Show More
@@ -1,438 +1,438 | |||||
1 | # hgweb/hgweb_mod.py - Web interface for a repository. |
|
1 | # hgweb/hgweb_mod.py - Web interface for a repository. | |
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 2005-2007 Matt Mackall <mpm@selenic.com> |
|
4 | # Copyright 2005-2007 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 contextlib |
|
11 | import contextlib | |
12 | import os |
|
12 | import os | |
13 |
|
13 | |||
14 | from .common import ( |
|
14 | from .common import ( | |
15 | ErrorResponse, |
|
15 | ErrorResponse, | |
16 | HTTP_BAD_REQUEST, |
|
16 | HTTP_BAD_REQUEST, | |
17 | HTTP_NOT_FOUND, |
|
17 | HTTP_NOT_FOUND, | |
18 | HTTP_NOT_MODIFIED, |
|
18 | HTTP_NOT_MODIFIED, | |
19 | HTTP_OK, |
|
19 | HTTP_OK, | |
20 | HTTP_SERVER_ERROR, |
|
20 | HTTP_SERVER_ERROR, | |
21 | caching, |
|
21 | caching, | |
22 | cspvalues, |
|
22 | cspvalues, | |
23 | permhooks, |
|
23 | permhooks, | |
24 | ) |
|
24 | ) | |
25 |
|
25 | |||
26 | from .. import ( |
|
26 | from .. import ( | |
27 | encoding, |
|
27 | encoding, | |
28 | error, |
|
28 | error, | |
29 | formatter, |
|
29 | formatter, | |
30 | hg, |
|
30 | hg, | |
31 | hook, |
|
31 | hook, | |
32 | profiling, |
|
32 | profiling, | |
33 | pycompat, |
|
33 | pycompat, | |
34 | repoview, |
|
34 | repoview, | |
35 | templatefilters, |
|
35 | templatefilters, | |
36 | templater, |
|
36 | templater, | |
37 | ui as uimod, |
|
37 | ui as uimod, | |
38 | util, |
|
38 | util, | |
39 | wireprotoserver, |
|
39 | wireprotoserver, | |
40 | ) |
|
40 | ) | |
41 |
|
41 | |||
42 | from . import ( |
|
42 | from . import ( | |
43 | request as requestmod, |
|
43 | request as requestmod, | |
44 | webcommands, |
|
44 | webcommands, | |
45 | webutil, |
|
45 | webutil, | |
46 | wsgicgi, |
|
46 | wsgicgi, | |
47 | ) |
|
47 | ) | |
48 |
|
48 | |||
49 | archivespecs = util.sortdict(( |
|
49 | archivespecs = util.sortdict(( | |
50 | ('zip', ('application/zip', 'zip', '.zip', None)), |
|
50 | ('zip', ('application/zip', 'zip', '.zip', None)), | |
51 | ('gz', ('application/x-gzip', 'tgz', '.tar.gz', None)), |
|
51 | ('gz', ('application/x-gzip', 'tgz', '.tar.gz', None)), | |
52 | ('bz2', ('application/x-bzip2', 'tbz2', '.tar.bz2', None)), |
|
52 | ('bz2', ('application/x-bzip2', 'tbz2', '.tar.bz2', None)), | |
53 | )) |
|
53 | )) | |
54 |
|
54 | |||
55 | def getstyle(req, configfn, templatepath): |
|
55 | def getstyle(req, configfn, templatepath): | |
56 | fromreq = req.form.get('style', [None])[0] |
|
56 | fromreq = req.form.get('style', [None])[0] | |
57 | styles = ( |
|
57 | styles = ( | |
58 | fromreq, |
|
58 | fromreq, | |
59 | configfn('web', 'style'), |
|
59 | configfn('web', 'style'), | |
60 | 'paper', |
|
60 | 'paper', | |
61 | ) |
|
61 | ) | |
62 | return styles, templater.stylemap(styles, templatepath) |
|
62 | return styles, templater.stylemap(styles, templatepath) | |
63 |
|
63 | |||
64 | def makebreadcrumb(url, prefix=''): |
|
64 | def makebreadcrumb(url, prefix=''): | |
65 | '''Return a 'URL breadcrumb' list |
|
65 | '''Return a 'URL breadcrumb' list | |
66 |
|
66 | |||
67 | A 'URL breadcrumb' is a list of URL-name pairs, |
|
67 | A 'URL breadcrumb' is a list of URL-name pairs, | |
68 | corresponding to each of the path items on a URL. |
|
68 | corresponding to each of the path items on a URL. | |
69 | This can be used to create path navigation entries. |
|
69 | This can be used to create path navigation entries. | |
70 | ''' |
|
70 | ''' | |
71 | if url.endswith('/'): |
|
71 | if url.endswith('/'): | |
72 | url = url[:-1] |
|
72 | url = url[:-1] | |
73 | if prefix: |
|
73 | if prefix: | |
74 | url = '/' + prefix + url |
|
74 | url = '/' + prefix + url | |
75 | relpath = url |
|
75 | relpath = url | |
76 | if relpath.startswith('/'): |
|
76 | if relpath.startswith('/'): | |
77 | relpath = relpath[1:] |
|
77 | relpath = relpath[1:] | |
78 |
|
78 | |||
79 | breadcrumb = [] |
|
79 | breadcrumb = [] | |
80 | urlel = url |
|
80 | urlel = url | |
81 | pathitems = [''] + relpath.split('/') |
|
81 | pathitems = [''] + relpath.split('/') | |
82 | for pathel in reversed(pathitems): |
|
82 | for pathel in reversed(pathitems): | |
83 | if not pathel or not urlel: |
|
83 | if not pathel or not urlel: | |
84 | break |
|
84 | break | |
85 | breadcrumb.append({'url': urlel, 'name': pathel}) |
|
85 | breadcrumb.append({'url': urlel, 'name': pathel}) | |
86 | urlel = os.path.dirname(urlel) |
|
86 | urlel = os.path.dirname(urlel) | |
87 | return reversed(breadcrumb) |
|
87 | return reversed(breadcrumb) | |
88 |
|
88 | |||
89 | class requestcontext(object): |
|
89 | class requestcontext(object): | |
90 | """Holds state/context for an individual request. |
|
90 | """Holds state/context for an individual request. | |
91 |
|
91 | |||
92 | Servers can be multi-threaded. Holding state on the WSGI application |
|
92 | Servers can be multi-threaded. Holding state on the WSGI application | |
93 | is prone to race conditions. Instances of this class exist to hold |
|
93 | is prone to race conditions. Instances of this class exist to hold | |
94 | mutable and race-free state for requests. |
|
94 | mutable and race-free state for requests. | |
95 | """ |
|
95 | """ | |
96 | def __init__(self, app, repo): |
|
96 | def __init__(self, app, repo): | |
97 | self.repo = repo |
|
97 | self.repo = repo | |
98 | self.reponame = app.reponame |
|
98 | self.reponame = app.reponame | |
99 |
|
99 | |||
100 | self.archivespecs = archivespecs |
|
100 | self.archivespecs = archivespecs | |
101 |
|
101 | |||
102 | self.maxchanges = self.configint('web', 'maxchanges') |
|
102 | self.maxchanges = self.configint('web', 'maxchanges') | |
103 | self.stripecount = self.configint('web', 'stripes') |
|
103 | self.stripecount = self.configint('web', 'stripes') | |
104 | self.maxshortchanges = self.configint('web', 'maxshortchanges') |
|
104 | self.maxshortchanges = self.configint('web', 'maxshortchanges') | |
105 | self.maxfiles = self.configint('web', 'maxfiles') |
|
105 | self.maxfiles = self.configint('web', 'maxfiles') | |
106 | self.allowpull = self.configbool('web', 'allow-pull') |
|
106 | self.allowpull = self.configbool('web', 'allow-pull') | |
107 |
|
107 | |||
108 | # we use untrusted=False to prevent a repo owner from using |
|
108 | # we use untrusted=False to prevent a repo owner from using | |
109 | # web.templates in .hg/hgrc to get access to any file readable |
|
109 | # web.templates in .hg/hgrc to get access to any file readable | |
110 | # by the user running the CGI script |
|
110 | # by the user running the CGI script | |
111 | self.templatepath = self.config('web', 'templates', untrusted=False) |
|
111 | self.templatepath = self.config('web', 'templates', untrusted=False) | |
112 |
|
112 | |||
113 | # This object is more expensive to build than simple config values. |
|
113 | # This object is more expensive to build than simple config values. | |
114 | # It is shared across requests. The app will replace the object |
|
114 | # It is shared across requests. The app will replace the object | |
115 | # if it is updated. Since this is a reference and nothing should |
|
115 | # if it is updated. Since this is a reference and nothing should | |
116 | # modify the underlying object, it should be constant for the lifetime |
|
116 | # modify the underlying object, it should be constant for the lifetime | |
117 | # of the request. |
|
117 | # of the request. | |
118 | self.websubtable = app.websubtable |
|
118 | self.websubtable = app.websubtable | |
119 |
|
119 | |||
120 | self.csp, self.nonce = cspvalues(self.repo.ui) |
|
120 | self.csp, self.nonce = cspvalues(self.repo.ui) | |
121 |
|
121 | |||
122 | # Trust the settings from the .hg/hgrc files by default. |
|
122 | # Trust the settings from the .hg/hgrc files by default. | |
123 | def config(self, section, name, default=uimod._unset, untrusted=True): |
|
123 | def config(self, section, name, default=uimod._unset, untrusted=True): | |
124 | return self.repo.ui.config(section, name, default, |
|
124 | return self.repo.ui.config(section, name, default, | |
125 | untrusted=untrusted) |
|
125 | untrusted=untrusted) | |
126 |
|
126 | |||
127 | def configbool(self, section, name, default=uimod._unset, untrusted=True): |
|
127 | def configbool(self, section, name, default=uimod._unset, untrusted=True): | |
128 | return self.repo.ui.configbool(section, name, default, |
|
128 | return self.repo.ui.configbool(section, name, default, | |
129 | untrusted=untrusted) |
|
129 | untrusted=untrusted) | |
130 |
|
130 | |||
131 | def configint(self, section, name, default=uimod._unset, untrusted=True): |
|
131 | def configint(self, section, name, default=uimod._unset, untrusted=True): | |
132 | return self.repo.ui.configint(section, name, default, |
|
132 | return self.repo.ui.configint(section, name, default, | |
133 | untrusted=untrusted) |
|
133 | untrusted=untrusted) | |
134 |
|
134 | |||
135 | def configlist(self, section, name, default=uimod._unset, untrusted=True): |
|
135 | def configlist(self, section, name, default=uimod._unset, untrusted=True): | |
136 | return self.repo.ui.configlist(section, name, default, |
|
136 | return self.repo.ui.configlist(section, name, default, | |
137 | untrusted=untrusted) |
|
137 | untrusted=untrusted) | |
138 |
|
138 | |||
139 | def archivelist(self, nodeid): |
|
139 | def archivelist(self, nodeid): | |
140 | allowed = self.configlist('web', 'allow_archive') |
|
140 | allowed = self.configlist('web', 'allow_archive') | |
141 | for typ, spec in self.archivespecs.iteritems(): |
|
141 | for typ, spec in self.archivespecs.iteritems(): | |
142 | if typ in allowed or self.configbool('web', 'allow%s' % typ): |
|
142 | if typ in allowed or self.configbool('web', 'allow%s' % typ): | |
143 | yield {'type': typ, 'extension': spec[2], 'node': nodeid} |
|
143 | yield {'type': typ, 'extension': spec[2], 'node': nodeid} | |
144 |
|
144 | |||
145 | def templater(self, wsgireq, req): |
|
145 | def templater(self, wsgireq, req): | |
146 | # determine scheme, port and server name |
|
146 | # determine scheme, port and server name | |
147 | # this is needed to create absolute urls |
|
147 | # this is needed to create absolute urls | |
148 | logourl = self.config('web', 'logourl') |
|
148 | logourl = self.config('web', 'logourl') | |
149 | logoimg = self.config('web', 'logoimg') |
|
149 | logoimg = self.config('web', 'logoimg') | |
150 | staticurl = (self.config('web', 'staticurl') |
|
150 | staticurl = (self.config('web', 'staticurl') | |
151 | or req.apppath + '/static/') |
|
151 | or req.apppath + '/static/') | |
152 | if not staticurl.endswith('/'): |
|
152 | if not staticurl.endswith('/'): | |
153 | staticurl += '/' |
|
153 | staticurl += '/' | |
154 |
|
154 | |||
155 | # some functions for the templater |
|
155 | # some functions for the templater | |
156 |
|
156 | |||
157 | def motd(**map): |
|
157 | def motd(**map): | |
158 | yield self.config('web', 'motd') |
|
158 | yield self.config('web', 'motd') | |
159 |
|
159 | |||
160 | # figure out which style to use |
|
160 | # figure out which style to use | |
161 |
|
161 | |||
162 | vars = {} |
|
162 | vars = {} | |
163 | styles, (style, mapfile) = getstyle(wsgireq, self.config, |
|
163 | styles, (style, mapfile) = getstyle(wsgireq, self.config, | |
164 | self.templatepath) |
|
164 | self.templatepath) | |
165 | if style == styles[0]: |
|
165 | if style == styles[0]: | |
166 | vars['style'] = style |
|
166 | vars['style'] = style | |
167 |
|
167 | |||
168 | sessionvars = webutil.sessionvars(vars, '?') |
|
168 | sessionvars = webutil.sessionvars(vars, '?') | |
169 |
|
169 | |||
170 | if not self.reponame: |
|
170 | if not self.reponame: | |
171 | self.reponame = (self.config('web', 'name', '') |
|
171 | self.reponame = (self.config('web', 'name', '') | |
172 | or wsgireq.env.get('REPO_NAME') |
|
172 | or wsgireq.env.get('REPO_NAME') | |
173 | or req.apppath or self.repo.root) |
|
173 | or req.apppath or self.repo.root) | |
174 |
|
174 | |||
175 | def websubfilter(text): |
|
175 | def websubfilter(text): | |
176 | return templatefilters.websub(text, self.websubtable) |
|
176 | return templatefilters.websub(text, self.websubtable) | |
177 |
|
177 | |||
178 | # create the templater |
|
178 | # create the templater | |
179 | # TODO: export all keywords: defaults = templatekw.keywords.copy() |
|
179 | # TODO: export all keywords: defaults = templatekw.keywords.copy() | |
180 | defaults = { |
|
180 | defaults = { | |
181 | 'url': req.apppath + '/', |
|
181 | 'url': req.apppath + '/', | |
182 | 'logourl': logourl, |
|
182 | 'logourl': logourl, | |
183 | 'logoimg': logoimg, |
|
183 | 'logoimg': logoimg, | |
184 | 'staticurl': staticurl, |
|
184 | 'staticurl': staticurl, | |
185 | 'urlbase': req.advertisedbaseurl, |
|
185 | 'urlbase': req.advertisedbaseurl, | |
186 | 'repo': self.reponame, |
|
186 | 'repo': self.reponame, | |
187 | 'encoding': encoding.encoding, |
|
187 | 'encoding': encoding.encoding, | |
188 | 'motd': motd, |
|
188 | 'motd': motd, | |
189 | 'sessionvars': sessionvars, |
|
189 | 'sessionvars': sessionvars, | |
190 | 'pathdef': makebreadcrumb(req.apppath), |
|
190 | 'pathdef': makebreadcrumb(req.apppath), | |
191 | 'style': style, |
|
191 | 'style': style, | |
192 | 'nonce': self.nonce, |
|
192 | 'nonce': self.nonce, | |
193 | } |
|
193 | } | |
194 | tres = formatter.templateresources(self.repo.ui, self.repo) |
|
194 | tres = formatter.templateresources(self.repo.ui, self.repo) | |
195 | tmpl = templater.templater.frommapfile(mapfile, |
|
195 | tmpl = templater.templater.frommapfile(mapfile, | |
196 | filters={'websub': websubfilter}, |
|
196 | filters={'websub': websubfilter}, | |
197 | defaults=defaults, |
|
197 | defaults=defaults, | |
198 | resources=tres) |
|
198 | resources=tres) | |
199 | return tmpl |
|
199 | return tmpl | |
200 |
|
200 | |||
201 |
|
201 | |||
202 | class hgweb(object): |
|
202 | class hgweb(object): | |
203 | """HTTP server for individual repositories. |
|
203 | """HTTP server for individual repositories. | |
204 |
|
204 | |||
205 | Instances of this class serve HTTP responses for a particular |
|
205 | Instances of this class serve HTTP responses for a particular | |
206 | repository. |
|
206 | repository. | |
207 |
|
207 | |||
208 | Instances are typically used as WSGI applications. |
|
208 | Instances are typically used as WSGI applications. | |
209 |
|
209 | |||
210 | Some servers are multi-threaded. On these servers, there may |
|
210 | Some servers are multi-threaded. On these servers, there may | |
211 | be multiple active threads inside __call__. |
|
211 | be multiple active threads inside __call__. | |
212 | """ |
|
212 | """ | |
213 | def __init__(self, repo, name=None, baseui=None): |
|
213 | def __init__(self, repo, name=None, baseui=None): | |
214 | if isinstance(repo, str): |
|
214 | if isinstance(repo, str): | |
215 | if baseui: |
|
215 | if baseui: | |
216 | u = baseui.copy() |
|
216 | u = baseui.copy() | |
217 | else: |
|
217 | else: | |
218 | u = uimod.ui.load() |
|
218 | u = uimod.ui.load() | |
219 | r = hg.repository(u, repo) |
|
219 | r = hg.repository(u, repo) | |
220 | else: |
|
220 | else: | |
221 | # we trust caller to give us a private copy |
|
221 | # we trust caller to give us a private copy | |
222 | r = repo |
|
222 | r = repo | |
223 |
|
223 | |||
224 | r.ui.setconfig('ui', 'report_untrusted', 'off', 'hgweb') |
|
224 | r.ui.setconfig('ui', 'report_untrusted', 'off', 'hgweb') | |
225 | r.baseui.setconfig('ui', 'report_untrusted', 'off', 'hgweb') |
|
225 | r.baseui.setconfig('ui', 'report_untrusted', 'off', 'hgweb') | |
226 | r.ui.setconfig('ui', 'nontty', 'true', 'hgweb') |
|
226 | r.ui.setconfig('ui', 'nontty', 'true', 'hgweb') | |
227 | r.baseui.setconfig('ui', 'nontty', 'true', 'hgweb') |
|
227 | r.baseui.setconfig('ui', 'nontty', 'true', 'hgweb') | |
228 | # resolve file patterns relative to repo root |
|
228 | # resolve file patterns relative to repo root | |
229 | r.ui.setconfig('ui', 'forcecwd', r.root, 'hgweb') |
|
229 | r.ui.setconfig('ui', 'forcecwd', r.root, 'hgweb') | |
230 | r.baseui.setconfig('ui', 'forcecwd', r.root, 'hgweb') |
|
230 | r.baseui.setconfig('ui', 'forcecwd', r.root, 'hgweb') | |
231 | # displaying bundling progress bar while serving feel wrong and may |
|
231 | # displaying bundling progress bar while serving feel wrong and may | |
232 | # break some wsgi implementation. |
|
232 | # break some wsgi implementation. | |
233 | r.ui.setconfig('progress', 'disable', 'true', 'hgweb') |
|
233 | r.ui.setconfig('progress', 'disable', 'true', 'hgweb') | |
234 | r.baseui.setconfig('progress', 'disable', 'true', 'hgweb') |
|
234 | r.baseui.setconfig('progress', 'disable', 'true', 'hgweb') | |
235 | self._repos = [hg.cachedlocalrepo(self._webifyrepo(r))] |
|
235 | self._repos = [hg.cachedlocalrepo(self._webifyrepo(r))] | |
236 | self._lastrepo = self._repos[0] |
|
236 | self._lastrepo = self._repos[0] | |
237 | hook.redirect(True) |
|
237 | hook.redirect(True) | |
238 | self.reponame = name |
|
238 | self.reponame = name | |
239 |
|
239 | |||
240 | def _webifyrepo(self, repo): |
|
240 | def _webifyrepo(self, repo): | |
241 | repo = getwebview(repo) |
|
241 | repo = getwebview(repo) | |
242 | self.websubtable = webutil.getwebsubs(repo) |
|
242 | self.websubtable = webutil.getwebsubs(repo) | |
243 | return repo |
|
243 | return repo | |
244 |
|
244 | |||
245 | @contextlib.contextmanager |
|
245 | @contextlib.contextmanager | |
246 | def _obtainrepo(self): |
|
246 | def _obtainrepo(self): | |
247 | """Obtain a repo unique to the caller. |
|
247 | """Obtain a repo unique to the caller. | |
248 |
|
248 | |||
249 | Internally we maintain a stack of cachedlocalrepo instances |
|
249 | Internally we maintain a stack of cachedlocalrepo instances | |
250 | to be handed out. If one is available, we pop it and return it, |
|
250 | to be handed out. If one is available, we pop it and return it, | |
251 | ensuring it is up to date in the process. If one is not available, |
|
251 | ensuring it is up to date in the process. If one is not available, | |
252 | we clone the most recently used repo instance and return it. |
|
252 | we clone the most recently used repo instance and return it. | |
253 |
|
253 | |||
254 | It is currently possible for the stack to grow without bounds |
|
254 | It is currently possible for the stack to grow without bounds | |
255 | if the server allows infinite threads. However, servers should |
|
255 | if the server allows infinite threads. However, servers should | |
256 | have a thread limit, thus establishing our limit. |
|
256 | have a thread limit, thus establishing our limit. | |
257 | """ |
|
257 | """ | |
258 | if self._repos: |
|
258 | if self._repos: | |
259 | cached = self._repos.pop() |
|
259 | cached = self._repos.pop() | |
260 | r, created = cached.fetch() |
|
260 | r, created = cached.fetch() | |
261 | else: |
|
261 | else: | |
262 | cached = self._lastrepo.copy() |
|
262 | cached = self._lastrepo.copy() | |
263 | r, created = cached.fetch() |
|
263 | r, created = cached.fetch() | |
264 | if created: |
|
264 | if created: | |
265 | r = self._webifyrepo(r) |
|
265 | r = self._webifyrepo(r) | |
266 |
|
266 | |||
267 | self._lastrepo = cached |
|
267 | self._lastrepo = cached | |
268 | self.mtime = cached.mtime |
|
268 | self.mtime = cached.mtime | |
269 | try: |
|
269 | try: | |
270 | yield r |
|
270 | yield r | |
271 | finally: |
|
271 | finally: | |
272 | self._repos.append(cached) |
|
272 | self._repos.append(cached) | |
273 |
|
273 | |||
274 | def run(self): |
|
274 | def run(self): | |
275 | """Start a server from CGI environment. |
|
275 | """Start a server from CGI environment. | |
276 |
|
276 | |||
277 | Modern servers should be using WSGI and should avoid this |
|
277 | Modern servers should be using WSGI and should avoid this | |
278 | method, if possible. |
|
278 | method, if possible. | |
279 | """ |
|
279 | """ | |
280 | if not encoding.environ.get('GATEWAY_INTERFACE', |
|
280 | if not encoding.environ.get('GATEWAY_INTERFACE', | |
281 | '').startswith("CGI/1."): |
|
281 | '').startswith("CGI/1."): | |
282 | raise RuntimeError("This function is only intended to be " |
|
282 | raise RuntimeError("This function is only intended to be " | |
283 | "called while running as a CGI script.") |
|
283 | "called while running as a CGI script.") | |
284 | wsgicgi.launch(self) |
|
284 | wsgicgi.launch(self) | |
285 |
|
285 | |||
286 | def __call__(self, env, respond): |
|
286 | def __call__(self, env, respond): | |
287 | """Run the WSGI application. |
|
287 | """Run the WSGI application. | |
288 |
|
288 | |||
289 | This may be called by multiple threads. |
|
289 | This may be called by multiple threads. | |
290 | """ |
|
290 | """ | |
291 | req = requestmod.wsgirequest(env, respond) |
|
291 | req = requestmod.wsgirequest(env, respond) | |
292 | return self.run_wsgi(req) |
|
292 | return self.run_wsgi(req) | |
293 |
|
293 | |||
294 | def run_wsgi(self, wsgireq): |
|
294 | def run_wsgi(self, wsgireq): | |
295 | """Internal method to run the WSGI application. |
|
295 | """Internal method to run the WSGI application. | |
296 |
|
296 | |||
297 | This is typically only called by Mercurial. External consumers |
|
297 | This is typically only called by Mercurial. External consumers | |
298 | should be using instances of this class as the WSGI application. |
|
298 | should be using instances of this class as the WSGI application. | |
299 | """ |
|
299 | """ | |
300 | with self._obtainrepo() as repo: |
|
300 | with self._obtainrepo() as repo: | |
301 | profile = repo.ui.configbool('profiling', 'enabled') |
|
301 | profile = repo.ui.configbool('profiling', 'enabled') | |
302 | with profiling.profile(repo.ui, enabled=profile): |
|
302 | with profiling.profile(repo.ui, enabled=profile): | |
303 | for r in self._runwsgi(wsgireq, repo): |
|
303 | for r in self._runwsgi(wsgireq, repo): | |
304 | yield r |
|
304 | yield r | |
305 |
|
305 | |||
306 | def _runwsgi(self, wsgireq, repo): |
|
306 | def _runwsgi(self, wsgireq, repo): | |
307 | req = requestmod.parserequestfromenv(wsgireq.env) |
|
307 | req = requestmod.parserequestfromenv(wsgireq.env) | |
308 | rctx = requestcontext(self, repo) |
|
308 | rctx = requestcontext(self, repo) | |
309 |
|
309 | |||
310 | # This state is global across all threads. |
|
310 | # This state is global across all threads. | |
311 | encoding.encoding = rctx.config('web', 'encoding') |
|
311 | encoding.encoding = rctx.config('web', 'encoding') | |
312 | rctx.repo.ui.environ = wsgireq.env |
|
312 | rctx.repo.ui.environ = wsgireq.env | |
313 |
|
313 | |||
314 | if rctx.csp: |
|
314 | if rctx.csp: | |
315 | # hgwebdir may have added CSP header. Since we generate our own, |
|
315 | # hgwebdir may have added CSP header. Since we generate our own, | |
316 | # replace it. |
|
316 | # replace it. | |
317 | wsgireq.headers = [h for h in wsgireq.headers |
|
317 | wsgireq.headers = [h for h in wsgireq.headers | |
318 | if h[0] != 'Content-Security-Policy'] |
|
318 | if h[0] != 'Content-Security-Policy'] | |
319 | wsgireq.headers.append(('Content-Security-Policy', rctx.csp)) |
|
319 | wsgireq.headers.append(('Content-Security-Policy', rctx.csp)) | |
320 |
|
320 | |||
321 | handled, res = wireprotoserver.handlewsgirequest( |
|
321 | handled, res = wireprotoserver.handlewsgirequest( | |
322 | rctx, wsgireq, req, self.check_perm) |
|
322 | rctx, wsgireq, req, self.check_perm) | |
323 | if handled: |
|
323 | if handled: | |
324 | return res |
|
324 | return res | |
325 |
|
325 | |||
326 | if req.havepathinfo: |
|
326 | if req.havepathinfo: | |
327 | query = req.dispatchpath |
|
327 | query = req.dispatchpath | |
328 | else: |
|
328 | else: | |
329 | query = req.querystring.partition('&')[0].partition(';')[0] |
|
329 | query = req.querystring.partition('&')[0].partition(';')[0] | |
330 |
|
330 | |||
331 | # translate user-visible url structure to internal structure |
|
331 | # translate user-visible url structure to internal structure | |
332 |
|
332 | |||
333 | args = query.split('/', 2) |
|
333 | args = query.split('/', 2) | |
334 | if 'cmd' not in wsgireq.form and args and args[0]: |
|
334 | if 'cmd' not in wsgireq.form and args and args[0]: | |
335 | cmd = args.pop(0) |
|
335 | cmd = args.pop(0) | |
336 | style = cmd.rfind('-') |
|
336 | style = cmd.rfind('-') | |
337 | if style != -1: |
|
337 | if style != -1: | |
338 | wsgireq.form['style'] = [cmd[:style]] |
|
338 | wsgireq.form['style'] = [cmd[:style]] | |
339 | cmd = cmd[style + 1:] |
|
339 | cmd = cmd[style + 1:] | |
340 |
|
340 | |||
341 | # avoid accepting e.g. style parameter as command |
|
341 | # avoid accepting e.g. style parameter as command | |
342 | if util.safehasattr(webcommands, cmd): |
|
342 | if util.safehasattr(webcommands, cmd): | |
343 | wsgireq.form['cmd'] = [cmd] |
|
343 | wsgireq.form['cmd'] = [cmd] | |
344 |
|
344 | |||
345 | if cmd == 'static': |
|
345 | if cmd == 'static': | |
346 | wsgireq.form['file'] = ['/'.join(args)] |
|
346 | wsgireq.form['file'] = ['/'.join(args)] | |
347 | else: |
|
347 | else: | |
348 | if args and args[0]: |
|
348 | if args and args[0]: | |
349 | node = args.pop(0).replace('%2F', '/') |
|
349 | node = args.pop(0).replace('%2F', '/') | |
350 | wsgireq.form['node'] = [node] |
|
350 | wsgireq.form['node'] = [node] | |
351 | if args: |
|
351 | if args: | |
352 | wsgireq.form['file'] = args |
|
352 | wsgireq.form['file'] = args | |
353 |
|
353 | |||
354 |
ua = |
|
354 | ua = req.headers.get('User-Agent', '') | |
355 | if cmd == 'rev' and 'mercurial' in ua: |
|
355 | if cmd == 'rev' and 'mercurial' in ua: | |
356 | wsgireq.form['style'] = ['raw'] |
|
356 | wsgireq.form['style'] = ['raw'] | |
357 |
|
357 | |||
358 | if cmd == 'archive': |
|
358 | if cmd == 'archive': | |
359 | fn = wsgireq.form['node'][0] |
|
359 | fn = wsgireq.form['node'][0] | |
360 | for type_, spec in rctx.archivespecs.iteritems(): |
|
360 | for type_, spec in rctx.archivespecs.iteritems(): | |
361 | ext = spec[2] |
|
361 | ext = spec[2] | |
362 | if fn.endswith(ext): |
|
362 | if fn.endswith(ext): | |
363 | wsgireq.form['node'] = [fn[:-len(ext)]] |
|
363 | wsgireq.form['node'] = [fn[:-len(ext)]] | |
364 | wsgireq.form['type'] = [type_] |
|
364 | wsgireq.form['type'] = [type_] | |
365 | else: |
|
365 | else: | |
366 | cmd = wsgireq.form.get('cmd', [''])[0] |
|
366 | cmd = wsgireq.form.get('cmd', [''])[0] | |
367 |
|
367 | |||
368 | # process the web interface request |
|
368 | # process the web interface request | |
369 |
|
369 | |||
370 | try: |
|
370 | try: | |
371 | tmpl = rctx.templater(wsgireq, req) |
|
371 | tmpl = rctx.templater(wsgireq, req) | |
372 | ctype = tmpl('mimetype', encoding=encoding.encoding) |
|
372 | ctype = tmpl('mimetype', encoding=encoding.encoding) | |
373 | ctype = templater.stringify(ctype) |
|
373 | ctype = templater.stringify(ctype) | |
374 |
|
374 | |||
375 | # check read permissions non-static content |
|
375 | # check read permissions non-static content | |
376 | if cmd != 'static': |
|
376 | if cmd != 'static': | |
377 | self.check_perm(rctx, wsgireq, None) |
|
377 | self.check_perm(rctx, wsgireq, None) | |
378 |
|
378 | |||
379 | if cmd == '': |
|
379 | if cmd == '': | |
380 | wsgireq.form['cmd'] = [tmpl.cache['default']] |
|
380 | wsgireq.form['cmd'] = [tmpl.cache['default']] | |
381 | cmd = wsgireq.form['cmd'][0] |
|
381 | cmd = wsgireq.form['cmd'][0] | |
382 |
|
382 | |||
383 | # Don't enable caching if using a CSP nonce because then it wouldn't |
|
383 | # Don't enable caching if using a CSP nonce because then it wouldn't | |
384 | # be a nonce. |
|
384 | # be a nonce. | |
385 | if rctx.configbool('web', 'cache') and not rctx.nonce: |
|
385 | if rctx.configbool('web', 'cache') and not rctx.nonce: | |
386 | caching(self, wsgireq) # sets ETag header or raises NOT_MODIFIED |
|
386 | caching(self, wsgireq) # sets ETag header or raises NOT_MODIFIED | |
387 | if cmd not in webcommands.__all__: |
|
387 | if cmd not in webcommands.__all__: | |
388 | msg = 'no such method: %s' % cmd |
|
388 | msg = 'no such method: %s' % cmd | |
389 | raise ErrorResponse(HTTP_BAD_REQUEST, msg) |
|
389 | raise ErrorResponse(HTTP_BAD_REQUEST, msg) | |
390 | elif cmd == 'file' and 'raw' in wsgireq.form.get('style', []): |
|
390 | elif cmd == 'file' and 'raw' in wsgireq.form.get('style', []): | |
391 | rctx.ctype = ctype |
|
391 | rctx.ctype = ctype | |
392 | content = webcommands.rawfile(rctx, wsgireq, tmpl) |
|
392 | content = webcommands.rawfile(rctx, wsgireq, tmpl) | |
393 | else: |
|
393 | else: | |
394 | content = getattr(webcommands, cmd)(rctx, wsgireq, tmpl) |
|
394 | content = getattr(webcommands, cmd)(rctx, wsgireq, tmpl) | |
395 | wsgireq.respond(HTTP_OK, ctype) |
|
395 | wsgireq.respond(HTTP_OK, ctype) | |
396 |
|
396 | |||
397 | return content |
|
397 | return content | |
398 |
|
398 | |||
399 | except (error.LookupError, error.RepoLookupError) as err: |
|
399 | except (error.LookupError, error.RepoLookupError) as err: | |
400 | wsgireq.respond(HTTP_NOT_FOUND, ctype) |
|
400 | wsgireq.respond(HTTP_NOT_FOUND, ctype) | |
401 | msg = pycompat.bytestr(err) |
|
401 | msg = pycompat.bytestr(err) | |
402 | if (util.safehasattr(err, 'name') and |
|
402 | if (util.safehasattr(err, 'name') and | |
403 | not isinstance(err, error.ManifestLookupError)): |
|
403 | not isinstance(err, error.ManifestLookupError)): | |
404 | msg = 'revision not found: %s' % err.name |
|
404 | msg = 'revision not found: %s' % err.name | |
405 | return tmpl('error', error=msg) |
|
405 | return tmpl('error', error=msg) | |
406 | except (error.RepoError, error.RevlogError) as inst: |
|
406 | except (error.RepoError, error.RevlogError) as inst: | |
407 | wsgireq.respond(HTTP_SERVER_ERROR, ctype) |
|
407 | wsgireq.respond(HTTP_SERVER_ERROR, ctype) | |
408 | return tmpl('error', error=pycompat.bytestr(inst)) |
|
408 | return tmpl('error', error=pycompat.bytestr(inst)) | |
409 | except ErrorResponse as inst: |
|
409 | except ErrorResponse as inst: | |
410 | wsgireq.respond(inst, ctype) |
|
410 | wsgireq.respond(inst, ctype) | |
411 | if inst.code == HTTP_NOT_MODIFIED: |
|
411 | if inst.code == HTTP_NOT_MODIFIED: | |
412 | # Not allowed to return a body on a 304 |
|
412 | # Not allowed to return a body on a 304 | |
413 | return [''] |
|
413 | return [''] | |
414 | return tmpl('error', error=pycompat.bytestr(inst)) |
|
414 | return tmpl('error', error=pycompat.bytestr(inst)) | |
415 |
|
415 | |||
416 | def check_perm(self, rctx, req, op): |
|
416 | def check_perm(self, rctx, req, op): | |
417 | for permhook in permhooks: |
|
417 | for permhook in permhooks: | |
418 | permhook(rctx, req, op) |
|
418 | permhook(rctx, req, op) | |
419 |
|
419 | |||
420 | def getwebview(repo): |
|
420 | def getwebview(repo): | |
421 | """The 'web.view' config controls changeset filter to hgweb. Possible |
|
421 | """The 'web.view' config controls changeset filter to hgweb. Possible | |
422 | values are ``served``, ``visible`` and ``all``. Default is ``served``. |
|
422 | values are ``served``, ``visible`` and ``all``. Default is ``served``. | |
423 | The ``served`` filter only shows changesets that can be pulled from the |
|
423 | The ``served`` filter only shows changesets that can be pulled from the | |
424 | hgweb instance. The``visible`` filter includes secret changesets but |
|
424 | hgweb instance. The``visible`` filter includes secret changesets but | |
425 | still excludes "hidden" one. |
|
425 | still excludes "hidden" one. | |
426 |
|
426 | |||
427 | See the repoview module for details. |
|
427 | See the repoview module for details. | |
428 |
|
428 | |||
429 | The option has been around undocumented since Mercurial 2.5, but no |
|
429 | The option has been around undocumented since Mercurial 2.5, but no | |
430 | user ever asked about it. So we better keep it undocumented for now.""" |
|
430 | user ever asked about it. So we better keep it undocumented for now.""" | |
431 | # experimental config: web.view |
|
431 | # experimental config: web.view | |
432 | viewconfig = repo.ui.config('web', 'view', untrusted=True) |
|
432 | viewconfig = repo.ui.config('web', 'view', untrusted=True) | |
433 | if viewconfig == 'all': |
|
433 | if viewconfig == 'all': | |
434 | return repo.unfiltered() |
|
434 | return repo.unfiltered() | |
435 | elif viewconfig in repoview.filtertable: |
|
435 | elif viewconfig in repoview.filtertable: | |
436 | return repo.filtered(viewconfig) |
|
436 | return repo.filtered(viewconfig) | |
437 | else: |
|
437 | else: | |
438 | return repo.filtered('served') |
|
438 | return repo.filtered('served') |
@@ -1,300 +1,315 | |||||
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 cgi |
|
11 | import cgi | |
12 | import errno |
|
12 | import errno | |
13 | import socket |
|
13 | import socket | |
|
14 | import wsgiref.headers as wsgiheaders | |||
14 | #import wsgiref.validate |
|
15 | #import wsgiref.validate | |
15 |
|
16 | |||
16 | from .common import ( |
|
17 | from .common import ( | |
17 | ErrorResponse, |
|
18 | ErrorResponse, | |
18 | HTTP_NOT_MODIFIED, |
|
19 | HTTP_NOT_MODIFIED, | |
19 | statusmessage, |
|
20 | statusmessage, | |
20 | ) |
|
21 | ) | |
21 |
|
22 | |||
22 | from ..thirdparty import ( |
|
23 | from ..thirdparty import ( | |
23 | attr, |
|
24 | attr, | |
24 | ) |
|
25 | ) | |
25 | from .. import ( |
|
26 | from .. import ( | |
26 | pycompat, |
|
27 | pycompat, | |
27 | util, |
|
28 | util, | |
28 | ) |
|
29 | ) | |
29 |
|
30 | |||
30 | shortcuts = { |
|
31 | shortcuts = { | |
31 | 'cl': [('cmd', ['changelog']), ('rev', None)], |
|
32 | 'cl': [('cmd', ['changelog']), ('rev', None)], | |
32 | 'sl': [('cmd', ['shortlog']), ('rev', None)], |
|
33 | 'sl': [('cmd', ['shortlog']), ('rev', None)], | |
33 | 'cs': [('cmd', ['changeset']), ('node', None)], |
|
34 | 'cs': [('cmd', ['changeset']), ('node', None)], | |
34 | 'f': [('cmd', ['file']), ('filenode', None)], |
|
35 | 'f': [('cmd', ['file']), ('filenode', None)], | |
35 | 'fl': [('cmd', ['filelog']), ('filenode', None)], |
|
36 | 'fl': [('cmd', ['filelog']), ('filenode', None)], | |
36 | 'fd': [('cmd', ['filediff']), ('node', None)], |
|
37 | 'fd': [('cmd', ['filediff']), ('node', None)], | |
37 | 'fa': [('cmd', ['annotate']), ('filenode', None)], |
|
38 | 'fa': [('cmd', ['annotate']), ('filenode', None)], | |
38 | 'mf': [('cmd', ['manifest']), ('manifest', None)], |
|
39 | 'mf': [('cmd', ['manifest']), ('manifest', None)], | |
39 | 'ca': [('cmd', ['archive']), ('node', None)], |
|
40 | 'ca': [('cmd', ['archive']), ('node', None)], | |
40 | 'tags': [('cmd', ['tags'])], |
|
41 | 'tags': [('cmd', ['tags'])], | |
41 | 'tip': [('cmd', ['changeset']), ('node', ['tip'])], |
|
42 | 'tip': [('cmd', ['changeset']), ('node', ['tip'])], | |
42 | 'static': [('cmd', ['static']), ('file', None)] |
|
43 | 'static': [('cmd', ['static']), ('file', None)] | |
43 | } |
|
44 | } | |
44 |
|
45 | |||
45 | def normalize(form): |
|
46 | def normalize(form): | |
46 | # first expand the shortcuts |
|
47 | # first expand the shortcuts | |
47 | for k in shortcuts: |
|
48 | for k in shortcuts: | |
48 | if k in form: |
|
49 | if k in form: | |
49 | for name, value in shortcuts[k]: |
|
50 | for name, value in shortcuts[k]: | |
50 | if value is None: |
|
51 | if value is None: | |
51 | value = form[k] |
|
52 | value = form[k] | |
52 | form[name] = value |
|
53 | form[name] = value | |
53 | del form[k] |
|
54 | del form[k] | |
54 | # And strip the values |
|
55 | # And strip the values | |
55 | bytesform = {} |
|
56 | bytesform = {} | |
56 | for k, v in form.iteritems(): |
|
57 | for k, v in form.iteritems(): | |
57 | bytesform[pycompat.bytesurl(k)] = [ |
|
58 | bytesform[pycompat.bytesurl(k)] = [ | |
58 | pycompat.bytesurl(i.strip()) for i in v] |
|
59 | pycompat.bytesurl(i.strip()) for i in v] | |
59 | return bytesform |
|
60 | return bytesform | |
60 |
|
61 | |||
61 | @attr.s(frozen=True) |
|
62 | @attr.s(frozen=True) | |
62 | class parsedrequest(object): |
|
63 | class parsedrequest(object): | |
63 | """Represents a parsed WSGI request / static HTTP request parameters.""" |
|
64 | """Represents a parsed WSGI request / static HTTP request parameters.""" | |
64 |
|
65 | |||
65 | # Full URL for this request. |
|
66 | # Full URL for this request. | |
66 | url = attr.ib() |
|
67 | url = attr.ib() | |
67 | # URL without any path components. Just <proto>://<host><port>. |
|
68 | # URL without any path components. Just <proto>://<host><port>. | |
68 | baseurl = attr.ib() |
|
69 | baseurl = attr.ib() | |
69 | # Advertised URL. Like ``url`` and ``baseurl`` but uses SERVER_NAME instead |
|
70 | # Advertised URL. Like ``url`` and ``baseurl`` but uses SERVER_NAME instead | |
70 | # of HTTP: Host header for hostname. This is likely what clients used. |
|
71 | # of HTTP: Host header for hostname. This is likely what clients used. | |
71 | advertisedurl = attr.ib() |
|
72 | advertisedurl = attr.ib() | |
72 | advertisedbaseurl = attr.ib() |
|
73 | advertisedbaseurl = attr.ib() | |
73 | # WSGI application path. |
|
74 | # WSGI application path. | |
74 | apppath = attr.ib() |
|
75 | apppath = attr.ib() | |
75 | # List of path parts to be used for dispatch. |
|
76 | # List of path parts to be used for dispatch. | |
76 | dispatchparts = attr.ib() |
|
77 | dispatchparts = attr.ib() | |
77 | # URL path component (no query string) used for dispatch. |
|
78 | # URL path component (no query string) used for dispatch. | |
78 | dispatchpath = attr.ib() |
|
79 | dispatchpath = attr.ib() | |
79 | # Whether there is a path component to this request. This can be true |
|
80 | # Whether there is a path component to this request. This can be true | |
80 | # when ``dispatchpath`` is empty due to REPO_NAME muckery. |
|
81 | # when ``dispatchpath`` is empty due to REPO_NAME muckery. | |
81 | havepathinfo = attr.ib() |
|
82 | havepathinfo = attr.ib() | |
82 | # Raw query string (part after "?" in URL). |
|
83 | # Raw query string (part after "?" in URL). | |
83 | querystring = attr.ib() |
|
84 | querystring = attr.ib() | |
84 | # List of 2-tuples of query string arguments. |
|
85 | # List of 2-tuples of query string arguments. | |
85 | querystringlist = attr.ib() |
|
86 | querystringlist = attr.ib() | |
86 | # Dict of query string arguments. Values are lists with at least 1 item. |
|
87 | # Dict of query string arguments. Values are lists with at least 1 item. | |
87 | querystringdict = attr.ib() |
|
88 | querystringdict = attr.ib() | |
|
89 | # wsgiref.headers.Headers instance. Operates like a dict with case | |||
|
90 | # insensitive keys. | |||
|
91 | headers = attr.ib() | |||
88 |
|
92 | |||
89 | def parserequestfromenv(env): |
|
93 | def parserequestfromenv(env): | |
90 | """Parse URL components from environment variables. |
|
94 | """Parse URL components from environment variables. | |
91 |
|
95 | |||
92 | WSGI defines request attributes via environment variables. This function |
|
96 | WSGI defines request attributes via environment variables. This function | |
93 | parses the environment variables into a data structure. |
|
97 | parses the environment variables into a data structure. | |
94 | """ |
|
98 | """ | |
95 | # PEP-0333 defines the WSGI spec and is a useful reference for this code. |
|
99 | # PEP-0333 defines the WSGI spec and is a useful reference for this code. | |
96 |
|
100 | |||
97 | # We first validate that the incoming object conforms with the WSGI spec. |
|
101 | # We first validate that the incoming object conforms with the WSGI spec. | |
98 | # We only want to be dealing with spec-conforming WSGI implementations. |
|
102 | # We only want to be dealing with spec-conforming WSGI implementations. | |
99 | # TODO enable this once we fix internal violations. |
|
103 | # TODO enable this once we fix internal violations. | |
100 | #wsgiref.validate.check_environ(env) |
|
104 | #wsgiref.validate.check_environ(env) | |
101 |
|
105 | |||
102 | # PEP-0333 states that environment keys and values are native strings |
|
106 | # PEP-0333 states that environment keys and values are native strings | |
103 | # (bytes on Python 2 and str on Python 3). The code points for the Unicode |
|
107 | # (bytes on Python 2 and str on Python 3). The code points for the Unicode | |
104 | # strings on Python 3 must be between \00000-\000FF. We deal with bytes |
|
108 | # strings on Python 3 must be between \00000-\000FF. We deal with bytes | |
105 | # in Mercurial, so mass convert string keys and values to bytes. |
|
109 | # in Mercurial, so mass convert string keys and values to bytes. | |
106 | if pycompat.ispy3: |
|
110 | if pycompat.ispy3: | |
107 | env = {k.encode('latin-1'): v for k, v in env.iteritems()} |
|
111 | env = {k.encode('latin-1'): v for k, v in env.iteritems()} | |
108 | env = {k: v.encode('latin-1') if isinstance(v, str) else v |
|
112 | env = {k: v.encode('latin-1') if isinstance(v, str) else v | |
109 | for k, v in env.iteritems()} |
|
113 | for k, v in env.iteritems()} | |
110 |
|
114 | |||
111 | # https://www.python.org/dev/peps/pep-0333/#environ-variables defines |
|
115 | # https://www.python.org/dev/peps/pep-0333/#environ-variables defines | |
112 | # the environment variables. |
|
116 | # the environment variables. | |
113 | # https://www.python.org/dev/peps/pep-0333/#url-reconstruction defines |
|
117 | # https://www.python.org/dev/peps/pep-0333/#url-reconstruction defines | |
114 | # how URLs are reconstructed. |
|
118 | # how URLs are reconstructed. | |
115 | fullurl = env['wsgi.url_scheme'] + '://' |
|
119 | fullurl = env['wsgi.url_scheme'] + '://' | |
116 | advertisedfullurl = fullurl |
|
120 | advertisedfullurl = fullurl | |
117 |
|
121 | |||
118 | def addport(s): |
|
122 | def addport(s): | |
119 | if env['wsgi.url_scheme'] == 'https': |
|
123 | if env['wsgi.url_scheme'] == 'https': | |
120 | if env['SERVER_PORT'] != '443': |
|
124 | if env['SERVER_PORT'] != '443': | |
121 | s += ':' + env['SERVER_PORT'] |
|
125 | s += ':' + env['SERVER_PORT'] | |
122 | else: |
|
126 | else: | |
123 | if env['SERVER_PORT'] != '80': |
|
127 | if env['SERVER_PORT'] != '80': | |
124 | s += ':' + env['SERVER_PORT'] |
|
128 | s += ':' + env['SERVER_PORT'] | |
125 |
|
129 | |||
126 | return s |
|
130 | return s | |
127 |
|
131 | |||
128 | if env.get('HTTP_HOST'): |
|
132 | if env.get('HTTP_HOST'): | |
129 | fullurl += env['HTTP_HOST'] |
|
133 | fullurl += env['HTTP_HOST'] | |
130 | else: |
|
134 | else: | |
131 | fullurl += env['SERVER_NAME'] |
|
135 | fullurl += env['SERVER_NAME'] | |
132 | fullurl = addport(fullurl) |
|
136 | fullurl = addport(fullurl) | |
133 |
|
137 | |||
134 | advertisedfullurl += env['SERVER_NAME'] |
|
138 | advertisedfullurl += env['SERVER_NAME'] | |
135 | advertisedfullurl = addport(advertisedfullurl) |
|
139 | advertisedfullurl = addport(advertisedfullurl) | |
136 |
|
140 | |||
137 | baseurl = fullurl |
|
141 | baseurl = fullurl | |
138 | advertisedbaseurl = advertisedfullurl |
|
142 | advertisedbaseurl = advertisedfullurl | |
139 |
|
143 | |||
140 | fullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) |
|
144 | fullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) | |
141 | advertisedfullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) |
|
145 | advertisedfullurl += util.urlreq.quote(env.get('SCRIPT_NAME', '')) | |
142 | fullurl += util.urlreq.quote(env.get('PATH_INFO', '')) |
|
146 | fullurl += util.urlreq.quote(env.get('PATH_INFO', '')) | |
143 | advertisedfullurl += util.urlreq.quote(env.get('PATH_INFO', '')) |
|
147 | advertisedfullurl += util.urlreq.quote(env.get('PATH_INFO', '')) | |
144 |
|
148 | |||
145 | if env.get('QUERY_STRING'): |
|
149 | if env.get('QUERY_STRING'): | |
146 | fullurl += '?' + env['QUERY_STRING'] |
|
150 | fullurl += '?' + env['QUERY_STRING'] | |
147 | advertisedfullurl += '?' + env['QUERY_STRING'] |
|
151 | advertisedfullurl += '?' + env['QUERY_STRING'] | |
148 |
|
152 | |||
149 | # When dispatching requests, we look at the URL components (PATH_INFO |
|
153 | # When dispatching requests, we look at the URL components (PATH_INFO | |
150 | # and QUERY_STRING) after the application root (SCRIPT_NAME). But hgwebdir |
|
154 | # and QUERY_STRING) after the application root (SCRIPT_NAME). But hgwebdir | |
151 | # has the concept of "virtual" repositories. This is defined via REPO_NAME. |
|
155 | # has the concept of "virtual" repositories. This is defined via REPO_NAME. | |
152 | # If REPO_NAME is defined, we append it to SCRIPT_NAME to form a new app |
|
156 | # If REPO_NAME is defined, we append it to SCRIPT_NAME to form a new app | |
153 | # root. We also exclude its path components from PATH_INFO when resolving |
|
157 | # root. We also exclude its path components from PATH_INFO when resolving | |
154 | # the dispatch path. |
|
158 | # the dispatch path. | |
155 |
|
159 | |||
156 | apppath = env['SCRIPT_NAME'] |
|
160 | apppath = env['SCRIPT_NAME'] | |
157 |
|
161 | |||
158 | if env.get('REPO_NAME'): |
|
162 | if env.get('REPO_NAME'): | |
159 | if not apppath.endswith('/'): |
|
163 | if not apppath.endswith('/'): | |
160 | apppath += '/' |
|
164 | apppath += '/' | |
161 |
|
165 | |||
162 | apppath += env.get('REPO_NAME') |
|
166 | apppath += env.get('REPO_NAME') | |
163 |
|
167 | |||
164 | if 'PATH_INFO' in env: |
|
168 | if 'PATH_INFO' in env: | |
165 | dispatchparts = env['PATH_INFO'].strip('/').split('/') |
|
169 | dispatchparts = env['PATH_INFO'].strip('/').split('/') | |
166 |
|
170 | |||
167 | # Strip out repo parts. |
|
171 | # Strip out repo parts. | |
168 | repoparts = env.get('REPO_NAME', '').split('/') |
|
172 | repoparts = env.get('REPO_NAME', '').split('/') | |
169 | if dispatchparts[:len(repoparts)] == repoparts: |
|
173 | if dispatchparts[:len(repoparts)] == repoparts: | |
170 | dispatchparts = dispatchparts[len(repoparts):] |
|
174 | dispatchparts = dispatchparts[len(repoparts):] | |
171 | else: |
|
175 | else: | |
172 | dispatchparts = [] |
|
176 | dispatchparts = [] | |
173 |
|
177 | |||
174 | dispatchpath = '/'.join(dispatchparts) |
|
178 | dispatchpath = '/'.join(dispatchparts) | |
175 |
|
179 | |||
176 | querystring = env.get('QUERY_STRING', '') |
|
180 | querystring = env.get('QUERY_STRING', '') | |
177 |
|
181 | |||
178 | # We store as a list so we have ordering information. We also store as |
|
182 | # We store as a list so we have ordering information. We also store as | |
179 | # a dict to facilitate fast lookup. |
|
183 | # a dict to facilitate fast lookup. | |
180 | querystringlist = util.urlreq.parseqsl(querystring, keep_blank_values=True) |
|
184 | querystringlist = util.urlreq.parseqsl(querystring, keep_blank_values=True) | |
181 |
|
185 | |||
182 | querystringdict = {} |
|
186 | querystringdict = {} | |
183 | for k, v in querystringlist: |
|
187 | for k, v in querystringlist: | |
184 | if k in querystringdict: |
|
188 | if k in querystringdict: | |
185 | querystringdict[k].append(v) |
|
189 | querystringdict[k].append(v) | |
186 | else: |
|
190 | else: | |
187 | querystringdict[k] = [v] |
|
191 | querystringdict[k] = [v] | |
188 |
|
192 | |||
|
193 | # HTTP_* keys contain HTTP request headers. The Headers structure should | |||
|
194 | # perform case normalization for us. We just rewrite underscore to dash | |||
|
195 | # so keys match what likely went over the wire. | |||
|
196 | headers = [] | |||
|
197 | for k, v in env.iteritems(): | |||
|
198 | if k.startswith('HTTP_'): | |||
|
199 | headers.append((k[len('HTTP_'):].replace('_', '-'), v)) | |||
|
200 | ||||
|
201 | headers = wsgiheaders.Headers(headers) | |||
|
202 | ||||
189 | return parsedrequest(url=fullurl, baseurl=baseurl, |
|
203 | return parsedrequest(url=fullurl, baseurl=baseurl, | |
190 | advertisedurl=advertisedfullurl, |
|
204 | advertisedurl=advertisedfullurl, | |
191 | advertisedbaseurl=advertisedbaseurl, |
|
205 | advertisedbaseurl=advertisedbaseurl, | |
192 | apppath=apppath, |
|
206 | apppath=apppath, | |
193 | dispatchparts=dispatchparts, dispatchpath=dispatchpath, |
|
207 | dispatchparts=dispatchparts, dispatchpath=dispatchpath, | |
194 | havepathinfo='PATH_INFO' in env, |
|
208 | havepathinfo='PATH_INFO' in env, | |
195 | querystring=querystring, |
|
209 | querystring=querystring, | |
196 | querystringlist=querystringlist, |
|
210 | querystringlist=querystringlist, | |
197 |
querystringdict=querystringdict |
|
211 | querystringdict=querystringdict, | |
|
212 | headers=headers) | |||
198 |
|
213 | |||
199 | class wsgirequest(object): |
|
214 | class wsgirequest(object): | |
200 | """Higher-level API for a WSGI request. |
|
215 | """Higher-level API for a WSGI request. | |
201 |
|
216 | |||
202 | WSGI applications are invoked with 2 arguments. They are used to |
|
217 | WSGI applications are invoked with 2 arguments. They are used to | |
203 | instantiate instances of this class, which provides higher-level APIs |
|
218 | instantiate instances of this class, which provides higher-level APIs | |
204 | for obtaining request parameters, writing HTTP output, etc. |
|
219 | for obtaining request parameters, writing HTTP output, etc. | |
205 | """ |
|
220 | """ | |
206 | def __init__(self, wsgienv, start_response): |
|
221 | def __init__(self, wsgienv, start_response): | |
207 | version = wsgienv[r'wsgi.version'] |
|
222 | version = wsgienv[r'wsgi.version'] | |
208 | if (version < (1, 0)) or (version >= (2, 0)): |
|
223 | if (version < (1, 0)) or (version >= (2, 0)): | |
209 | raise RuntimeError("Unknown and unsupported WSGI version %d.%d" |
|
224 | raise RuntimeError("Unknown and unsupported WSGI version %d.%d" | |
210 | % version) |
|
225 | % version) | |
211 | self.inp = wsgienv[r'wsgi.input'] |
|
226 | self.inp = wsgienv[r'wsgi.input'] | |
212 | self.err = wsgienv[r'wsgi.errors'] |
|
227 | self.err = wsgienv[r'wsgi.errors'] | |
213 | self.threaded = wsgienv[r'wsgi.multithread'] |
|
228 | self.threaded = wsgienv[r'wsgi.multithread'] | |
214 | self.multiprocess = wsgienv[r'wsgi.multiprocess'] |
|
229 | self.multiprocess = wsgienv[r'wsgi.multiprocess'] | |
215 | self.run_once = wsgienv[r'wsgi.run_once'] |
|
230 | self.run_once = wsgienv[r'wsgi.run_once'] | |
216 | self.env = wsgienv |
|
231 | self.env = wsgienv | |
217 | self.form = normalize(cgi.parse(self.inp, |
|
232 | self.form = normalize(cgi.parse(self.inp, | |
218 | self.env, |
|
233 | self.env, | |
219 | keep_blank_values=1)) |
|
234 | keep_blank_values=1)) | |
220 | self._start_response = start_response |
|
235 | self._start_response = start_response | |
221 | self.server_write = None |
|
236 | self.server_write = None | |
222 | self.headers = [] |
|
237 | self.headers = [] | |
223 |
|
238 | |||
224 | def __iter__(self): |
|
239 | def __iter__(self): | |
225 | return iter([]) |
|
240 | return iter([]) | |
226 |
|
241 | |||
227 | def read(self, count=-1): |
|
242 | def read(self, count=-1): | |
228 | return self.inp.read(count) |
|
243 | return self.inp.read(count) | |
229 |
|
244 | |||
230 | def drain(self): |
|
245 | def drain(self): | |
231 | '''need to read all data from request, httplib is half-duplex''' |
|
246 | '''need to read all data from request, httplib is half-duplex''' | |
232 | length = int(self.env.get('CONTENT_LENGTH') or 0) |
|
247 | length = int(self.env.get('CONTENT_LENGTH') or 0) | |
233 | for s in util.filechunkiter(self.inp, limit=length): |
|
248 | for s in util.filechunkiter(self.inp, limit=length): | |
234 | pass |
|
249 | pass | |
235 |
|
250 | |||
236 | def respond(self, status, type, filename=None, body=None): |
|
251 | def respond(self, status, type, filename=None, body=None): | |
237 | if not isinstance(type, str): |
|
252 | if not isinstance(type, str): | |
238 | type = pycompat.sysstr(type) |
|
253 | type = pycompat.sysstr(type) | |
239 | if self._start_response is not None: |
|
254 | if self._start_response is not None: | |
240 | self.headers.append((r'Content-Type', type)) |
|
255 | self.headers.append((r'Content-Type', type)) | |
241 | if filename: |
|
256 | if filename: | |
242 | filename = (filename.rpartition('/')[-1] |
|
257 | filename = (filename.rpartition('/')[-1] | |
243 | .replace('\\', '\\\\').replace('"', '\\"')) |
|
258 | .replace('\\', '\\\\').replace('"', '\\"')) | |
244 | self.headers.append(('Content-Disposition', |
|
259 | self.headers.append(('Content-Disposition', | |
245 | 'inline; filename="%s"' % filename)) |
|
260 | 'inline; filename="%s"' % filename)) | |
246 | if body is not None: |
|
261 | if body is not None: | |
247 | self.headers.append((r'Content-Length', str(len(body)))) |
|
262 | self.headers.append((r'Content-Length', str(len(body)))) | |
248 |
|
263 | |||
249 | for k, v in self.headers: |
|
264 | for k, v in self.headers: | |
250 | if not isinstance(v, str): |
|
265 | if not isinstance(v, str): | |
251 | raise TypeError('header value must be string: %r' % (v,)) |
|
266 | raise TypeError('header value must be string: %r' % (v,)) | |
252 |
|
267 | |||
253 | if isinstance(status, ErrorResponse): |
|
268 | if isinstance(status, ErrorResponse): | |
254 | self.headers.extend(status.headers) |
|
269 | self.headers.extend(status.headers) | |
255 | if status.code == HTTP_NOT_MODIFIED: |
|
270 | if status.code == HTTP_NOT_MODIFIED: | |
256 | # RFC 2616 Section 10.3.5: 304 Not Modified has cases where |
|
271 | # RFC 2616 Section 10.3.5: 304 Not Modified has cases where | |
257 | # it MUST NOT include any headers other than these and no |
|
272 | # it MUST NOT include any headers other than these and no | |
258 | # body |
|
273 | # body | |
259 | self.headers = [(k, v) for (k, v) in self.headers if |
|
274 | self.headers = [(k, v) for (k, v) in self.headers if | |
260 | k in ('Date', 'ETag', 'Expires', |
|
275 | k in ('Date', 'ETag', 'Expires', | |
261 | 'Cache-Control', 'Vary')] |
|
276 | 'Cache-Control', 'Vary')] | |
262 | status = statusmessage(status.code, pycompat.bytestr(status)) |
|
277 | status = statusmessage(status.code, pycompat.bytestr(status)) | |
263 | elif status == 200: |
|
278 | elif status == 200: | |
264 | status = '200 Script output follows' |
|
279 | status = '200 Script output follows' | |
265 | elif isinstance(status, int): |
|
280 | elif isinstance(status, int): | |
266 | status = statusmessage(status) |
|
281 | status = statusmessage(status) | |
267 |
|
282 | |||
268 | self.server_write = self._start_response( |
|
283 | self.server_write = self._start_response( | |
269 | pycompat.sysstr(status), self.headers) |
|
284 | pycompat.sysstr(status), self.headers) | |
270 | self._start_response = None |
|
285 | self._start_response = None | |
271 | self.headers = [] |
|
286 | self.headers = [] | |
272 | if body is not None: |
|
287 | if body is not None: | |
273 | self.write(body) |
|
288 | self.write(body) | |
274 | self.server_write = None |
|
289 | self.server_write = None | |
275 |
|
290 | |||
276 | def write(self, thing): |
|
291 | def write(self, thing): | |
277 | if thing: |
|
292 | if thing: | |
278 | try: |
|
293 | try: | |
279 | self.server_write(thing) |
|
294 | self.server_write(thing) | |
280 | except socket.error as inst: |
|
295 | except socket.error as inst: | |
281 | if inst[0] != errno.ECONNRESET: |
|
296 | if inst[0] != errno.ECONNRESET: | |
282 | raise |
|
297 | raise | |
283 |
|
298 | |||
284 | def writelines(self, lines): |
|
299 | def writelines(self, lines): | |
285 | for line in lines: |
|
300 | for line in lines: | |
286 | self.write(line) |
|
301 | self.write(line) | |
287 |
|
302 | |||
288 | def flush(self): |
|
303 | def flush(self): | |
289 | return None |
|
304 | return None | |
290 |
|
305 | |||
291 | def close(self): |
|
306 | def close(self): | |
292 | return None |
|
307 | return None | |
293 |
|
308 | |||
294 | def wsgiapplication(app_maker): |
|
309 | def wsgiapplication(app_maker): | |
295 | '''For compatibility with old CGI scripts. A plain hgweb() or hgwebdir() |
|
310 | '''For compatibility with old CGI scripts. A plain hgweb() or hgwebdir() | |
296 | can and should now be used as a WSGI application.''' |
|
311 | can and should now be used as a WSGI application.''' | |
297 | application = app_maker() |
|
312 | application = app_maker() | |
298 | def run_wsgi(env, respond): |
|
313 | def run_wsgi(env, respond): | |
299 | return application(env, respond) |
|
314 | return application(env, respond) | |
300 | return run_wsgi |
|
315 | return run_wsgi |
General Comments 0
You need to be logged in to leave comments.
Login now