Lib/BaseHTTPServer.py

   1 """HTTP server base class.
   2
   3 Note: the class in this module doesn't implement any HTTP request; see
   4 SimpleHTTPServer for simple implementations of GET, HEAD and POST
   5 (including CGI scripts).  It does, however, optionally implement HTTP/1.1
   6 persistent connections, as of version 0.3.
   7
   8 Contents:
   9
  10 - BaseHTTPRequestHandler: HTTP request handler base class
  11 - test: test function
  12
  13 XXX To do:
  14
  15 - log requests even later (to capture byte count)
  16 - log user-agent header and other interesting goodies
  17 - send error log to separate file
  18 """
  19
  20
  21 # See also:
  22 #
  23 # HTTP Working Group                                        T. Berners-Lee
  24 # INTERNET-DRAFT                                            R. T. Fielding
  25 # <draft-ietf-http-v10-spec-00.txt>                     H. Frystyk Nielsen
  26 # Expires September 8, 1995                                  March 8, 1995
  27 #
  28 # URL: http://www.ics.uci.edu/pub/ietf/http/draft-ietf-http-v10-spec-00.txt
  29 #
  30 # and
  31 #
  32 # Network Working Group                                      R. Fielding
  33 # Request for Comments: 2616                                       et al
  34 # Obsoletes: 2068                                              June 1999
  35 # Category: Standards Track
  36 #
  37 # URL: http://www.faqs.org/rfcs/rfc2616.html
  38
  39 # Log files
  40 # ---------
  41 #
  42 # Here's a quote from the NCSA httpd docs about log file format.
  43 #
  44 # | The logfile format is as follows. Each line consists of:
  45 # |
  46 # | host rfc931 authuser [DD/Mon/YYYY:hh:mm:ss] "request" ddd bbbb
  47 # |
  48 # |        host: Either the DNS name or the IP number of the remote client
  49 # |        rfc931: Any information returned by identd for this person,
  50 # |                - otherwise.
  51 # |        authuser: If user sent a userid for authentication, the user name,
  52 # |                  - otherwise.
  53 # |        DD: Day
  54 # |        Mon: Month (calendar name)
  55 # |        YYYY: Year
  56 # |        hh: hour (24-hour format, the machine's timezone)
  57 # |        mm: minutes
  58 # |        ss: seconds
  59 # |        request: The first line of the HTTP request as sent by the client.
  60 # |        ddd: the status code returned by the server, - if not available.
  61 # |        bbbb: the total number of bytes sent,
  62 # |              *not including the HTTP/1.0 header*, - if not available
  63 # |
  64 # | You can determine the name of the file accessed through request.
  65 #
  66 # (Actually, the latter is only true if you know the server configuration
  67 # at the time the request was made!)
  68
  69 __version__ = "0.3"
  70
  71 __all__ = ["HTTPServer", "BaseHTTPRequestHandler"]
  72
  73 import sys
  74 import time
  75 import socket # For gethostbyaddr()
  76 import mimetools
  77 import SocketServer
  78
  79 # Default error message
  80 DEFAULT_ERROR_MESSAGE = """\
  81 <head>
  82 <title>Error response</title>
  83 </head>
  84 <body>
  85 <h1>Error response</h1>
  86 <p>Error code %(code)d.
  87 <p>Message: %(message)s.
  88 <p>Error code explanation: %(code)s = %(explain)s.
  89 </body>
  90 """
  91
  92 def _quote_html(html):
  93     return html.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
  94
  95 class HTTPServer(SocketServer.TCPServer):
  96
  97     allow_reuse_address = 1    # Seems to make sense in testing environment
  98
  99     def server_bind(self):
 100         """Override server_bind to store the server name."""
 101         SocketServer.TCPServer.server_bind(self)
 102         host, port = self.socket.getsockname()[:2]
 103         self.server_name = socket.getfqdn(host)
 104         self.server_port = port
 105
 106
 107 class BaseHTTPRequestHandler(SocketServer.StreamRequestHandler):
 108
 109     """HTTP request handler base class.
 110
 111     The following explanation of HTTP serves to guide you through the
 112     code as well as to expose any misunderstandings I may have about
 113     HTTP (so you don't need to read the code to figure out I'm wrong
 114     :-).
 115
 116     HTTP (HyperText Transfer Protocol) is an extensible protocol on
 117     top of a reliable stream transport (e.g. TCP/IP).  The protocol
 118     recognizes three parts to a request:
 119
 120     1. One line identifying the request type and path
 121     2. An optional set of RFC-822-style headers
 122     3. An optional data part
 123
 124     The headers and data are separated by a blank line.
 125
 126     The first line of the request has the form
 127
 128     <command> <path> <version>
 129
 130     where <command> is a (case-sensitive) keyword such as GET or POST,
 131     <path> is a string containing path information for the request,
 132     and <version> should be the string "HTTP/1.0" or "HTTP/1.1".
 133     <path> is encoded using the URL encoding scheme (using %xx to signify
 134     the ASCII character with hex code xx).
 135
 136     The specification specifies that lines are separated by CRLF but
 137     for compatibility with the widest range of clients recommends
 138     servers also handle LF.  Similarly, whitespace in the request line
 139     is treated sensibly (allowing multiple spaces between components
 140     and allowing trailing whitespace).
 141
 142     Similarly, for output, lines ought to be separated by CRLF pairs
 143     but most clients grok LF characters just fine.
 144
 145     If the first line of the request has the form
 146
 147     <command> <path>
 148
 149     (i.e. <version> is left out) then this is assumed to be an HTTP
 150     0.9 request; this form has no optional headers and data part and
 151     the reply consists of just the data.
 152
 153     The reply form of the HTTP 1.x protocol again has three parts:
 154
 155     1. One line giving the response code
 156     2. An optional set of RFC-822-style headers
 157     3. The data
 158
 159     Again, the headers and data are separated by a blank line.
 160
 161     The response code line has the form
 162
 163     <version> <responsecode> <responsestring>
 164
 165     where <version> is the protocol version ("HTTP/1.0" or "HTTP/1.1"),
 166     <responsecode> is a 3-digit response code indicating success or
 167     failure of the request, and <responsestring> is an optional
 168     human-readable string explaining what the response code means.
 169
 170     This server parses the request and the headers, and then calls a
 171     function specific to the request type (<command>).  Specifically,
 172     a request SPAM will be handled by a method do_SPAM().  If no
 173     such method exists the server sends an error response to the
 174     client.  If it exists, it is called with no arguments:
 175
 176     do_SPAM()
 177
 178     Note that the request name is case sensitive (i.e. SPAM and spam
 179     are different requests).
 180
 181     The various request details are stored in instance variables:
 182
 183     - client_address is the client IP address in the form (host,
 184     port);
 185
 186     - command, path and version are the broken-down request line;
 187
 188     - headers is an instance of mimetools.Message (or a derived
 189     class) containing the header information;
 190
 191     - rfile is a file object open for reading positioned at the
 192     start of the optional input data part;
 193
 194     - wfile is a file object open for writing.
 195
 196     IT IS IMPORTANT TO ADHERE TO THE PROTOCOL FOR WRITING!
 197
 198     The first thing to be written must be the response line.  Then
 199     follow 0 or more header lines, then a blank line, and then the
 200     actual data (if any).  The meaning of the header lines depends on
 201     the command executed by the server; in most cases, when data is
 202     returned, there should be at least one header line of the form
 203
 204     Content-type: <type>/<subtype>
 205
 206     where <type> and <subtype> should be registered MIME types,
 207     e.g. "text/html" or "text/plain".
 208
 209     """
 210
 211     # The Python system version, truncated to its first component.
 212     sys_version = "Python/" + sys.version.split()[0]
 213
 214     # The server software version.  You may want to override this.
 215     # The format is multiple whitespace-separated strings,
 216     # where each string is of the form name[/version].
 217     server_version = "BaseHTTP/" + __version__
 218
 219     def parse_request(self):
 220         """Parse a request (internal).
 221
 222         The request should be stored in self.raw_requestline; the results
 223         are in self.command, self.path, self.request_version and
 224         self.headers.
 225
 226         Return True for success, False for failure; on failure, an
 227         error is sent back.
 228
 229         """
 230         self.command = None  # set in case of error on the first line
 231         self.request_version = version = "HTTP/0.9" # Default
 232         self.close_connection = 1
 233         requestline = self.raw_requestline
 234         if requestline[-2:] == '\r\n':
 235             requestline = requestline[:-2]
 236         elif requestline[-1:] == '\n':
 237             requestline = requestline[:-1]
 238         self.requestline = requestline
 239         words = requestline.split()
 240         if len(words) == 3:
 241             [command, path, version] = words
 242             if version[:5] != 'HTTP/':
 243                 self.send_error(400, "Bad request version (%r)" % version)
 244                 return False
 245             try:
 246                 base_version_number = version.split('/', 1)[1]
 247                 version_number = base_version_number.split(".")
 248                 # RFC 2145 section 3.1 says there can be only one "." and
 249                 #   - major and minor numbers MUST be treated as
 250                 #      separate integers;
 251                 #   - HTTP/2.4 is a lower version than HTTP/2.13, which in
 252                 #      turn is lower than HTTP/12.3;
 253                 #   - Leading zeros MUST be ignored by recipients.
 254                 if len(version_number) != 2:
 255                     raise ValueError
 256                 version_number = int(version_number[0]), int(version_number[1])
 257             except (ValueError, IndexError):
 258                 self.send_error(400, "Bad request version (%r)" % version)
 259                 return False
 260             if version_number >= (1, 1) and self.protocol_version >= "HTTP/1.1":
 261                 self.close_connection = 0
 262             if version_number >= (2, 0):
 263                 self.send_error(505,
 264                           "Invalid HTTP Version (%s)" % base_version_number)
 265                 return False
 266         elif len(words) == 2:
 267             [command, path] = words
 268             self.close_connection = 1
 269             if command != 'GET':
 270                 self.send_error(400,
 271                                 "Bad HTTP/0.9 request type (%r)" % command)
 272                 return False
 273         elif not words:
 274             return False
 275         else:
 276             self.send_error(400, "Bad request syntax (%r)" % requestline)
 277             return False
 278         self.command, self.path, self.request_version = command, path, version
 279
 280         # Examine the headers and look for a Connection directive
 281         self.headers = self.MessageClass(self.rfile, 0)
 282
 283         conntype = self.headers.get('Connection', "")
 284         if conntype.lower() == 'close':
 285             self.close_connection = 1
 286         elif (conntype.lower() == 'keep-alive' and
 287               self.protocol_version >= "HTTP/1.1"):
 288             self.close_connection = 0
 289         return True
 290
 291     def handle_one_request(self):
 292         """Handle a single HTTP request.
 293
 294         You normally don't need to override this method; see the class
 295         __doc__ string for information on how to handle specific HTTP
 296         commands such as GET and POST.
 297
 298         """
 299         self.raw_requestline = self.rfile.readline()
 300         if not self.raw_requestline:
 301             self.close_connection = 1
 302             return
 303         if not self.parse_request(): # An error code has been sent, just exit
 304             return
 305         mname = 'do_' + self.command
 306         if not hasattr(self, mname):
 307             self.send_error(501, "Unsupported method (%r)" % self.command)
 308             return
 309         method = getattr(self, mname)
 310         method()
 311
 312     def handle(self):
 313         """Handle multiple requests if necessary."""
 314         self.close_connection = 1
 315
 316         self.handle_one_request()
 317         while not self.close_connection:
 318             self.handle_one_request()
 319
 320     def send_error(self, code, message=None):
 321         """Send and log an error reply.
 322
 323         Arguments are the error code, and a detailed message.
 324         The detailed message defaults to the short entry matching the
 325         response code.
 326
 327         This sends an error response (so it must be called before any
 328         output has been generated), logs the error, and finally sends
 329         a piece of HTML explaining the error to the user.
 330
 331         """
 332
 333         try:
 334             short, long = self.responses[code]
 335         except KeyError:
 336             short, long = '???', '???'
 337         if message is None:
 338             message = short
 339         explain = long
 340         self.log_error("code %d, message %s", code, message)
 341         # using _quote_html to prevent Cross Site Scripting attacks (see bug #1100201)
 342         content = (self.error_message_format %
 343                    {'code': code, 'message': _quote_html(message), 'explain': explain})
 344         self.send_response(code, message)
 345         self.send_header("Content-Type", "text/html")
 346         self.send_header('Connection', 'close')
 347         self.end_headers()
 348         if self.command != 'HEAD' and code >= 200 and code not in (204, 304):
 349             self.wfile.write(content)
 350
 351     error_message_format = DEFAULT_ERROR_MESSAGE
 352
 353     def send_response(self, code, message=None):
 354         """Send the response header and log the response code.
 355
 356         Also send two standard headers with the server software
 357         version and the current date.
 358
 359         """
 360         self.log_request(code)
 361         if message is None:
 362             if code in self.responses:
 363                 message = self.responses[code][0]
 364             else:
 365                 message = ''
 366         if self.request_version != 'HTTP/0.9':
 367             self.wfile.write("%s %d %s\r\n" %
 368                              (self.protocol_version, code, message))
 369             # print (self.protocol_version, code, message)
 370         self.send_header('Server', self.version_string())
 371         self.send_header('Date', self.date_time_string())
 372
 373     def send_header(self, keyword, value):
 374         """Send a MIME header."""
 375         if self.request_version != 'HTTP/0.9':
 376             self.wfile.write("%s: %s\r\n" % (keyword, value))
 377
 378         if keyword.lower() == 'connection':
 379             if value.lower() == 'close':
 380                 self.close_connection = 1
 381             elif value.lower() == 'keep-alive':
 382                 self.close_connection = 0
 383
 384     def end_headers(self):
 385         """Send the blank line ending the MIME headers."""
 386         if self.request_version != 'HTTP/0.9':
 387             self.wfile.write("\r\n")
 388
 389     def log_request(self, code='-', size='-'):
 390         """Log an accepted request.
 391
 392         This is called by send_response().
 393
 394         """
 395
 396         self.log_message('"%s" %s %s',
 397                          self.requestline, str(code), str(size))
 398
 399     def log_error(self, format, *args):
 400         """Log an error.
 401
 402         This is called when a request cannot be fulfilled.  By
 403         default it passes the message on to log_message().
 404
 405         Arguments are the same as for log_message().
 406
 407         XXX This should go to the separate error log.
 408
 409         """
 410
 411         self.log_message(format, *args)
 412
 413     def log_message(self, format, *args):
 414         """Log an arbitrary message.
 415
 416         This is used by all other logging functions.  Override
 417         it if you have specific logging wishes.
 418
 419         The first argument, FORMAT, is a format string for the
 420         message to be logged.  If the format string contains
 421         any % escapes requiring parameters, they should be
 422         specified as subsequent arguments (it's just like
 423         printf!).
 424
 425         The client host and current date/time are prefixed to
 426         every message.
 427
 428         """
 429
 430         sys.stderr.write("%s - - [%s] %s\n" %
 431                          (self.address_string(),
 432                           self.log_date_time_string(),
 433                           format%args))
 434
 435     def version_string(self):
 436         """Return the server software version string."""
 437         return self.server_version + ' ' + self.sys_version
 438
 439     def date_time_string(self, timestamp=None):
 440         """Return the current date and time formatted for a message header."""
 441         if timestamp is None:
 442             timestamp = time.time()
 443         year, month, day, hh, mm, ss, wd, y, z = time.gmtime(timestamp)
 444         s = "%s, %02d %3s %4d %02d:%02d:%02d GMT" % (
 445                 self.weekdayname[wd],
 446                 day, self.monthname[month], year,
 447                 hh, mm, ss)
 448         return s
 449
 450     def log_date_time_string(self):
 451         """Return the current time formatted for logging."""
 452         now = time.time()
 453         year, month, day, hh, mm, ss, x, y, z = time.localtime(now)
 454         s = "%02d/%3s/%04d %02d:%02d:%02d" % (
 455                 day, self.monthname[month], year, hh, mm, ss)
 456         return s
 457
 458     weekdayname = ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']
 459
 460     monthname = [None,
 461                  'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
 462                  'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
 463
 464     def address_string(self):
 465         """Return the client address formatted for logging.
 466
 467         This version looks up the full hostname using gethostbyaddr(),
 468         and tries to find a name that contains at least one dot.
 469
 470         """
 471
 472         host, port = self.client_address[:2]
 473         return socket.getfqdn(host)
 474
 475     # Essentially static class variables
 476
 477     # The version of the HTTP protocol we support.
 478     # Set this to HTTP/1.1 to enable automatic keepalive
 479     protocol_version = "HTTP/1.0"
 480
 481     # The Message-like class used to parse headers
 482     MessageClass = mimetools.Message
 483
 484     # Table mapping response codes to messages; entries have the
 485     # form {code: (shortmessage, longmessage)}.
 486     # See RFC 2616.
 487     responses = {
 488         100: ('Continue', 'Request received, please continue'),
 489         101: ('Switching Protocols',
 490               'Switching to new protocol; obey Upgrade header'),
 491
 492         200: ('OK', 'Request fulfilled, document follows'),
 493         201: ('Created', 'Document created, URL follows'),
 494         202: ('Accepted',
 495               'Request accepted, processing continues off-line'),
 496         203: ('Non-Authoritative Information', 'Request fulfilled from cache'),
 497         204: ('No Content', 'Request fulfilled, nothing follows'),
 498         205: ('Reset Content', 'Clear input form for further input.'),
 499         206: ('Partial Content', 'Partial content follows.'),
 500
 501         300: ('Multiple Choices',
 502               'Object has several resources -- see URI list'),
 503         301: ('Moved Permanently', 'Object moved permanently -- see URI list'),
 504         302: ('Found', 'Object moved temporarily -- see URI list'),
 505         303: ('See Other', 'Object moved -- see Method and URL list'),
 506         304: ('Not Modified',
 507               'Document has not changed since given time'),
 508         305: ('Use Proxy',
 509               'You must use proxy specified in Location to access this '
 510               'resource.'),
 511         307: ('Temporary Redirect',
 512               'Object moved temporarily -- see URI list'),
 513
 514         400: ('Bad Request',
 515               'Bad request syntax or unsupported method'),
 516         401: ('Unauthorized',
 517               'No permission -- see authorization schemes'),
 518         402: ('Payment Required',
 519               'No payment -- see charging schemes'),
 520         403: ('Forbidden',
 521               'Request forbidden -- authorization will not help'),
 522         404: ('Not Found', 'Nothing matches the given URI'),
 523         405: ('Method Not Allowed',
 524               'Specified method is invalid for this server.'),
 525         406: ('Not Acceptable', 'URI not available in preferred format.'),
 526         407: ('Proxy Authentication Required', 'You must authenticate with '
 527               'this proxy before proceeding.'),
 528         408: ('Request Timeout', 'Request timed out; try again later.'),
 529         409: ('Conflict', 'Request conflict.'),
 530         410: ('Gone',
 531               'URI no longer exists and has been permanently removed.'),
 532         411: ('Length Required', 'Client must specify Content-Length.'),
 533         412: ('Precondition Failed', 'Precondition in headers is false.'),
 534         413: ('Request Entity Too Large', 'Entity is too large.'),
 535         414: ('Request-URI Too Long', 'URI is too long.'),
 536         415: ('Unsupported Media Type', 'Entity body in unsupported format.'),
 537         416: ('Requested Range Not Satisfiable',
 538               'Cannot satisfy request range.'),
 539         417: ('Expectation Failed',
 540               'Expect condition could not be satisfied.'),
 541
 542         500: ('Internal Server Error', 'Server got itself in trouble'),
 543         501: ('Not Implemented',
 544               'Server does not support this operation'),
 545         502: ('Bad Gateway', 'Invalid responses from another server/proxy.'),
 546         503: ('Service Unavailable',
 547               'The server cannot process the request due to a high load'),
 548         504: ('Gateway Timeout',
 549               'The gateway server did not receive a timely response'),
 550         505: ('HTTP Version Not Supported', 'Cannot fulfill request.'),
 551         }
 552
 553
 554 def test(HandlerClass = BaseHTTPRequestHandler,
 555          ServerClass = HTTPServer, protocol="HTTP/1.0"):
 556     """Test the HTTP request handler class.
 557
 558     This runs an HTTP server on port 8000 (or the first command line
 559     argument).
 560
 561     """
 562
 563     if sys.argv[1:]:
 564         port = int(sys.argv[1])
 565     else:
 566         port = 8000
 567     server_address = ('', port)
 568
 569     HandlerClass.protocol_version = protocol
 570     httpd = ServerClass(server_address, HandlerClass)
 571
 572     sa = httpd.socket.getsockname()
 573     print "Serving HTTP on", sa[0], "port", sa[1], "..."
 574     httpd.serve_forever()
 575
 576
 577 if __name__ == '__main__':
 578     test()