |
|
171 | r'SCRIPT_NAME': r'/script1/script2', | |
171 | r'PATH_INFO': r'/path1/path2', |
|
172 | r'PATH_INFO': r'/path1/path2', | |
172 | }) |
|
173 | }) | |
173 |
|
174 | |||
174 | self.assertEqual(r.url, |
|
175 | self.assertEqual(r.url, | |
175 | b'http://testserver/script1/script2/path1/path2') |
|
176 | b'http://testserver/script1/script2/path1/path2') | |
176 | self.assertEqual(r.baseurl, b'http://testserver') |
|
177 | self.assertEqual(r.baseurl, b'http://testserver') | |
177 | self.assertEqual(r.advertisedurl, |
|
178 | self.assertEqual(r.advertisedurl, | |
178 | b'http://testserver/script1/script2/path1/path2') |
|
179 | b'http://testserver/script1/script2/path1/path2') | |
179 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
180 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
180 | self.assertEqual(r.apppath, b'/script1/script2') |
|
181 | self.assertEqual(r.apppath, b'/script1/script2') | |
181 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) |
|
182 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |
182 | self.assertEqual(r.dispatchpath, b'path1/path2') |
|
183 | self.assertEqual(r.dispatchpath, b'path1/path2') | |
183 |
|
184 | |||
184 | r = parse(DEFAULT_ENV, extra={ |
|
185 | r = parse(DEFAULT_ENV, extra={ | |
185 | r'HTTP_HOST': r'hostserver', |
|
186 | r'HTTP_HOST': r'hostserver', | |
186 | r'SCRIPT_NAME': r'/script', |
|
187 | r'SCRIPT_NAME': r'/script', | |
187 | r'PATH_INFO': r'/pathinfo', |
|
188 | r'PATH_INFO': r'/pathinfo', | |
188 | }) |
|
189 | }) | |
189 |
|
190 | |||
190 | self.assertEqual(r.url, b'http://hostserver/script/pathinfo') |
|
191 | self.assertEqual(r.url, b'http://hostserver/script/pathinfo') | |
191 | self.assertEqual(r.baseurl, b'http://hostserver') |
|
192 | self.assertEqual(r.baseurl, b'http://hostserver') | |
192 | self.assertEqual(r.advertisedurl, b'http://testserver/script/pathinfo') |
|
193 | self.assertEqual(r.advertisedurl, b'http://testserver/script/pathinfo') | |
193 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
194 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
194 | self.assertEqual(r.apppath, b'/script') |
|
195 | self.assertEqual(r.apppath, b'/script') | |
195 | self.assertEqual(r.dispatchparts, [b'pathinfo']) |
|
196 | self.assertEqual(r.dispatchparts, [b'pathinfo']) | |
196 | self.assertEqual(r.dispatchpath, b'pathinfo') |
|
197 | self.assertEqual(r.dispatchpath, b'pathinfo') | |
197 |
|
198 | |||
198 | def testreponame(self): |
|
199 | def testreponame(self): | |
199 | """repository path components get stripped from URL.""" |
|
200 | """repository path components get stripped from URL.""" | |
200 |
|
201 | |||
201 | with self.assertRaisesRegexp(error.ProgrammingError, |
|
202 | with self.assertRaisesRegexp(error.ProgrammingError, | |
202 | b'reponame requires PATH_INFO'): |
|
203 | b'reponame requires PATH_INFO'): | |
203 | parse(DEFAULT_ENV, reponame=b'repo') |
|
204 | parse(DEFAULT_ENV, reponame=b'repo') | |
204 |
|
205 | |||
205 | with self.assertRaisesRegexp(error.ProgrammingError, |
|
206 | with self.assertRaisesRegexp(error.ProgrammingError, | |
206 | b'PATH_INFO does not begin with repo ' |
|
207 | b'PATH_INFO does not begin with repo ' | |
207 | b'name'): |
|
208 | b'name'): | |
208 | parse(DEFAULT_ENV, reponame=b'repo', extra={ |
|
209 | parse(DEFAULT_ENV, reponame=b'repo', extra={ | |
209 | r'PATH_INFO': r'/pathinfo', |
|
210 | r'PATH_INFO': r'/pathinfo', | |
210 | }) |
|
211 | }) | |
211 |
|
212 | |||
212 | with self.assertRaisesRegexp(error.ProgrammingError, |
|
213 | with self.assertRaisesRegexp(error.ProgrammingError, | |
213 | b'reponame prefix of PATH_INFO'): |
|
214 | b'reponame prefix of PATH_INFO'): | |
214 | parse(DEFAULT_ENV, reponame=b'repo', extra={ |
|
215 | parse(DEFAULT_ENV, reponame=b'repo', extra={ | |
215 | r'PATH_INFO': r'/repoextra/path', |
|
216 | r'PATH_INFO': r'/repoextra/path', | |
216 | }) |
|
217 | }) | |
217 |
|
218 | |||
218 | r = parse(DEFAULT_ENV, reponame=b'repo', extra={ |
|
219 | r = parse(DEFAULT_ENV, reponame=b'repo', extra={ | |
219 | r'PATH_INFO': r'/repo/path1/path2', |
|
220 | r'PATH_INFO': r'/repo/path1/path2', | |
220 | }) |
|
221 | }) | |
221 |
|
222 | |||
222 | self.assertEqual(r.url, b'http://testserver/repo/path1/path2') |
|
223 | self.assertEqual(r.url, b'http://testserver/repo/path1/path2') | |
223 | self.assertEqual(r.baseurl, b'http://testserver') |
|
224 | self.assertEqual(r.baseurl, b'http://testserver') | |
224 | self.assertEqual(r.advertisedurl, b'http://testserver/repo/path1/path2') |
|
225 | self.assertEqual(r.advertisedurl, b'http://testserver/repo/path1/path2') | |
225 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
226 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
226 | self.assertEqual(r.apppath, b'/repo') |
|
227 | self.assertEqual(r.apppath, b'/repo') | |
227 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) |
|
228 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |
228 | self.assertEqual(r.dispatchpath, b'path1/path2') |
|
229 | self.assertEqual(r.dispatchpath, b'path1/path2') | |
229 | self.assertEqual(r.reponame, b'repo') |
|
230 | self.assertEqual(r.reponame, b'repo') | |
230 |
|
231 | |||
231 | r = parse(DEFAULT_ENV, reponame=b'prefix/repo', extra={ |
|
232 | r = parse(DEFAULT_ENV, reponame=b'prefix/repo', extra={ | |
232 | r'PATH_INFO': r'/prefix/repo/path1/path2', |
|
233 | r'PATH_INFO': r'/prefix/repo/path1/path2', | |
233 | }) |
|
234 | }) | |
234 |
|
235 | |||
235 | self.assertEqual(r.url, b'http://testserver/prefix/repo/path1/path2') |
|
236 | self.assertEqual(r.url, b'http://testserver/prefix/repo/path1/path2') | |
236 | self.assertEqual(r.baseurl, b'http://testserver') |
|
237 | self.assertEqual(r.baseurl, b'http://testserver') | |
237 | self.assertEqual(r.advertisedurl, |
|
238 | self.assertEqual(r.advertisedurl, | |
238 | b'http://testserver/prefix/repo/path1/path2') |
|
239 | b'http://testserver/prefix/repo/path1/path2') | |
239 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') |
|
240 | self.assertEqual(r.advertisedbaseurl, b'http://testserver') | |
240 | self.assertEqual(r.apppath, b'/prefix/repo') |
|
241 | self.assertEqual(r.apppath, b'/prefix/repo') | |
241 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) |
|
242 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |
242 | self.assertEqual(r.dispatchpath, b'path1/path2') |
|
243 | self.assertEqual(r.dispatchpath, b'path1/path2') | |
243 | self.assertEqual(r.reponame, b'prefix/repo') |
|
244 | self.assertEqual(r.reponame, b'prefix/repo') | |
244 |
|
245 | |||
|
246 | def testaltbaseurl(self): | |||
|
247 | # Simple hostname remap. | |||
|
248 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver') | |||
|
249 | ||||
|
250 | self.assertEqual(r.url, b'http://testserver') | |||
|
251 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
252 | self.assertEqual(r.advertisedurl, b'http://altserver') | |||
|
253 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
254 | self.assertEqual(r.urlscheme, b'http') | |||
|
255 | self.assertEqual(r.apppath, b'') | |||
|
256 | self.assertEqual(r.dispatchparts, []) | |||
|
257 | self.assertIsNone(r.dispatchpath) | |||
|
258 | self.assertIsNone(r.reponame) | |||
|
259 | ||||
|
260 | # With a custom port. | |||
|
261 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver:8000') | |||
|
262 | self.assertEqual(r.url, b'http://testserver') | |||
|
263 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
264 | self.assertEqual(r.advertisedurl, b'http://altserver:8000') | |||
|
265 | self.assertEqual(r.advertisedbaseurl, b'http://altserver:8000') | |||
|
266 | self.assertEqual(r.urlscheme, b'http') | |||
|
267 | self.assertEqual(r.apppath, b'') | |||
|
268 | self.assertEqual(r.dispatchparts, []) | |||
|
269 | self.assertIsNone(r.dispatchpath) | |||
|
270 | self.assertIsNone(r.reponame) | |||
|
271 | ||||
|
272 | # With a changed protocol. | |||
|
273 | r = parse(DEFAULT_ENV, altbaseurl='https://altserver') | |||
|
274 | self.assertEqual(r.url, b'http://testserver') | |||
|
275 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
276 | self.assertEqual(r.advertisedurl, b'https://altserver') | |||
|
277 | self.assertEqual(r.advertisedbaseurl, b'https://altserver') | |||
|
278 | # URL scheme is defined as the actual scheme, not advertised. | |||
|
279 | self.assertEqual(r.urlscheme, b'http') | |||
|
280 | self.assertEqual(r.apppath, b'') | |||
|
281 | self.assertEqual(r.dispatchparts, []) | |||
|
282 | self.assertIsNone(r.dispatchpath) | |||
|
283 | self.assertIsNone(r.reponame) | |||
|
284 | ||||
|
285 | # Need to specify explicit port number for proper https:// alt URLs. | |||
|
286 | r = parse(DEFAULT_ENV, altbaseurl='https://altserver:443') | |||
|
287 | self.assertEqual(r.url, b'http://testserver') | |||
|
288 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
289 | self.assertEqual(r.advertisedurl, b'https://altserver') | |||
|
290 | self.assertEqual(r.advertisedbaseurl, b'https://altserver') | |||
|
291 | self.assertEqual(r.urlscheme, b'http') | |||
|
292 | self.assertEqual(r.apppath, b'') | |||
|
293 | self.assertEqual(r.dispatchparts, []) | |||
|
294 | self.assertIsNone(r.dispatchpath) | |||
|
295 | self.assertIsNone(r.reponame) | |||
|
296 | ||||
|
297 | # With only PATH_INFO defined. | |||
|
298 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver', extra={ | |||
|
299 | r'PATH_INFO': r'/path1/path2', | |||
|
300 | }) | |||
|
301 | self.assertEqual(r.url, b'http://testserver/path1/path2') | |||
|
302 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
303 | self.assertEqual(r.advertisedurl, b'http://altserver/path1/path2') | |||
|
304 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
305 | self.assertEqual(r.urlscheme, b'http') | |||
|
306 | self.assertEqual(r.apppath, b'') | |||
|
307 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |||
|
308 | self.assertEqual(r.dispatchpath, b'path1/path2') | |||
|
309 | self.assertIsNone(r.reponame) | |||
|
310 | ||||
|
311 | # Path on alt URL. | |||
|
312 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver/altpath') | |||
|
313 | self.assertEqual(r.url, b'http://testserver') | |||
|
314 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
315 | self.assertEqual(r.advertisedurl, b'http://altserver/altpath') | |||
|
316 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
317 | self.assertEqual(r.urlscheme, b'http') | |||
|
318 | self.assertEqual(r.apppath, b'/altpath') | |||
|
319 | self.assertEqual(r.dispatchparts, []) | |||
|
320 | self.assertIsNone(r.dispatchpath) | |||
|
321 | self.assertIsNone(r.reponame) | |||
|
322 | ||||
|
323 | # With a trailing slash. | |||
|
324 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver/altpath/') | |||
|
325 | self.assertEqual(r.url, b'http://testserver') | |||
|
326 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
327 | self.assertEqual(r.advertisedurl, b'http://altserver/altpath/') | |||
|
328 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
329 | self.assertEqual(r.urlscheme, b'http') | |||
|
330 | self.assertEqual(r.apppath, b'/altpath/') | |||
|
331 | self.assertEqual(r.dispatchparts, []) | |||
|
332 | self.assertIsNone(r.dispatchpath) | |||
|
333 | self.assertIsNone(r.reponame) | |||
|
334 | ||||
|
335 | # PATH_INFO + path on alt URL. | |||
|
336 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver/altpath', extra={ | |||
|
337 | r'PATH_INFO': r'/path1/path2', | |||
|
338 | }) | |||
|
339 | self.assertEqual(r.url, b'http://testserver/path1/path2') | |||
|
340 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
341 | self.assertEqual(r.advertisedurl, | |||
|
342 | b'http://altserver/altpath/path1/path2') | |||
|
343 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
344 | self.assertEqual(r.urlscheme, b'http') | |||
|
345 | self.assertEqual(r.apppath, b'/altpath') | |||
|
346 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |||
|
347 | self.assertEqual(r.dispatchpath, b'path1/path2') | |||
|
348 | self.assertIsNone(r.reponame) | |||
|
349 | ||||
|
350 | # PATH_INFO + path on alt URL with trailing slash. | |||
|
351 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver/altpath/', extra={ | |||
|
352 | r'PATH_INFO': r'/path1/path2', | |||
|
353 | }) | |||
|
354 | self.assertEqual(r.url, b'http://testserver/path1/path2') | |||
|
355 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
356 | self.assertEqual(r.advertisedurl, | |||
|
357 | b'http://altserver/altpath//path1/path2') | |||
|
358 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
359 | self.assertEqual(r.urlscheme, b'http') | |||
|
360 | self.assertEqual(r.apppath, b'/altpath/') | |||
|
361 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |||
|
362 | self.assertEqual(r.dispatchpath, b'path1/path2') | |||
|
363 | self.assertIsNone(r.reponame) | |||
|
364 | ||||
|
365 | # Local SCRIPT_NAME is ignored. | |||
|
366 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver', extra={ | |||
|
367 | r'SCRIPT_NAME': r'/script', | |||
|
368 | r'PATH_INFO': r'/path1/path2', | |||
|
369 | }) | |||
|
370 | self.assertEqual(r.url, b'http://testserver/script/path1/path2') | |||
|
371 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
372 | self.assertEqual(r.advertisedurl, b'http://altserver/path1/path2') | |||
|
373 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
374 | self.assertEqual(r.urlscheme, b'http') | |||
|
375 | self.assertEqual(r.apppath, b'') | |||
|
376 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |||
|
377 | self.assertEqual(r.dispatchpath, b'path1/path2') | |||
|
378 | self.assertIsNone(r.reponame) | |||
|
379 | ||||
|
380 | # Use remote's path for script name, app path | |||
|
381 | r = parse(DEFAULT_ENV, altbaseurl='http://altserver/altroot', extra={ | |||
|
382 | r'SCRIPT_NAME': r'/script', | |||
|
383 | r'PATH_INFO': r'/path1/path2', | |||
|
384 | }) | |||
|
385 | self.assertEqual(r.url, b'http://testserver/script/path1/path2') | |||
|
386 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
387 | self.assertEqual(r.advertisedurl, | |||
|
388 | b'http://altserver/altroot/path1/path2') | |||
|
389 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
390 | self.assertEqual(r.urlscheme, b'http') | |||
|
391 | self.assertEqual(r.apppath, b'/altroot') | |||
|
392 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |||
|
393 | self.assertEqual(r.dispatchpath, b'path1/path2') | |||
|
394 | self.assertIsNone(r.reponame) | |||
|
395 | ||||
|
396 | # reponame is factored in properly. | |||
|
397 | r = parse(DEFAULT_ENV, reponame=b'repo', | |||
|
398 | altbaseurl='http://altserver/altroot', | |||
|
399 | extra={ | |||
|
400 | r'SCRIPT_NAME': r'/script', | |||
|
401 | r'PATH_INFO': r'/repo/path1/path2', | |||
|
402 | }) | |||
|
403 | ||||
|
404 | self.assertEqual(r.url, b'http://testserver/script/repo/path1/path2') | |||
|
405 | self.assertEqual(r.baseurl, b'http://testserver') | |||
|
406 | self.assertEqual(r.advertisedurl, | |||
|
407 | b'http://altserver/altroot/repo/path1/path2') | |||
|
408 | self.assertEqual(r.advertisedbaseurl, b'http://altserver') | |||
|
409 | self.assertEqual(r.apppath, b'/altroot/repo') | |||
|
410 | self.assertEqual(r.dispatchparts, [b'path1', b'path2']) | |||
|
411 | self.assertEqual(r.dispatchpath, b'path1/path2') | |||
|
412 | self.assertEqual(r.reponame, b'repo') | |||
|
413 | ||||
245 | if __name__ == '__main__': |
|
414 | if __name__ == '__main__': | |
246 | import silenttestrunner |
|
415 | import silenttestrunner | |
247 | silenttestrunner.main(__name__) |
|
416 | silenttestrunner.main(__name__) |
General Comments 0
You need to be logged in to leave comments.
Login now