~bzr-pqm/bzr/bzr.dev

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
# Copyright (C) 2006-2010 Canonical Ltd
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# 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 General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA

import errno
import httplib
import os
import posixpath
import random
import re
import select
import SimpleHTTPServer
import socket
import SocketServer
import sys
import threading
import time
import urllib
import urlparse

from bzrlib import (
    osutils,
    tests,
    transport,
    )
from bzrlib.tests import test_server
from bzrlib.transport import local


class BadWebserverPath(ValueError):
    def __str__(self):
        return 'path %s is not in %s' % self.args


class TestingHTTPRequestHandler(SimpleHTTPServer.SimpleHTTPRequestHandler):
    """Handles one request.

    A TestingHTTPRequestHandler is instantiated for every request received by
    the associated server. Note that 'request' here is inherited from the base
    TCPServer class, for the HTTP server it is really a connection which itself
    will handle one or several HTTP requests.
    """
    # Default protocol version
    protocol_version = 'HTTP/1.1'

    # The Message-like class used to parse the request headers
    MessageClass = httplib.HTTPMessage

    def setup(self):
        SimpleHTTPServer.SimpleHTTPRequestHandler.setup(self)
        self._cwd = self.server._home_dir
        tcs = self.server.test_case_server
        if tcs.protocol_version is not None:
            # If the test server forced a protocol version, use it
            self.protocol_version = tcs.protocol_version

    def log_message(self, format, *args):
        tcs = self.server.test_case_server
        tcs.log('webserver - %s - - [%s] %s "%s" "%s"',
                self.address_string(),
                self.log_date_time_string(),
                format % args,
                self.headers.get('referer', '-'),
                self.headers.get('user-agent', '-'))

    def handle(self):
        SimpleHTTPServer.SimpleHTTPRequestHandler.handle(self)
        # Some client (pycurl, I'm looking at you) are more picky than others
        # and require that the socket itself is closed
        # (SocketServer.StreamRequestHandler only close the two associated
        # 'makefile' objects)
        self.connection.close()

    def handle_one_request(self):
        """Handle a single HTTP request.

        We catch all socket errors occurring when the client close the
        connection early to avoid polluting the test results.
        """
        try:
            self._handle_one_request()
        except socket.error, e:
            # Any socket error should close the connection, but some errors are
            # due to the client closing early and we don't want to pollute test
            # results, so we raise only the others.
            self.close_connection = 1
            if (len(e.args) == 0
                or e.args[0] not in (errno.EPIPE, errno.ECONNRESET,
                                     errno.ECONNABORTED, errno.EBADF)):
                raise

    error_content_type = 'text/plain'
    error_message_format = '''\
Error code: %(code)s.
Message: %(message)s.
'''

    def send_error(self, code, message=None):
        """Send and log an error reply.

        We redefine the python-provided version to be able to set a 
        ``Content-Length`` header as some http/1.1 clients complain otherwise
        (see bug #568421).

        :param code: The HTTP error code.

        :param message: The explanation of the error code, Defaults to a short
             entry.
        """

        if message is None:
            try:
                message = self.responses[code][0]
            except KeyError:
                message = '???'
        self.log_error("code %d, message %s", code, message)
        content = (self.error_message_format %
                   {'code': code, 'message': message})
        self.send_response(code, message)
        self.send_header("Content-Type", self.error_content_type)
        self.send_header("Content-Length", "%d" % len(content))
        self.send_header('Connection', 'close')
        self.end_headers()
        if self.command != 'HEAD' and code >= 200 and code not in (204, 304):
            self.wfile.write(content)

    def _handle_one_request(self):
        SimpleHTTPServer.SimpleHTTPRequestHandler.handle_one_request(self)

    _range_regexp = re.compile(r'^(?P<start>\d+)-(?P<end>\d+)$')
    _tail_regexp = re.compile(r'^-(?P<tail>\d+)$')

    def parse_ranges(self, ranges_header):
        """Parse the range header value and returns ranges and tail.

        RFC2616 14.35 says that syntactically invalid range
        specifiers MUST be ignored. In that case, we return 0 for
        tail and [] for ranges.
        """
        tail = 0
        ranges = []
        if not ranges_header.startswith('bytes='):
            # Syntactically invalid header
            return 0, []

        ranges_header = ranges_header[len('bytes='):]
        for range_str in ranges_header.split(','):
            # FIXME: RFC2616 says end is optional and default to file_size
            range_match = self._range_regexp.match(range_str)
            if range_match is not None:
                start = int(range_match.group('start'))
                end = int(range_match.group('end'))
                if start > end:
                    # Syntactically invalid range
                    return 0, []
                ranges.append((start, end))
            else:
                tail_match = self._tail_regexp.match(range_str)
                if tail_match is not None:
                    tail = int(tail_match.group('tail'))
                else:
                    # Syntactically invalid range
                    return 0, []
        return tail, ranges

    def _header_line_length(self, keyword, value):
        header_line = '%s: %s\r\n' % (keyword, value)
        return len(header_line)

    def send_head(self):
        """Overrides base implementation to work around a bug in python2.5."""
        path = self.translate_path(self.path)
        if os.path.isdir(path) and not self.path.endswith('/'):
            # redirect browser - doing basically what apache does when
            # DirectorySlash option is On which is quite common (braindead, but
            # common)
            self.send_response(301)
            self.send_header("Location", self.path + "/")
            # Indicates that the body is empty for HTTP/1.1 clients
            self.send_header('Content-Length', '0')
            self.end_headers()
            return None

        return SimpleHTTPServer.SimpleHTTPRequestHandler.send_head(self)

    def send_range_content(self, file, start, length):
        file.seek(start)
        self.wfile.write(file.read(length))

    def get_single_range(self, file, file_size, start, end):
        self.send_response(206)
        length = end - start + 1
        self.send_header('Accept-Ranges', 'bytes')
        self.send_header("Content-Length", "%d" % length)

        self.send_header("Content-Type", 'application/octet-stream')
        self.send_header("Content-Range", "bytes %d-%d/%d" % (start,
                                                              end,
                                                              file_size))
        self.end_headers()
        self.send_range_content(file, start, length)

    def get_multiple_ranges(self, file, file_size, ranges):
        self.send_response(206)
        self.send_header('Accept-Ranges', 'bytes')
        boundary = '%d' % random.randint(0,0x7FFFFFFF)
        self.send_header('Content-Type',
                         'multipart/byteranges; boundary=%s' % boundary)
        boundary_line = '--%s\r\n' % boundary
        # Calculate the Content-Length
        content_length = 0
        for (start, end) in ranges:
            content_length += len(boundary_line)
            content_length += self._header_line_length(
                'Content-type', 'application/octet-stream')
            content_length += self._header_line_length(
                'Content-Range', 'bytes %d-%d/%d' % (start, end, file_size))
            content_length += len('\r\n') # end headers
            content_length += end - start + 1
        content_length += len(boundary_line)
        self.send_header('Content-length', content_length)
        self.end_headers()

        # Send the multipart body
        for (start, end) in ranges:
            self.wfile.write(boundary_line)
            self.send_header('Content-type', 'application/octet-stream')
            self.send_header('Content-Range', 'bytes %d-%d/%d'
                             % (start, end, file_size))
            self.end_headers()
            self.send_range_content(file, start, end - start + 1)
        # Final boundary
        self.wfile.write(boundary_line)

    def do_GET(self):
        """Serve a GET request.

        Handles the Range header.
        """
        # Update statistics
        self.server.test_case_server.GET_request_nb += 1

        path = self.translate_path(self.path)
        ranges_header_value = self.headers.get('Range')
        if ranges_header_value is None or os.path.isdir(path):
            # Let the mother class handle most cases
            return SimpleHTTPServer.SimpleHTTPRequestHandler.do_GET(self)

        try:
            # Always read in binary mode. Opening files in text
            # mode may cause newline translations, making the
            # actual size of the content transmitted *less* than
            # the content-length!
            f = open(path, 'rb')
        except IOError:
            self.send_error(404, "File not found")
            return

        file_size = os.fstat(f.fileno())[6]
        tail, ranges = self.parse_ranges(ranges_header_value)
        # Normalize tail into ranges
        if tail != 0:
            ranges.append((file_size - tail, file_size))

        self._satisfiable_ranges = True
        if len(ranges) == 0:
            self._satisfiable_ranges = False
        else:
            def check_range(range_specifier):
                start, end = range_specifier
                # RFC2616 14.35, ranges are invalid if start >= file_size
                if start >= file_size:
                    self._satisfiable_ranges = False # Side-effect !
                    return 0, 0
                # RFC2616 14.35, end values should be truncated
                # to file_size -1 if they exceed it
                end = min(end, file_size - 1)
                return start, end

            ranges = map(check_range, ranges)

        if not self._satisfiable_ranges:
            # RFC2616 14.16 and 14.35 says that when a server
            # encounters unsatisfiable range specifiers, it
            # SHOULD return a 416.
            f.close()
            # FIXME: We SHOULD send a Content-Range header too,
            # but the implementation of send_error does not
            # allows that. So far.
            self.send_error(416, "Requested range not satisfiable")
            return

        if len(ranges) == 1:
            (start, end) = ranges[0]
            self.get_single_range(f, file_size, start, end)
        else:
            self.get_multiple_ranges(f, file_size, ranges)
        f.close()

    def translate_path(self, path):
        """Translate a /-separated PATH to the local filename syntax.

        If the server requires it, proxy the path before the usual translation
        """
        if self.server.test_case_server.proxy_requests:
            # We need to act as a proxy and accept absolute urls,
            # which SimpleHTTPRequestHandler (parent) is not
            # ready for. So we just drop the protocol://host:port
            # part in front of the request-url (because we know
            # we would not forward the request to *another*
            # proxy).

            # So we do what SimpleHTTPRequestHandler.translate_path
            # do beginning with python 2.4.3: abandon query
            # parameters, scheme, host port, etc (which ensure we
            # provide the right behaviour on all python versions).
            path = urlparse.urlparse(path)[2]
            # And now, we can apply *our* trick to proxy files
            path += '-proxied'

        return self._translate_path(path)

    def _translate_path(self, path):
        """Translate a /-separated PATH to the local filename syntax.

        Note that we're translating http URLs here, not file URLs.
        The URL root location is the server's startup directory.
        Components that mean special things to the local file system
        (e.g. drive or directory names) are ignored.  (XXX They should
        probably be diagnosed.)

        Override from python standard library to stop it calling os.getcwd()
        """
        # abandon query parameters
        path = urlparse.urlparse(path)[2]
        path = posixpath.normpath(urllib.unquote(path))
        path = path.decode('utf-8')
        words = path.split('/')
        words = filter(None, words)
        path = self._cwd
        for num, word in enumerate(words):
            if num == 0:
                drive, word = os.path.splitdrive(word)
            head, word = os.path.split(word)
            if word in (os.curdir, os.pardir): continue
            path = os.path.join(path, word)
        return path


class TestingHTTPServerMixin:

    def __init__(self, test_case_server):
        # test_case_server can be used to communicate between the
        # tests and the server (or the request handler and the
        # server), allowing dynamic behaviors to be defined from
        # the tests cases.
        self.test_case_server = test_case_server
        self._home_dir = test_case_server._home_dir


class TestingHTTPServer(test_server.TestingTCPServer, TestingHTTPServerMixin):

    def __init__(self, server_address, request_handler_class,
                 test_case_server):
        test_server.TestingTCPServer.__init__(self, server_address,
                                              request_handler_class)
        TestingHTTPServerMixin.__init__(self, test_case_server)


class TestingThreadingHTTPServer(test_server.TestingThreadingTCPServer,
                                 TestingHTTPServerMixin):
    """A threading HTTP test server for HTTP 1.1.

    Since tests can initiate several concurrent connections to the same http
    server, we need an independent connection for each of them. We achieve that
    by spawning a new thread for each connection.
    """
    def __init__(self, server_address, request_handler_class,
                 test_case_server):
        test_server.TestingThreadingTCPServer.__init__(self, server_address,
                                                       request_handler_class)
        TestingHTTPServerMixin.__init__(self, test_case_server)


class HttpServer(test_server.TestingTCPServerInAThread):
    """A test server for http transports.

    Subclasses can provide a specific request handler.
    """

    # The real servers depending on the protocol
    http_server_class = {'HTTP/1.0': TestingHTTPServer,
                         'HTTP/1.1': TestingThreadingHTTPServer,
                         }

    # Whether or not we proxy the requests (see
    # TestingHTTPRequestHandler.translate_path).
    proxy_requests = False

    # used to form the url that connects to this server
    _url_protocol = 'http'

    def __init__(self, request_handler=TestingHTTPRequestHandler,
                 protocol_version=None):
        """Constructor.

        :param request_handler: a class that will be instantiated to handle an
            http connection (one or several requests).

        :param protocol_version: if specified, will override the protocol
            version of the request handler.
        """
        # Depending on the protocol version, we will create the approriate
        # server
        if protocol_version is None:
            # Use the request handler one
            proto_vers = request_handler.protocol_version
        else:
            # Use our own, it will be used to override the request handler
            # one too.
            proto_vers = protocol_version
        # Get the appropriate server class for the required protocol
        serv_cls = self.http_server_class.get(proto_vers, None)
        if serv_cls is None:
            raise httplib.UnknownProtocol(proto_vers)
        self.host = 'localhost'
        self.port = 0
        super(HttpServer, self).__init__((self.host, self.port),
                                         serv_cls,
                                         request_handler)
        self.protocol_version = proto_vers
        # Allows tests to verify number of GET requests issued
        self.GET_request_nb = 0
        self._http_base_url = None
        self.logs = []

    def create_server(self):
        return self.server_class(
            (self.host, self.port), self.request_handler_class, self)

    def _get_remote_url(self, path):
        path_parts = path.split(os.path.sep)
        if os.path.isabs(path):
            if path_parts[:len(self._local_path_parts)] != \
                   self._local_path_parts:
                raise BadWebserverPath(path, self.test_dir)
            remote_path = '/'.join(path_parts[len(self._local_path_parts):])
        else:
            remote_path = '/'.join(path_parts)

        return self._http_base_url + remote_path

    def log(self, format, *args):
        """Capture Server log output."""
        self.logs.append(format % args)

    def start_server(self, backing_transport_server=None):
        """See bzrlib.transport.Server.start_server.

        :param backing_transport_server: The transport that requests over this
            protocol should be forwarded to. Note that this is currently not
            supported for HTTP.
        """
        # XXX: TODO: make the server back onto vfs_server rather than local
        # disk.
        if not (backing_transport_server is None
                or isinstance(backing_transport_server,
                              test_server.LocalURLServer)):
            raise AssertionError(
                "HTTPServer currently assumes local transport, got %s" %
                backing_transport_server)
        self._home_dir = os.getcwdu()
        self._local_path_parts = self._home_dir.split(os.path.sep)
        self.logs = []

        super(HttpServer, self).start_server()
        self._http_base_url = '%s://%s:%s/' % (
            self._url_protocol, self.host, self.port)

    def get_url(self):
        """See bzrlib.transport.Server.get_url."""
        return self._get_remote_url(self._home_dir)

    def get_bogus_url(self):
        """See bzrlib.transport.Server.get_bogus_url."""
        # this is chosen to try to prevent trouble with proxies, weird dns,
        # etc
        return self._url_protocol + '://127.0.0.1:1/'


class HttpServer_urllib(HttpServer):
    """Subclass of HttpServer that gives http+urllib urls.

    This is for use in testing: connections to this server will always go
    through urllib where possible.
    """

    # urls returned by this server should require the urllib client impl
    _url_protocol = 'http+urllib'


class HttpServer_PyCurl(HttpServer):
    """Subclass of HttpServer that gives http+pycurl urls.

    This is for use in testing: connections to this server will always go
    through pycurl where possible.
    """

    # We don't care about checking the pycurl availability as
    # this server will be required only when pycurl is present

    # urls returned by this server should require the pycurl client impl
    _url_protocol = 'http+pycurl'