svn: support proxy-prefix properly fixes #5521

rhodecode/lib/middleware/simplesvn.py

@@ -1,225 +1,226 b''
 # -*- coding: utf-8 -*-
 
 # Copyright (C) 2010-2018 RhodeCode GmbH
 #
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU Affero General Public License, version 3
 # (only), as published by the Free Software Foundation.
 #
 # This program is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 # GNU General Public License for more details.
 #
 # You should have received a copy of the GNU Affero General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 #
 # This program is dual-licensed. If you wish to learn more about the
 # RhodeCode Enterprise Edition, including its added features, Support services,
 # and proprietary license terms, please see https://rhodecode.com/licenses/
 
 import base64
 import logging
 import urllib
 import urlparse
 
 import requests
 from pyramid.httpexceptions import HTTPNotAcceptable
 
 from rhodecode.lib import rc_cache
 from rhodecode.lib.middleware import simplevcs
 from rhodecode.lib.utils import is_valid_repo
 from rhodecode.lib.utils2 import str2bool, safe_int
 from rhodecode.lib.ext_json import json
 from rhodecode.lib.hooks_daemon import store_txn_id_data
 
 
 log = logging.getLogger(__name__)
 
 
 class SimpleSvnApp(object):
     IGNORED_HEADERS = [
         'connection', 'keep-alive', 'content-encoding',
         'transfer-encoding', 'content-length']
     rc_extras = {}
 
     def __init__(self, config):
         self.config = config
 
     def __call__(self, environ, start_response):
         request_headers = self._get_request_headers(environ)
         data = environ['wsgi.input']
         req_method = environ['REQUEST_METHOD']
         has_content_length = 'CONTENT_LENGTH' in environ
-        path_info = self._get_url(environ['PATH_INFO'])
+        path_info = self._get_url(
+            self.config.get('subversion_http_server_url', ''), environ['PATH_INFO'])
         transfer_encoding = environ.get('HTTP_TRANSFER_ENCODING', '')
         log.debug('Handling: %s method via `%s`', req_method, path_info)
 
         # stream control flag, based on request and content type...
         stream = False
 
         if req_method in ['MKCOL'] or has_content_length:
             data_processed = False
             # read chunk to check if we have txn-with-props
             initial_data = data.read(1024)
             if initial_data.startswith('(create-txn-with-props'):
                 data = initial_data + data.read()
                 # store on-the-fly our rc_extra using svn revision properties
                 # those can be read later on in hooks executed so we have a way
                 # to pass in the data into svn hooks
                 rc_data = base64.urlsafe_b64encode(json.dumps(self.rc_extras))
                 rc_data_len = len(rc_data)
                 # header defines data length, and serialized data
                 skel = ' rc-scm-extras {} {}'.format(rc_data_len, rc_data)
                 data = data[:-2] + skel + '))'
                 data_processed = True
 
             if not data_processed:
                 # NOTE(johbo): Avoid that we end up with sending the request in chunked
                 # transfer encoding (mainly on Gunicorn). If we know the content
                 # length, then we should transfer the payload in one request.
                 data = initial_data + data.read()
 
         if req_method in ['GET', 'PUT'] or transfer_encoding == 'chunked':
             # NOTE(marcink): when getting/uploading files we want to STREAM content
             # back to the client/proxy instead of buffering it here...
             stream = True
 
         stream = stream
         log.debug(
             'Calling SVN PROXY: method:%s via `%s`, Stream: %s',
             req_method, path_info, stream)
         response = requests.request(
             req_method, path_info,
             data=data, headers=request_headers, stream=stream)
 
         if response.status_code not in [200, 401]:
             if response.status_code >= 500:
                 log.error('Got SVN response:%s with text:\n`%s`',
                           response, response.text)
             else:
                 log.debug('Got SVN response:%s with text:\n`%s`',
                           response, response.text)
         else:
             log.debug('got response code: %s', response.status_code)
 
         response_headers = self._get_response_headers(response.headers)
 
         if response.headers.get('SVN-Txn-name'):
             svn_tx_id = response.headers.get('SVN-Txn-name')
             txn_id = rc_cache.utils.compute_key_from_params(
                 self.config['repository'], svn_tx_id)
             port = safe_int(self.rc_extras['hooks_uri'].split(':')[-1])
             store_txn_id_data(txn_id, {'port': port})
 
         start_response(
             '{} {}'.format(response.status_code, response.reason),
             response_headers)
         return response.iter_content(chunk_size=1024)
 
-    def _get_url(self, path):
-        url_path = urlparse.urljoin(
-            self.config.get('subversion_http_server_url', ''), path)
+    def _get_url(self, svn_http_server, path):
+        svn_http_server_url = (svn_http_server or '').rstrip('/')
+        url_path = urlparse.urljoin(svn_http_server_url + '/', (path or '').lstrip('/'))
         url_path = urllib.quote(url_path, safe="/:=~+!$,;'")
         return url_path
 
     def _get_request_headers(self, environ):
         headers = {}
 
         for key in environ:
             if not key.startswith('HTTP_'):
                 continue
             new_key = key.split('_')
             new_key = [k.capitalize() for k in new_key[1:]]
             new_key = '-'.join(new_key)
             headers[new_key] = environ[key]
 
         if 'CONTENT_TYPE' in environ:
             headers['Content-Type'] = environ['CONTENT_TYPE']
 
         if 'CONTENT_LENGTH' in environ:
             headers['Content-Length'] = environ['CONTENT_LENGTH']
 
         return headers
 
     def _get_response_headers(self, headers):
         headers = [
             (h, headers[h])
             for h in headers
             if h.lower() not in self.IGNORED_HEADERS
         ]
 
         return headers
 
 
 class DisabledSimpleSvnApp(object):
     def __init__(self, config):
         self.config = config
 
     def __call__(self, environ, start_response):
         reason = 'Cannot handle SVN call because: SVN HTTP Proxy is not enabled'
         log.warning(reason)
         return HTTPNotAcceptable(reason)(environ, start_response)
 
 
 class SimpleSvn(simplevcs.SimpleVCS):
 
     SCM = 'svn'
     READ_ONLY_COMMANDS = ('OPTIONS', 'PROPFIND', 'GET', 'REPORT')
     DEFAULT_HTTP_SERVER = 'http://localhost:8090'
 
     def _get_repository_name(self, environ):
         """
         Gets repository name out of PATH_INFO header
 
         :param environ: environ where PATH_INFO is stored
         """
         path = environ['PATH_INFO'].split('!')
         repo_name = path[0].strip('/')
 
         # SVN includes the whole path in it's requests, including
         # subdirectories inside the repo. Therefore we have to search for
         # the repo root directory.
         if not is_valid_repo(
                 repo_name, self.base_path, explicit_scm=self.SCM):
             current_path = ''
             for component in repo_name.split('/'):
                 current_path += component
                 if is_valid_repo(
                         current_path, self.base_path, explicit_scm=self.SCM):
                     return current_path
                 current_path += '/'
 
         return repo_name
 
     def _get_action(self, environ):
         return (
             'pull'
             if environ['REQUEST_METHOD'] in self.READ_ONLY_COMMANDS
             else 'push')
 
     def _should_use_callback_daemon(self, extras, environ, action):
         # only MERGE command triggers hooks, so we don't want to start
         # hooks server too many times. POST however starts the svn transaction
         # so we also need to run the init of callback daemon of POST
         if environ['REQUEST_METHOD'] in ['MERGE', 'POST']:
             return True
         return False
 
     def _create_wsgi_app(self, repo_path, repo_name, config):
         if self._is_svn_enabled():
             return SimpleSvnApp(config)
         # we don't have http proxy enabled return dummy request handler
         return DisabledSimpleSvnApp(config)
 
     def _is_svn_enabled(self):
         conf = self.repo_vcs_config
         return str2bool(conf.get('vcs_svn_proxy', 'http_requests_enabled'))
 
     def _create_config(self, extras, repo_name):
         conf = self.repo_vcs_config
         server_url = conf.get('vcs_svn_proxy', 'http_server_url')
         server_url = server_url or self.DEFAULT_HTTP_SERVER
 
         extras['subversion_http_server_url'] = server_url
         return extras

rhodecode/tests/lib/middleware/test_simplesvn.py

@@ -1,201 +1,209 b''
 # -*- coding: utf-8 -*-
 
 # Copyright (C) 2010-2018 RhodeCode GmbH
 #
 # This program is free software: you can redistribute it and/or modify
 # it under the terms of the GNU Affero General Public License, version 3
 # (only), as published by the Free Software Foundation.
 #
 # This program is distributed in the hope that it will be useful,
 # but WITHOUT ANY WARRANTY; without even the implied warranty of
 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 # GNU General Public License for more details.
 #
 # You should have received a copy of the GNU Affero General Public License
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 #
 # This program is dual-licensed. If you wish to learn more about the
 # RhodeCode Enterprise Edition, including its added features, Support services,
 # and proprietary license terms, please see https://rhodecode.com/licenses/
 
 from StringIO import StringIO
 
 import pytest
 from mock import patch, Mock
 
 from rhodecode.lib.middleware.simplesvn import SimpleSvn, SimpleSvnApp
 from rhodecode.lib.utils import get_rhodecode_base_path
 
 
 class TestSimpleSvn(object):
     @pytest.fixture(autouse=True)
     def simple_svn(self, baseapp, request_stub):
         base_path = get_rhodecode_base_path()
         self.app = SimpleSvn(
             config={'auth_ret_code': '', 'base_path': base_path},
             registry=request_stub.registry)
 
     def test_get_config(self):
         extras = {'foo': 'FOO', 'bar': 'BAR'}
         config = self.app._create_config(extras, repo_name='test-repo')
         assert config == extras
 
     @pytest.mark.parametrize(
         'method', ['OPTIONS', 'PROPFIND', 'GET', 'REPORT'])
     def test_get_action_returns_pull(self, method):
         environment = {'REQUEST_METHOD': method}
         action = self.app._get_action(environment)
         assert action == 'pull'
 
     @pytest.mark.parametrize(
         'method', [
             'MKACTIVITY', 'PROPPATCH', 'PUT', 'CHECKOUT', 'MKCOL', 'MOVE',
             'COPY', 'DELETE', 'LOCK', 'UNLOCK', 'MERGE'
         ])
     def test_get_action_returns_push(self, method):
         environment = {'REQUEST_METHOD': method}
         action = self.app._get_action(environment)
         assert action == 'push'
 
     @pytest.mark.parametrize(
         'path, expected_name', [
             ('/hello-svn', 'hello-svn'),
             ('/hello-svn/', 'hello-svn'),
             ('/group/hello-svn/', 'group/hello-svn'),
             ('/group/hello-svn/!svn/vcc/default', 'group/hello-svn'),
         ])
     def test_get_repository_name(self, path, expected_name):
         environment = {'PATH_INFO': path}
         name = self.app._get_repository_name(environment)
         assert name == expected_name
 
     def test_get_repository_name_subfolder(self, backend_svn):
         repo = backend_svn.repo
         environment = {
             'PATH_INFO': '/{}/path/with/subfolders'.format(repo.repo_name)}
         name = self.app._get_repository_name(environment)
         assert name == repo.repo_name
 
     def test_create_wsgi_app(self):
         with patch.object(SimpleSvn, '_is_svn_enabled') as mock_method:
             mock_method.return_value = False
             with patch('rhodecode.lib.middleware.simplesvn.DisabledSimpleSvnApp') as (
                     wsgi_app_mock):
                 config = Mock()
                 wsgi_app = self.app._create_wsgi_app(
                     repo_path='', repo_name='', config=config)
 
             wsgi_app_mock.assert_called_once_with(config)
             assert wsgi_app == wsgi_app_mock()
 
     def test_create_wsgi_app_when_enabled(self):
         with patch.object(SimpleSvn, '_is_svn_enabled') as mock_method:
             mock_method.return_value = True
             with patch('rhodecode.lib.middleware.simplesvn.SimpleSvnApp') as (
                     wsgi_app_mock):
                 config = Mock()
                 wsgi_app = self.app._create_wsgi_app(
                     repo_path='', repo_name='', config=config)
 
             wsgi_app_mock.assert_called_once_with(config)
             assert wsgi_app == wsgi_app_mock()
 
 
 class TestSimpleSvnApp(object):
     data = '<xml></xml>'
     path = '/group/my-repo'
     wsgi_input = StringIO(data)
     environment = {
         'HTTP_DAV': (
             'http://subversion.tigris.org/xmlns/dav/svn/depth,'
             ' http://subversion.tigris.org/xmlns/dav/svn/mergeinfo'),
         'HTTP_USER_AGENT': 'SVN/1.8.11 (x86_64-linux) serf/1.3.8',
         'REQUEST_METHOD': 'OPTIONS',
         'PATH_INFO': path,
         'wsgi.input': wsgi_input,
         'CONTENT_TYPE': 'text/xml',
         'CONTENT_LENGTH': '130'
     }
 
     def setup_method(self, method):
         self.host = 'http://localhost/'
         base_path = get_rhodecode_base_path()
         self.app = SimpleSvnApp(
             config={'subversion_http_server_url': self.host,
                     'base_path': base_path})
 
     def test_get_request_headers_with_content_type(self):
         expected_headers = {
             'Dav': self.environment['HTTP_DAV'],
             'User-Agent': self.environment['HTTP_USER_AGENT'],
             'Content-Type': self.environment['CONTENT_TYPE'],
             'Content-Length': self.environment['CONTENT_LENGTH']
         }
         headers = self.app._get_request_headers(self.environment)
         assert headers == expected_headers
 
     def test_get_request_headers_without_content_type(self):
         environment = self.environment.copy()
         environment.pop('CONTENT_TYPE')
         expected_headers = {
             'Dav': environment['HTTP_DAV'],
             'Content-Length': self.environment['CONTENT_LENGTH'],
             'User-Agent': environment['HTTP_USER_AGENT'],
         }
         request_headers = self.app._get_request_headers(environment)
         assert request_headers == expected_headers
 
     def test_get_response_headers(self):
         headers = {
             'Connection': 'keep-alive',
             'Keep-Alive': 'timeout=5, max=100',
             'Transfer-Encoding': 'chunked',
             'Content-Encoding': 'gzip',
             'MS-Author-Via': 'DAV',
             'SVN-Supported-Posts': 'create-txn-with-props'
         }
         expected_headers = [
             ('MS-Author-Via', 'DAV'),
             ('SVN-Supported-Posts', 'create-txn-with-props'),
         ]
         response_headers = self.app._get_response_headers(headers)
         assert sorted(response_headers) == sorted(expected_headers)
 
-    def test_get_url(self):
-        url = self.app._get_url(self.path)
-        expected_url = '{}{}'.format(self.host.strip('/'), self.path)
+    @pytest.mark.parametrize('svn_http_url, path_info, expected_url', [
+        ('http://localhost:8200', '/repo_name', 'http://localhost:8200/repo_name'),
+        ('http://localhost:8200///', '/repo_name', 'http://localhost:8200/repo_name'),
+        ('http://localhost:8200', '/group/repo_name', 'http://localhost:8200/group/repo_name'),
+        ('http://localhost:8200/', '/group/repo_name', 'http://localhost:8200/group/repo_name'),
+        ('http://localhost:8200/prefix', '/repo_name', 'http://localhost:8200/prefix/repo_name'),
+        ('http://localhost:8200/prefix', 'repo_name', 'http://localhost:8200/prefix/repo_name'),
+        ('http://localhost:8200/prefix', '/group/repo_name', 'http://localhost:8200/prefix/group/repo_name')
+    ])
+    def test_get_url(self, svn_http_url, path_info, expected_url):
+        url = self.app._get_url(svn_http_url, path_info)
         assert url == expected_url
 
     def test_call(self):
         start_response = Mock()
         response_mock = Mock()
         response_mock.headers = {
             'Content-Encoding': 'gzip',
             'MS-Author-Via': 'DAV',
             'SVN-Supported-Posts': 'create-txn-with-props'
         }
         response_mock.status_code = 200
         response_mock.reason = 'OK'
         with patch('rhodecode.lib.middleware.simplesvn.requests.request') as (
                 request_mock):
             request_mock.return_value = response_mock
             self.app(self.environment, start_response)
 
         expected_url = '{}{}'.format(self.host.strip('/'), self.path)
         expected_request_headers = {
             'Dav': self.environment['HTTP_DAV'],
             'User-Agent': self.environment['HTTP_USER_AGENT'],
             'Content-Type': self.environment['CONTENT_TYPE'],
             'Content-Length': self.environment['CONTENT_LENGTH']
         }
         expected_response_headers = [
             ('SVN-Supported-Posts', 'create-txn-with-props'),
             ('MS-Author-Via', 'DAV'),
         ]
         request_mock.assert_called_once_with(
             self.environment['REQUEST_METHOD'], expected_url,
             data=self.data, headers=expected_request_headers, stream=False)
         response_mock.iter_content.assert_called_once_with(chunk_size=1024)
         args, _ = start_response.call_args
         assert args[0] == '200 OK'
         assert sorted(args[1]) == sorted(expected_response_headers)