Show More
@@ -1,242 +1,242 b'' | |||
|
1 | 1 | |
|
2 | 2 | # Copyright (C) 2010-2023 RhodeCode GmbH |
|
3 | 3 | # |
|
4 | 4 | # This program is free software: you can redistribute it and/or modify |
|
5 | 5 | # it under the terms of the GNU Affero General Public License, version 3 |
|
6 | 6 | # (only), as published by the Free Software Foundation. |
|
7 | 7 | # |
|
8 | 8 | # This program is distributed in the hope that it will be useful, |
|
9 | 9 | # but WITHOUT ANY WARRANTY; without even the implied warranty of |
|
10 | 10 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
|
11 | 11 | # GNU General Public License for more details. |
|
12 | 12 | # |
|
13 | 13 | # You should have received a copy of the GNU Affero General Public License |
|
14 | 14 | # along with this program. If not, see <http://www.gnu.org/licenses/>. |
|
15 | 15 | # |
|
16 | 16 | # This program is dual-licensed. If you wish to learn more about the |
|
17 | 17 | # RhodeCode Enterprise Edition, including its added features, Support services, |
|
18 | 18 | # and proprietary license terms, please see https://rhodecode.com/licenses/ |
|
19 | 19 | |
|
20 | 20 | import base64 |
|
21 | 21 | import logging |
|
22 | 22 | import urllib.request |
|
23 | 23 | import urllib.parse |
|
24 | 24 | import urllib.error |
|
25 | 25 | import urllib.parse |
|
26 | 26 | |
|
27 | 27 | import requests |
|
28 | 28 | from pyramid.httpexceptions import HTTPNotAcceptable |
|
29 | 29 | |
|
30 | 30 | from rhodecode import ConfigGet |
|
31 | 31 | from rhodecode.lib import rc_cache |
|
32 | 32 | from rhodecode.lib.middleware import simplevcs |
|
33 | 33 | from rhodecode.lib.middleware.utils import get_path_info |
|
34 | 34 | from rhodecode.lib.utils import is_valid_repo |
|
35 | 35 | from rhodecode.lib.str_utils import safe_str, safe_int, safe_bytes |
|
36 | 36 | from rhodecode.lib.type_utils import str2bool |
|
37 | 37 | from rhodecode.lib.ext_json import json |
|
38 | 38 | from rhodecode.lib.hook_daemon.base import store_txn_id_data |
|
39 | 39 | |
|
40 | 40 | log = logging.getLogger(__name__) |
|
41 | 41 | |
|
42 | 42 | |
|
43 | 43 | class SimpleSvnApp(object): |
|
44 | 44 | IGNORED_HEADERS = [ |
|
45 | 45 | 'connection', 'keep-alive', 'content-encoding', |
|
46 | 46 | 'transfer-encoding', 'content-length'] |
|
47 | 47 | rc_extras = {} |
|
48 | 48 | |
|
49 | 49 | def __init__(self, config): |
|
50 | 50 | self.config = config |
|
51 | 51 | self.session = requests.Session() |
|
52 | 52 | |
|
53 | 53 | def __call__(self, environ, start_response): |
|
54 | 54 | request_headers = self._get_request_headers(environ) |
|
55 | 55 | data_io = environ['wsgi.input'] |
|
56 | 56 | req_method: str = environ['REQUEST_METHOD'] |
|
57 | has_content_length = 'CONTENT_LENGTH' in environ | |
|
57 | has_content_length: bool = 'CONTENT_LENGTH' in environ | |
|
58 | 58 | |
|
59 | 59 | path_info = self._get_url( |
|
60 | 60 | self.config.get('subversion_http_server_url', ''), get_path_info(environ)) |
|
61 | 61 | transfer_encoding = environ.get('HTTP_TRANSFER_ENCODING', '') |
|
62 | log.debug('Handling: %s method via `%s`', req_method, path_info) | |
|
62 | log.debug('Handling: %s method via `%s` has_content_length:%s', req_method, path_info, has_content_length) | |
|
63 | 63 | |
|
64 | 64 | # stream control flag, based on request and content type... |
|
65 | 65 | stream = False |
|
66 | 66 | |
|
67 | 67 | if req_method in ['MKCOL'] or has_content_length: |
|
68 | 68 | data_processed = False |
|
69 | 69 | # read chunk to check if we have txn-with-props |
|
70 | 70 | initial_data: bytes = data_io.read(1024) |
|
71 | 71 | if initial_data.startswith(b'(create-txn-with-props'): |
|
72 | 72 | data_io = initial_data + data_io.read() |
|
73 | 73 | # store on-the-fly our rc_extra using svn revision properties |
|
74 | 74 | # those can be read later on in hooks executed so we have a way |
|
75 | 75 | # to pass in the data into svn hooks |
|
76 | 76 | rc_data = base64.urlsafe_b64encode(json.dumps(self.rc_extras)) |
|
77 | 77 | rc_data_len = str(len(rc_data)) |
|
78 | 78 | # header defines data length, and serialized data |
|
79 | 79 | skel = b' rc-scm-extras %b %b' % (safe_bytes(rc_data_len), safe_bytes(rc_data)) |
|
80 | 80 | data_io = data_io[:-2] + skel + b'))' |
|
81 | 81 | data_processed = True |
|
82 | 82 | |
|
83 | 83 | if not data_processed: |
|
84 | 84 | # NOTE(johbo): Avoid that we end up with sending the request in chunked |
|
85 | 85 | # transfer encoding (mainly on Gunicorn). If we know the content |
|
86 | 86 | # length, then we should transfer the payload in one request. |
|
87 | 87 | data_io = initial_data + data_io.read() |
|
88 | 88 | |
|
89 | 89 | if req_method in ['GET', 'PUT'] or transfer_encoding == 'chunked': |
|
90 | 90 | # NOTE(marcink): when getting/uploading files, we want to STREAM content |
|
91 | 91 | # back to the client/proxy instead of buffering it here... |
|
92 | 92 | stream = True |
|
93 | 93 | |
|
94 | 94 | stream = stream |
|
95 | 95 | log.debug('Calling SVN PROXY at `%s`, using method:%s. Stream: %s', |
|
96 | 96 | path_info, req_method, stream) |
|
97 | 97 | |
|
98 | 98 | call_kwargs = dict( |
|
99 | 99 | data=data_io, |
|
100 | 100 | headers=request_headers, |
|
101 | 101 | stream=stream |
|
102 | 102 | ) |
|
103 | 103 | if req_method in ['HEAD', 'DELETE']: |
|
104 | 104 | del call_kwargs['data'] |
|
105 | 105 | |
|
106 | 106 | try: |
|
107 | 107 | response = self.session.request( |
|
108 | 108 | req_method, path_info, **call_kwargs) |
|
109 | 109 | except requests.ConnectionError: |
|
110 | 110 | log.exception('ConnectionError occurred for endpoint %s', path_info) |
|
111 | 111 | raise |
|
112 | 112 | |
|
113 | 113 | if response.status_code not in [200, 401]: |
|
114 | 114 | text = '\n{}'.format(safe_str(response.text)) if response.text else '' |
|
115 | 115 | if response.status_code >= 500: |
|
116 | 116 | log.error('Got SVN response:%s with text:`%s`', response, text) |
|
117 | 117 | else: |
|
118 | 118 | log.debug('Got SVN response:%s with text:`%s`', response, text) |
|
119 | 119 | else: |
|
120 | 120 | log.debug('got response code: %s', response.status_code) |
|
121 | 121 | |
|
122 | 122 | response_headers = self._get_response_headers(response.headers) |
|
123 | 123 | |
|
124 | 124 | if response.headers.get('SVN-Txn-name'): |
|
125 | 125 | svn_tx_id = response.headers.get('SVN-Txn-name') |
|
126 | 126 | txn_id = rc_cache.utils.compute_key_from_params( |
|
127 | 127 | self.config['repository'], svn_tx_id) |
|
128 | 128 | port = safe_int(self.rc_extras['hooks_uri'].split(':')[-1]) |
|
129 | 129 | store_txn_id_data(txn_id, {'port': port}) |
|
130 | 130 | |
|
131 | 131 | start_response(f'{response.status_code} {response.reason}', response_headers) |
|
132 | 132 | return response.iter_content(chunk_size=1024) |
|
133 | 133 | |
|
134 | 134 | def _get_url(self, svn_http_server, path): |
|
135 | 135 | svn_http_server_url = (svn_http_server or '').rstrip('/') |
|
136 | 136 | url_path = urllib.parse.urljoin(svn_http_server_url + '/', (path or '').lstrip('/')) |
|
137 | 137 | url_path = urllib.parse.quote(url_path, safe="/:=~+!$,;'") |
|
138 | 138 | return url_path |
|
139 | 139 | |
|
140 | 140 | def _get_request_headers(self, environ): |
|
141 | 141 | headers = {} |
|
142 | 142 | whitelist = { |
|
143 | 143 | 'Authorization': {} |
|
144 | 144 | } |
|
145 | 145 | for key in environ: |
|
146 | 146 | if key in whitelist: |
|
147 | 147 | headers[key] = environ[key] |
|
148 | 148 | elif not key.startswith('HTTP_'): |
|
149 | 149 | continue |
|
150 | 150 | else: |
|
151 | 151 | new_key = key.split('_') |
|
152 | 152 | new_key = [k.capitalize() for k in new_key[1:]] |
|
153 | 153 | new_key = '-'.join(new_key) |
|
154 | 154 | headers[new_key] = environ[key] |
|
155 | 155 | |
|
156 | 156 | if 'CONTENT_TYPE' in environ: |
|
157 | 157 | headers['Content-Type'] = environ['CONTENT_TYPE'] |
|
158 | 158 | |
|
159 | 159 | if 'CONTENT_LENGTH' in environ: |
|
160 | 160 | headers['Content-Length'] = environ['CONTENT_LENGTH'] |
|
161 | 161 | |
|
162 | 162 | return headers |
|
163 | 163 | |
|
164 | 164 | def _get_response_headers(self, headers): |
|
165 | 165 | headers = [ |
|
166 | 166 | (h, headers[h]) |
|
167 | 167 | for h in headers |
|
168 | 168 | if h.lower() not in self.IGNORED_HEADERS |
|
169 | 169 | ] |
|
170 | 170 | |
|
171 | 171 | return headers |
|
172 | 172 | |
|
173 | 173 | |
|
174 | 174 | class DisabledSimpleSvnApp(object): |
|
175 | 175 | def __init__(self, config): |
|
176 | 176 | self.config = config |
|
177 | 177 | |
|
178 | 178 | def __call__(self, environ, start_response): |
|
179 | 179 | reason = 'Cannot handle SVN call because: SVN HTTP Proxy is not enabled' |
|
180 | 180 | log.warning(reason) |
|
181 | 181 | return HTTPNotAcceptable(reason)(environ, start_response) |
|
182 | 182 | |
|
183 | 183 | |
|
184 | 184 | class SimpleSvn(simplevcs.SimpleVCS): |
|
185 | 185 | |
|
186 | 186 | SCM = 'svn' |
|
187 | 187 | READ_ONLY_COMMANDS = ('OPTIONS', 'PROPFIND', 'GET', 'REPORT') |
|
188 | 188 | DEFAULT_HTTP_SERVER = 'http://localhost:8090' |
|
189 | 189 | |
|
190 | 190 | def _get_repository_name(self, environ): |
|
191 | 191 | """ |
|
192 | 192 | Gets repository name out of PATH_INFO header |
|
193 | 193 | |
|
194 | 194 | :param environ: environ where PATH_INFO is stored |
|
195 | 195 | """ |
|
196 | 196 | path = get_path_info(environ).split('!') |
|
197 | 197 | repo_name = path[0].strip('/') |
|
198 | 198 | |
|
199 | 199 | # SVN includes the whole path in it's requests, including |
|
200 | 200 | # subdirectories inside the repo. Therefore we have to search for |
|
201 | 201 | # the repo root directory. |
|
202 | 202 | if not is_valid_repo( |
|
203 | 203 | repo_name, self.base_path, explicit_scm=self.SCM): |
|
204 | 204 | current_path = '' |
|
205 | 205 | for component in repo_name.split('/'): |
|
206 | 206 | current_path += component |
|
207 | 207 | if is_valid_repo( |
|
208 | 208 | current_path, self.base_path, explicit_scm=self.SCM): |
|
209 | 209 | return current_path |
|
210 | 210 | current_path += '/' |
|
211 | 211 | |
|
212 | 212 | return repo_name |
|
213 | 213 | |
|
214 | 214 | def _get_action(self, environ): |
|
215 | 215 | return ( |
|
216 | 216 | 'pull' |
|
217 | 217 | if environ['REQUEST_METHOD'] in self.READ_ONLY_COMMANDS |
|
218 | 218 | else 'push') |
|
219 | 219 | |
|
220 | 220 | def _should_use_callback_daemon(self, extras, environ, action): |
|
221 | 221 | # only MERGE command triggers hooks, so we don't want to start |
|
222 | 222 | # hooks server too many times. POST however starts the svn transaction |
|
223 | 223 | # so we also need to run the init of callback daemon of POST |
|
224 | 224 | if environ['REQUEST_METHOD'] in ['MERGE', 'POST']: |
|
225 | 225 | return True |
|
226 | 226 | return False |
|
227 | 227 | |
|
228 | 228 | def _create_wsgi_app(self, repo_path, repo_name, config): |
|
229 | 229 | if self._is_svn_enabled(): |
|
230 | 230 | return SimpleSvnApp(config) |
|
231 | 231 | # we don't have http proxy enabled return dummy request handler |
|
232 | 232 | return DisabledSimpleSvnApp(config) |
|
233 | 233 | |
|
234 | 234 | def _is_svn_enabled(self): |
|
235 | 235 | return ConfigGet().get_bool('vcs.svn.proxy.enabled') |
|
236 | 236 | |
|
237 | 237 | def _create_config(self, extras, repo_name, scheme='http'): |
|
238 | 238 | server_url = ConfigGet().get_str('vcs.svn.proxy.host') |
|
239 | 239 | server_url = server_url or self.DEFAULT_HTTP_SERVER |
|
240 | 240 | |
|
241 | 241 | extras['subversion_http_server_url'] = server_url |
|
242 | 242 | return extras |
General Comments 0
You need to be logged in to leave comments.
Login now