Lib/SocketServer.py

   1 """Generic socket server classes.
   2
   3 This module tries to capture the various aspects of defining a server:
   4
   5 For socket-based servers:
   6
   7 - address family:
   8         - AF_INET{,6}: IP (Internet Protocol) sockets (default)
   9         - AF_UNIX: Unix domain sockets
  10         - others, e.g. AF_DECNET are conceivable (see <socket.h>
  11 - socket type:
  12         - SOCK_STREAM (reliable stream, e.g. TCP)
  13         - SOCK_DGRAM (datagrams, e.g. UDP)
  14
  15 For request-based servers (including socket-based):
  16
  17 - client address verification before further looking at the request
  18         (This is actually a hook for any processing that needs to look
  19          at the request before anything else, e.g. logging)
  20 - how to handle multiple requests:
  21         - synchronous (one request is handled at a time)
  22         - forking (each request is handled by a new process)
  23         - threading (each request is handled by a new thread)
  24
  25 The classes in this module favor the server type that is simplest to
  26 write: a synchronous TCP/IP server.  This is bad class design, but
  27 save some typing.  (There's also the issue that a deep class hierarchy
  28 slows down method lookups.)
  29
  30 There are five classes in an inheritance diagram, four of which represent
  31 synchronous servers of four types:
  32
  33         +------------+
  34         | BaseServer |
  35         +------------+
  36               |
  37               v
  38         +-----------+        +------------------+
  39         | TCPServer |------->| UnixStreamServer |
  40         +-----------+        +------------------+
  41               |
  42               v
  43         +-----------+        +--------------------+
  44         | UDPServer |------->| UnixDatagramServer |
  45         +-----------+        +--------------------+
  46
  47 Note that UnixDatagramServer derives from UDPServer, not from
  48 UnixStreamServer -- the only difference between an IP and a Unix
  49 stream server is the address family, which is simply repeated in both
  50 unix server classes.
  51
  52 Forking and threading versions of each type of server can be created
  53 using the ForkingMixIn and ThreadingMixIn mix-in classes.  For
  54 instance, a threading UDP server class is created as follows:
  55
  56         class ThreadingUDPServer(ThreadingMixIn, UDPServer): pass
  57
  58 The Mix-in class must come first, since it overrides a method defined
  59 in UDPServer! Setting the various member variables also changes
  60 the behavior of the underlying server mechanism.
  61
  62 To implement a service, you must derive a class from
  63 BaseRequestHandler and redefine its handle() method.  You can then run
  64 various versions of the service by combining one of the server classes
  65 with your request handler class.
  66
  67 The request handler class must be different for datagram or stream
  68 services.  This can be hidden by using the request handler
  69 subclasses StreamRequestHandler or DatagramRequestHandler.
  70
  71 Of course, you still have to use your head!
  72
  73 For instance, it makes no sense to use a forking server if the service
  74 contains state in memory that can be modified by requests (since the
  75 modifications in the child process would never reach the initial state
  76 kept in the parent process and passed to each child).  In this case,
  77 you can use a threading server, but you will probably have to use
  78 locks to avoid two requests that come in nearly simultaneous to apply
  79 conflicting changes to the server state.
  80
  81 On the other hand, if you are building e.g. an HTTP server, where all
  82 data is stored externally (e.g. in the file system), a synchronous
  83 class will essentially render the service "deaf" while one request is
  84 being handled -- which may be for a very long time if a client is slow
  85 to reqd all the data it has requested.  Here a threading or forking
  86 server is appropriate.
  87
  88 In some cases, it may be appropriate to process part of a request
  89 synchronously, but to finish processing in a forked child depending on
  90 the request data.  This can be implemented by using a synchronous
  91 server and doing an explicit fork in the request handler class
  92 handle() method.
  93
  94 Another approach to handling multiple simultaneous requests in an
  95 environment that supports neither threads nor fork (or where these are
  96 too expensive or inappropriate for the service) is to maintain an
  97 explicit table of partially finished requests and to use select() to
  98 decide which request to work on next (or whether to handle a new
  99 incoming request).  This is particularly important for stream services
 100 where each client can potentially be connected for a long time (if
 101 threads or subprocesses cannot be used).
 102
 103 Future work:
 104 - Standard classes for Sun RPC (which uses either UDP or TCP)
 105 - Standard mix-in classes to implement various authentication
 106   and encryption schemes
 107 - Standard framework for select-based multiplexing
 108
 109 XXX Open problems:
 110 - What to do with out-of-band data?
 111
 112 BaseServer:
 113 - split generic "request" functionality out into BaseServer class.
 114   Copyright (C) 2000  Luke Kenneth Casson Leighton <lkcl@samba.org>
 115
 116   example: read entries from a SQL database (requires overriding
 117   get_request() to return a table entry from the database).
 118   entry is processed by a RequestHandlerClass.
 119
 120 """
 121
 122 # Author of the BaseServer patch: Luke Kenneth Casson Leighton
 123
 124 # XXX Warning!
 125 # There is a test suite for this module, but it cannot be run by the
 126 # standard regression test.
 127 # To run it manually, run Lib/test/test_socketserver.py.
 128
 129 __version__ = "0.4"
 130
 131
 132 import socket
 133 import sys
 134 import os
 135
 136 __all__ = ["TCPServer","UDPServer","ForkingUDPServer","ForkingTCPServer",
 137            "ThreadingUDPServer","ThreadingTCPServer","BaseRequestHandler",
 138            "StreamRequestHandler","DatagramRequestHandler",
 139            "ThreadingMixIn", "ForkingMixIn"]
 140 if hasattr(socket, "AF_UNIX"):
 141     __all__.extend(["UnixStreamServer","UnixDatagramServer",
 142                     "ThreadingUnixStreamServer",
 143                     "ThreadingUnixDatagramServer"])
 144
 145 class BaseServer:
 146
 147     """Base class for server classes.
 148
 149     Methods for the caller:
 150
 151     - __init__(server_address, RequestHandlerClass)
 152     - serve_forever()
 153     - handle_request()  # if you do not use serve_forever()
 154     - fileno() -> int   # for select()
 155
 156     Methods that may be overridden:
 157
 158     - server_bind()
 159     - server_activate()
 160     - get_request() -> request, client_address
 161     - verify_request(request, client_address)
 162     - server_close()
 163     - process_request(request, client_address)
 164     - close_request(request)
 165     - handle_error()
 166
 167     Methods for derived classes:
 168
 169     - finish_request(request, client_address)
 170
 171     Class variables that may be overridden by derived classes or
 172     instances:
 173
 174     - address_family
 175     - socket_type
 176     - allow_reuse_address
 177
 178     Instance variables:
 179
 180     - RequestHandlerClass
 181     - socket
 182
 183     """
 184
 185     def __init__(self, server_address, RequestHandlerClass):
 186         """Constructor.  May be extended, do not override."""
 187         self.server_address = server_address
 188         self.RequestHandlerClass = RequestHandlerClass
 189
 190     def server_activate(self):
 191         """Called by constructor to activate the server.
 192
 193         May be overridden.
 194
 195         """
 196         pass
 197
 198     def serve_forever(self):
 199         """Handle one request at a time until doomsday."""
 200         while 1:
 201             self.handle_request()
 202
 203     # The distinction between handling, getting, processing and
 204     # finishing a request is fairly arbitrary.  Remember:
 205     #
 206     # - handle_request() is the top-level call.  It calls
 207     #   get_request(), verify_request() and process_request()
 208     # - get_request() is different for stream or datagram sockets
 209     # - process_request() is the place that may fork a new process
 210     #   or create a new thread to finish the request
 211     # - finish_request() instantiates the request handler class;
 212     #   this constructor will handle the request all by itself
 213
 214     def handle_request(self):
 215         """Handle one request, possibly blocking."""
 216         try:
 217             request, client_address = self.get_request()
 218         except socket.error:
 219             return
 220         if self.verify_request(request, client_address):
 221             try:
 222                 self.process_request(request, client_address)
 223             except:
 224                 self.handle_error(request, client_address)
 225                 self.close_request(request)
 226
 227     def verify_request(self, request, client_address):
 228         """Verify the request.  May be overridden.
 229
 230         Return True if we should proceed with this request.
 231
 232         """
 233         return True
 234
 235     def process_request(self, request, client_address):
 236         """Call finish_request.
 237
 238         Overridden by ForkingMixIn and ThreadingMixIn.
 239
 240         """
 241         self.finish_request(request, client_address)
 242         self.close_request(request)
 243
 244     def server_close(self):
 245         """Called to clean-up the server.
 246
 247         May be overridden.
 248
 249         """
 250         pass
 251
 252     def finish_request(self, request, client_address):
 253         """Finish one request by instantiating RequestHandlerClass."""
 254         self.RequestHandlerClass(request, client_address, self)
 255
 256     def close_request(self, request):
 257         """Called to clean up an individual request."""
 258         pass
 259
 260     def handle_error(self, request, client_address):
 261         """Handle an error gracefully.  May be overridden.
 262
 263         The default is to print a traceback and continue.
 264
 265         """
 266         print '-'*40
 267         print 'Exception happened during processing of request from',
 268         print client_address
 269         import traceback
 270         traceback.print_exc() # XXX But this goes to stderr!
 271         print '-'*40
 272
 273
 274 class TCPServer(BaseServer):
 275
 276     """Base class for various socket-based server classes.
 277
 278     Defaults to synchronous IP stream (i.e., TCP).
 279
 280     Methods for the caller:
 281
 282     - __init__(server_address, RequestHandlerClass)
 283     - serve_forever()
 284     - handle_request()  # if you don't use serve_forever()
 285     - fileno() -> int   # for select()
 286
 287     Methods that may be overridden:
 288
 289     - server_bind()
 290     - server_activate()
 291     - get_request() -> request, client_address
 292     - verify_request(request, client_address)
 293     - process_request(request, client_address)
 294     - close_request(request)
 295     - handle_error()
 296
 297     Methods for derived classes:
 298
 299     - finish_request(request, client_address)
 300
 301     Class variables that may be overridden by derived classes or
 302     instances:
 303
 304     - address_family
 305     - socket_type
 306     - request_queue_size (only for stream sockets)
 307     - allow_reuse_address
 308
 309     Instance variables:
 310
 311     - server_address
 312     - RequestHandlerClass
 313     - socket
 314
 315     """
 316
 317     address_family = socket.AF_INET
 318
 319     socket_type = socket.SOCK_STREAM
 320
 321     request_queue_size = 5
 322
 323     allow_reuse_address = False
 324
 325     def __init__(self, server_address, RequestHandlerClass):
 326         """Constructor.  May be extended, do not override."""
 327         BaseServer.__init__(self, server_address, RequestHandlerClass)
 328         self.socket = socket.socket(self.address_family,
 329                                     self.socket_type)
 330         self.server_bind()
 331         self.server_activate()
 332
 333     def server_bind(self):
 334         """Called by constructor to bind the socket.
 335
 336         May be overridden.
 337
 338         """
 339         if self.allow_reuse_address:
 340             self.socket.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
 341         self.socket.bind(self.server_address)
 342
 343     def server_activate(self):
 344         """Called by constructor to activate the server.
 345
 346         May be overridden.
 347
 348         """
 349         self.socket.listen(self.request_queue_size)
 350
 351     def server_close(self):
 352         """Called to clean-up the server.
 353
 354         May be overridden.
 355
 356         """
 357         self.socket.close()
 358
 359     def fileno(self):
 360         """Return socket file number.
 361
 362         Interface required by select().
 363
 364         """
 365         return self.socket.fileno()
 366
 367     def get_request(self):
 368         """Get the request and client address from the socket.
 369
 370         May be overridden.
 371
 372         """
 373         return self.socket.accept()
 374
 375     def close_request(self, request):
 376         """Called to clean up an individual request."""
 377         request.close()
 378
 379
 380 class UDPServer(TCPServer):
 381
 382     """UDP server class."""
 383
 384     allow_reuse_address = False
 385
 386     socket_type = socket.SOCK_DGRAM
 387
 388     max_packet_size = 8192
 389
 390     def get_request(self):
 391         data, client_addr = self.socket.recvfrom(self.max_packet_size)
 392         return (data, self.socket), client_addr
 393
 394     def server_activate(self):
 395         # No need to call listen() for UDP.
 396         pass
 397
 398     def close_request(self, request):
 399         # No need to close anything.
 400         pass
 401
 402 class ForkingMixIn:
 403
 404     """Mix-in class to handle each request in a new process."""
 405
 406     active_children = None
 407     max_children = 40
 408
 409     def collect_children(self):
 410         """Internal routine to wait for died children."""
 411         while self.active_children:
 412             if len(self.active_children) < self.max_children:
 413                 options = os.WNOHANG
 414             else:
 415                 # If the maximum number of children are already
 416                 # running, block while waiting for a child to exit
 417                 options = 0
 418             try:
 419                 pid, status = os.waitpid(0, options)
 420             except os.error:
 421                 pid = None
 422             if not pid: break
 423             self.active_children.remove(pid)
 424
 425     def process_request(self, request, client_address):
 426         """Fork a new subprocess to process the request."""
 427         self.collect_children()
 428         pid = os.fork()
 429         if pid:
 430             # Parent process
 431             if self.active_children is None:
 432                 self.active_children = []
 433             self.active_children.append(pid)
 434             self.close_request(request)
 435             return
 436         else:
 437             # Child process.
 438             # This must never return, hence os._exit()!
 439             try:
 440                 self.finish_request(request, client_address)
 441                 os._exit(0)
 442             except:
 443                 try:
 444                     self.handle_error(request, client_address)
 445                 finally:
 446                     os._exit(1)
 447
 448
 449 class ThreadingMixIn:
 450     """Mix-in class to handle each request in a new thread."""
 451
 452     # Decides how threads will act upon termination of the
 453     # main process
 454     daemon_threads = False
 455
 456     def process_request_thread(self, request, client_address):
 457         """Same as in BaseServer but as a thread.
 458
 459         In addition, exception handling is done here.
 460
 461         """
 462         try:
 463             self.finish_request(request, client_address)
 464             self.close_request(request)
 465         except:
 466             self.handle_error(request, client_address)
 467             self.close_request(request)
 468
 469     def process_request(self, request, client_address):
 470         """Start a new thread to process the request."""
 471         import threading
 472         t = threading.Thread(target = self.process_request_thread,
 473                              args = (request, client_address))
 474         if self.daemon_threads:
 475             t.setDaemon (1)
 476         t.start()
 477
 478
 479 class ForkingUDPServer(ForkingMixIn, UDPServer): pass
 480 class ForkingTCPServer(ForkingMixIn, TCPServer): pass
 481
 482 class ThreadingUDPServer(ThreadingMixIn, UDPServer): pass
 483 class ThreadingTCPServer(ThreadingMixIn, TCPServer): pass
 484
 485 if hasattr(socket, 'AF_UNIX'):
 486
 487     class UnixStreamServer(TCPServer):
 488         address_family = socket.AF_UNIX
 489
 490     class UnixDatagramServer(UDPServer):
 491         address_family = socket.AF_UNIX
 492
 493     class ThreadingUnixStreamServer(ThreadingMixIn, UnixStreamServer): pass
 494
 495     class ThreadingUnixDatagramServer(ThreadingMixIn, UnixDatagramServer): pass
 496
 497 class BaseRequestHandler:
 498
 499     """Base class for request handler classes.
 500
 501     This class is instantiated for each request to be handled.  The
 502     constructor sets the instance variables request, client_address
 503     and server, and then calls the handle() method.  To implement a
 504     specific service, all you need to do is to derive a class which
 505     defines a handle() method.
 506
 507     The handle() method can find the request as self.request, the
 508     client address as self.client_address, and the server (in case it
 509     needs access to per-server information) as self.server.  Since a
 510     separate instance is created for each request, the handle() method
 511     can define arbitrary other instance variariables.
 512
 513     """
 514
 515     def __init__(self, request, client_address, server):
 516         self.request = request
 517         self.client_address = client_address
 518         self.server = server
 519         try:
 520             self.setup()
 521             self.handle()
 522             self.finish()
 523         finally:
 524             sys.exc_traceback = None    # Help garbage collection
 525
 526     def setup(self):
 527         pass
 528
 529     def handle(self):
 530         pass
 531
 532     def finish(self):
 533         pass
 534
 535
 536 # The following two classes make it possible to use the same service
 537 # class for stream or datagram servers.
 538 # Each class sets up these instance variables:
 539 # - rfile: a file object from which receives the request is read
 540 # - wfile: a file object to which the reply is written
 541 # When the handle() method returns, wfile is flushed properly
 542
 543
 544 class StreamRequestHandler(BaseRequestHandler):
 545
 546     """Define self.rfile and self.wfile for stream sockets."""
 547
 548     # Default buffer sizes for rfile, wfile.
 549     # We default rfile to buffered because otherwise it could be
 550     # really slow for large data (a getc() call per byte); we make
 551     # wfile unbuffered because (a) often after a write() we want to
 552     # read and we need to flush the line; (b) big writes to unbuffered
 553     # files are typically optimized by stdio even when big reads
 554     # aren't.
 555     rbufsize = -1
 556     wbufsize = 0
 557
 558     def setup(self):
 559         self.connection = self.request
 560         self.rfile = self.connection.makefile('rb', self.rbufsize)
 561         self.wfile = self.connection.makefile('wb', self.wbufsize)
 562
 563     def finish(self):
 564         if not self.wfile.closed:
 565             self.wfile.flush()
 566         self.wfile.close()
 567         self.rfile.close()
 568
 569
 570 class DatagramRequestHandler(BaseRequestHandler):
 571
 572     # XXX Regrettably, I cannot get this working on Linux;
 573     # s.recvfrom() doesn't return a meaningful client address.
 574
 575     """Define self.rfile and self.wfile for datagram sockets."""
 576
 577     def setup(self):
 578         try:
 579             from cStringIO import StringIO
 580         except ImportError:
 581             from StringIO import StringIO
 582         self.packet, self.socket = self.request
 583         self.rfile = StringIO(self.packet)
 584         self.wfile = StringIO()
 585
 586     def finish(self):
 587         self.socket.sendto(self.wfile.getvalue(), self.client_address)