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).
   6
   7 Contents:
   8
   9 - BaseHTTPRequestHandler: HTTP request handler base class
  10 - test: test function
  11
  12 XXX To do:
  13
  14 - send server version
  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 - are request names really case sensitive?
  19
  20 """
  21
  22
  23 # See also:
  24 #
  25 # HTTP Working Group                                        T. Berners-Lee
  26 # INTERNET-DRAFT                                            R. T. Fielding
  27 # <draft-ietf-http-v10-spec-00.txt>                     H. Frystyk Nielsen
  28 # Expires September 8, 1995                                  March 8, 1995
  29 #
  30 # URL: http://www.ics.uci.edu/pub/ietf/http/draft-ietf-http-v10-spec-00.txt
  31
  32
  33 # Log files
  34 # ---------
  35 #
  36 # Here's a quote from the NCSA httpd docs about log file format.
  37 #
  38 # | The logfile format is as follows. Each line consists of:
  39 # |
  40 # | host rfc931 authuser [DD/Mon/YYYY:hh:mm:ss] "request" ddd bbbb
  41 # |
  42 # |        host: Either the DNS name or the IP number of the remote client
  43 # |        rfc931: Any information returned by identd for this person,
  44 # |                - otherwise.
  45 # |        authuser: If user sent a userid for authentication, the user name,
  46 # |                  - otherwise.
  47 # |        DD: Day
  48 # |        Mon: Month (calendar name)
  49 # |        YYYY: Year
  50 # |        hh: hour (24-hour format, the machine's timezone)
  51 # |        mm: minutes
  52 # |        ss: seconds
  53 # |        request: The first line of the HTTP request as sent by the client.
  54 # |        ddd: the status code returned by the server, - if not available.
  55 # |        bbbb: the total number of bytes sent,
  56 # |              *not including the HTTP/1.0 header*, - if not available
  57 # |
  58 # | You can determine the name of the file accessed through request.
  59 #
  60 # (Actually, the latter is only true if you know the server configuration
  61 # at the time the request was made!)
  62
  63
  64 __version__ = "0.2"
  65
  66
  67 import sys
  68 import time
  69 import socket # For gethostbyaddr()
  70 import string
  71 import mimetools
  72 import SocketServer
  73
  74 # Default error message
  75 DEFAULT_ERROR_MESSAGE = """\
  76 <head>
  77 <title>Error response</title>
  78 </head>
  79 <body>
  80 <h1>Error response</h1>
  81 <p>Error code %(code)d.
  82 <p>Message: %(message)s.
  83 <p>Error code explanation: %(code)s = %(explain)s.
  84 </body>
  85 """
  86
  87
  88 class HTTPServer(SocketServer.TCPServer):
  89
  90     allow_reuse_address = 1    # Seems to make sense in testing environment
  91
  92     def server_bind(self):
  93         """Override server_bind to store the server name."""
  94         SocketServer.TCPServer.server_bind(self)
  95         host, port = self.socket.getsockname()
  96         if not host or host == '0.0.0.0':
  97             host = socket.gethostname()
  98         try:
  99             hostname, hostnames, hostaddrs = socket.gethostbyaddr(host)
 100         except socket.error:
 101             hostname = host
 102         else:
 103             if '.' not in hostname:
 104                 for host in hostnames:
 105                     if '.' in host:
 106                         hostname = host
 107                         break
 108         self.server_name = hostname
 109         self.server_port = port
 110
 111
 112 class BaseHTTPRequestHandler(SocketServer.StreamRequestHandler):
 113
 114     """HTTP request handler base class.
 115
 116     The following explanation of HTTP serves to guide you through the
 117     code as well as to expose any misunderstandings I may have about
 118     HTTP (so you don't need to read the code to figure out I'm wrong
 119     :-).
 120
 121     HTTP (HyperText Transfer Protocol) is an extensible protocol on
 122     top of a reliable stream transport (e.g. TCP/IP).  The protocol
 123     recognizes three parts to a request:
 124
 125     1. One line identifying the request type and path
 126     2. An optional set of RFC-822-style headers
 127     3. An optional data part
 128
 129     The headers and data are separated by a blank line.
 130
 131     The first line of the request has the form
 132
 133     <command> <path> <version>
 134
 135     where <command> is a (case-sensitive) keyword such as GET or POST,
 136     <path> is a string containing path information for the request,
 137     and <version> should be the string "HTTP/1.0".  <path> is encoded
 138     using the URL encoding scheme (using %xx to signify the ASCII
 139     character with hex code xx).
 140
 141     The protocol is vague about whether lines are separated by LF
 142     characters or by CRLF pairs -- for compatibility with the widest
 143     range of clients, both should be accepted.  Similarly, whitespace
 144     in the request line should be treated sensibly (allowing multiple
 145     spaces between components and allowing trailing whitespace).
 146
 147     Similarly, for output, lines ought to be separated by CRLF pairs
 148     but most clients grok LF characters just fine.
 149
 150     If the first line of the request has the form
 151
 152     <command> <path>
 153
 154     (i.e. <version> is left out) then this is assumed to be an HTTP
 155     0.9 request; this form has no optional headers and data part and
 156     the reply consists of just the data.
 157
 158     The reply form of the HTTP 1.0 protocol again has three parts:
 159
 160     1. One line giving the response code
 161     2. An optional set of RFC-822-style headers
 162     3. The data
 163
 164     Again, the headers and data are separated by a blank line.
 165
 166     The response code line has the form
 167
 168     <version> <responsecode> <responsestring>
 169
 170     where <version> is the protocol version (always "HTTP/1.0"),
 171     <responsecode> is a 3-digit response code indicating success or
 172     failure of the request, and <responsestring> is an optional
 173     human-readable string explaining what the response code means.
 174
 175     This server parses the request and the headers, and then calls a
 176     function specific to the request type (<command>).  Specifically,
 177     a request SPAM will be handled by a method do_SPAM().  If no
 178     such method exists the server sends an error response to the
 179     client.  If it exists, it is called with no arguments:
 180
 181     do_SPAM()
 182
 183     Note that the request name is case sensitive (i.e. SPAM and spam
 184     are different requests).
 185
 186     The various request details are stored in instance variables:
 187
 188     - client_address is the client IP address in the form (host,
 189     port);
 190
 191     - command, path and version are the broken-down request line;
 192
 193     - headers is an instance of mimetools.Message (or a derived
 194     class) containing the header information;
 195
 196     - rfile is a file object open for reading positioned at the
 197     start of the optional input data part;
 198
 199     - wfile is a file object open for writing.
 200
 201     IT IS IMPORTANT TO ADHERE TO THE PROTOCOL FOR WRITING!
 202
 203     The first thing to be written must be the response line.  Then
 204     follow 0 or more header lines, then a blank line, and then the
 205     actual data (if any).  The meaning of the header lines depends on
 206     the command executed by the server; in most cases, when data is
 207     returned, there should be at least one header line of the form
 208
 209     Content-type: <type>/<subtype>
 210
 211     where <type> and <subtype> should be registered MIME types,
 212     e.g. "text/html" or "text/plain".
 213
 214     """
 215
 216     # The Python system version, truncated to its first component.
 217     sys_version = "Python/" + string.split(sys.version)[0]
 218
 219     # The server software version.  You may want to override this.
 220     # The format is multiple whitespace-separated strings,
 221     # where each string is of the form name[/version].
 222     server_version = "BaseHTTP/" + __version__
 223
 224     def parse_request(self):
 225         """Parse a request (internal).
 226
 227         The request should be stored in self.raw_request; the results
 228         are in self.command, self.path, self.request_version and
 229         self.headers.
 230
 231         Return value is 1 for success, 0 for failure; on failure, an
 232         error is sent back.
 233
 234         """
 235         self.request_version = version = "HTTP/0.9" # Default
 236         requestline = self.raw_requestline
 237         if requestline[-2:] == '\r\n':
 238             requestline = requestline[:-2]
 239         elif requestline[-1:] == '\n':
 240             requestline = requestline[:-1]
 241         self.requestline = requestline
 242         words = string.split(requestline)
 243         if len(words) == 3:
 244             [command, path, version] = words
 245             if version[:5] != 'HTTP/':
 246                 self.send_error(400, "Bad request version (%s)" % `version`)
 247                 return 0
 248         elif len(words) == 2:
 249             [command, path] = words
 250             if command != 'GET':
 251                 self.send_error(400,
 252                                 "Bad HTTP/0.9 request type (%s)" % `command`)
 253                 return 0
 254         else:
 255             self.send_error(400, "Bad request syntax (%s)" % `requestline`)
 256             return 0
 257         self.command, self.path, self.request_version = command, path, version
 258         self.headers = self.MessageClass(self.rfile, 0)
 259         return 1
 260
 261     def handle(self):
 262         """Handle a single HTTP request.
 263
 264         You normally don't need to override this method; see the class
 265         __doc__ string for information on how to handle specific HTTP
 266         commands such as GET and POST.
 267
 268         """
 269
 270         self.raw_requestline = self.rfile.readline()
 271         if not self.parse_request(): # An error code has been sent, just exit
 272             return
 273         mname = 'do_' + self.command
 274         if not hasattr(self, mname):
 275             self.send_error(501, "Unsupported method (%s)" % `self.command`)
 276             return
 277         method = getattr(self, mname)
 278         method()
 279
 280     def send_error(self, code, message=None):
 281         """Send and log an error reply.
 282
 283         Arguments are the error code, and a detailed message.
 284         The detailed message defaults to the short entry matching the
 285         response code.
 286
 287         This sends an error response (so it must be called before any
 288         output has been generated), logs the error, and finally sends
 289         a piece of HTML explaining the error to the user.
 290
 291         """
 292
 293         try:
 294             short, long = self.responses[code]
 295         except KeyError:
 296             short, long = '???', '???'
 297         if not message:
 298             message = short
 299         explain = long
 300         self.log_error("code %d, message %s", code, message)
 301         self.send_response(code, message)
 302         self.end_headers()
 303         self.wfile.write(self.error_message_format %
 304                          {'code': code,
 305                           'message': message,
 306                           'explain': explain})
 307
 308     error_message_format = DEFAULT_ERROR_MESSAGE
 309
 310     def send_response(self, code, message=None):
 311         """Send the response header and log the response code.
 312
 313         Also send two standard headers with the server software
 314         version and the current date.
 315
 316         """
 317         self.log_request(code)
 318         if message is None:
 319             if self.responses.has_key(code):
 320                 message = self.responses[code][0]
 321             else:
 322                 message = ''
 323         if self.request_version != 'HTTP/0.9':
 324             self.wfile.write("%s %s %s\r\n" %
 325                              (self.protocol_version, str(code), message))
 326         self.send_header('Server', self.version_string())
 327         self.send_header('Date', self.date_time_string())
 328
 329     def send_header(self, keyword, value):
 330         """Send a MIME header."""
 331         if self.request_version != 'HTTP/0.9':
 332             self.wfile.write("%s: %s\r\n" % (keyword, value))
 333
 334     def end_headers(self):
 335         """Send the blank line ending the MIME headers."""
 336         if self.request_version != 'HTTP/0.9':
 337             self.wfile.write("\r\n")
 338
 339     def log_request(self, code='-', size='-'):
 340         """Log an accepted request.
 341
 342         This is called by send_reponse().
 343
 344         """
 345
 346         self.log_message('"%s" %s %s',
 347                          self.requestline, str(code), str(size))
 348
 349     def log_error(self, *args):
 350         """Log an error.
 351
 352         This is called when a request cannot be fulfilled.  By
 353         default it passes the message on to log_message().
 354
 355         Arguments are the same as for log_message().
 356
 357         XXX This should go to the separate error log.
 358
 359         """
 360
 361         apply(self.log_message, args)
 362
 363     def log_message(self, format, *args):
 364         """Log an arbitrary message.
 365
 366         This is used by all other logging functions.  Override
 367         it if you have specific logging wishes.
 368
 369         The first argument, FORMAT, is a format string for the
 370         message to be logged.  If the format string contains
 371         any % escapes requiring parameters, they should be
 372         specified as subsequent arguments (it's just like
 373         printf!).
 374
 375         The client host and current date/time are prefixed to
 376         every message.
 377
 378         """
 379
 380         sys.stderr.write("%s - - [%s] %s\n" %
 381                          (self.address_string(),
 382                           self.log_date_time_string(),
 383                           format%args))
 384
 385     def version_string(self):
 386         """Return the server software version string."""
 387         return self.server_version + ' ' + self.sys_version
 388
 389     def date_time_string(self):
 390         """Return the current date and time formatted for a message header."""
 391         now = time.time()
 392         year, month, day, hh, mm, ss, wd, y, z = time.gmtime(now)
 393         s = "%s, %02d %3s %4d %02d:%02d:%02d GMT" % (
 394                 self.weekdayname[wd],
 395                 day, self.monthname[month], year,
 396                 hh, mm, ss)
 397         return s
 398
 399     def log_date_time_string(self):
 400         """Return the current time formatted for logging."""
 401         now = time.time()
 402         year, month, day, hh, mm, ss, x, y, z = time.localtime(now)
 403         s = "%02d/%3s/%04d %02d:%02d:%02d" % (
 404                 day, self.monthname[month], year, hh, mm, ss)
 405         return s
 406
 407     weekdayname = ['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun']
 408
 409     monthname = [None,
 410                  'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun',
 411                  'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
 412
 413     def address_string(self):
 414         """Return the client address formatted for logging.
 415
 416         This version looks up the full hostname using gethostbyaddr(),
 417         and tries to find a name that contains at least one dot.
 418
 419         """
 420
 421         (host, port) = self.client_address
 422         try:
 423             name, names, addresses = socket.gethostbyaddr(host)
 424         except socket.error, msg:
 425             return host
 426         names.insert(0, name)
 427         for name in names:
 428             if '.' in name: return name
 429         return names[0]
 430
 431
 432     # Essentially static class variables
 433
 434     # The version of the HTTP protocol we support.
 435     # Don't override unless you know what you're doing (hint: incoming
 436     # requests are required to have exactly this version string).
 437     protocol_version = "HTTP/1.0"
 438
 439     # The Message-like class used to parse headers
 440     MessageClass = mimetools.Message
 441
 442     # Table mapping response codes to messages; entries have the
 443     # form {code: (shortmessage, longmessage)}.
 444     # See http://www.w3.org/hypertext/WWW/Protocols/HTTP/HTRESP.html
 445     responses = {
 446         200: ('OK', 'Request fulfilled, document follows'),
 447         201: ('Created', 'Document created, URL follows'),
 448         202: ('Accepted',
 449               'Request accepted, processing continues off-line'),
 450         203: ('Partial information', 'Request fulfilled from cache'),
 451         204: ('No response', 'Request fulfilled, nothing follows'),
 452
 453         301: ('Moved', 'Object moved permanently -- see URI list'),
 454         302: ('Found', 'Object moved temporarily -- see URI list'),
 455         303: ('Method', 'Object moved -- see Method and URL list'),
 456         304: ('Not modified',
 457               'Document has not changed singe given time'),
 458
 459         400: ('Bad request',
 460               'Bad request syntax or unsupported method'),
 461         401: ('Unauthorized',
 462               'No permission -- see authorization schemes'),
 463         402: ('Payment required',
 464               'No payment -- see charging schemes'),
 465         403: ('Forbidden',
 466               'Request forbidden -- authorization will not help'),
 467         404: ('Not found', 'Nothing matches the given URI'),
 468
 469         500: ('Internal error', 'Server got itself in trouble'),
 470         501: ('Not implemented',
 471               'Server does not support this operation'),
 472         502: ('Service temporarily overloaded',
 473               'The server cannot process the request due to a high load'),
 474         503: ('Gateway timeout',
 475               'The gateway server did not receive a timely response'),
 476
 477         }
 478
 479
 480 def test(HandlerClass = BaseHTTPRequestHandler,
 481          ServerClass = HTTPServer):
 482     """Test the HTTP request handler class.
 483
 484     This runs an HTTP server on port 8000 (or the first command line
 485     argument).
 486
 487     """
 488
 489     if sys.argv[1:]:
 490         port = string.atoi(sys.argv[1])
 491     else:
 492         port = 8000
 493     server_address = ('', port)
 494
 495     httpd = ServerClass(server_address, HandlerClass)
 496
 497     print "Serving HTTP on port", port, "..."
 498     httpd.serve_forever()
 499
 500
 501 if __name__ == '__main__':
 502     test()