1 """Generic socket server classes.
3 This module tries to capture the various aspects of defining a server:
5 For socket-based servers:
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>
12 - SOCK_STREAM (reliable stream, e.g. TCP)
13 - SOCK_DGRAM (datagrams, e.g. UDP)
15 For request-based servers (including socket-based):
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)
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.)
30 There are five classes in an inheritance diagram, four of which represent
31 synchronous servers of four types:
38 +-----------+ +------------------+
39 | TCPServer |------->| UnixStreamServer |
40 +-----------+ +------------------+
43 +-----------+ +--------------------+
44 | UDPServer |------->| UnixDatagramServer |
45 +-----------+ +--------------------+
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
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:
56 class ThreadingUDPServer(ThreadingMixIn, UDPServer): pass
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.
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.
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.
71 Of course, you still have to use your head!
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.
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.
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
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).
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
110 - What to do with out-of-band data?
113 - split generic "request" functionality out into BaseServer class.
114 Copyright (C) 2000 Luke Kenneth Casson Leighton <lkcl@samba.org>
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.
122 # Author of the BaseServer patch: Luke Kenneth Casson Leighton
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.
139 import dummy_threading
as threading
141 __all__
= ["TCPServer","UDPServer","ForkingUDPServer","ForkingTCPServer",
142 "ThreadingUDPServer","ThreadingTCPServer","BaseRequestHandler",
143 "StreamRequestHandler","DatagramRequestHandler",
144 "ThreadingMixIn", "ForkingMixIn"]
145 if hasattr(socket
, "AF_UNIX"):
146 __all__
.extend(["UnixStreamServer","UnixDatagramServer",
147 "ThreadingUnixStreamServer",
148 "ThreadingUnixDatagramServer"])
152 """Base class for server classes.
154 Methods for the caller:
156 - __init__(server_address, RequestHandlerClass)
157 - serve_forever(poll_interval=0.5)
159 - handle_request() # if you do not use serve_forever()
160 - fileno() -> int # for select()
162 Methods that may be overridden:
166 - get_request() -> request, client_address
168 - verify_request(request, client_address)
170 - process_request(request, client_address)
171 - close_request(request)
174 Methods for derived classes:
176 - finish_request(request, client_address)
178 Class variables that may be overridden by derived classes or
184 - allow_reuse_address
188 - RequestHandlerClass
195 def __init__(self
, server_address
, RequestHandlerClass
):
196 """Constructor. May be extended, do not override."""
197 self
.server_address
= server_address
198 self
.RequestHandlerClass
= RequestHandlerClass
199 self
.__is
_shut
_down
= threading
.Event()
200 self
.__shutdown
_request
= False
202 def server_activate(self
):
203 """Called by constructor to activate the server.
210 def serve_forever(self
, poll_interval
=0.5):
211 """Handle one request at a time until shutdown.
213 Polls for shutdown every poll_interval seconds. Ignores
214 self.timeout. If you need to do periodic tasks, do them in
217 self
.__is
_shut
_down
.clear()
219 while not self
.__shutdown
_request
:
220 # XXX: Consider using another file descriptor or
221 # connecting to the socket to wake this up instead of
222 # polling. Polling reduces our responsiveness to a
223 # shutdown request and wastes cpu at all other times.
224 r
, w
, e
= select
.select([self
], [], [], poll_interval
)
226 self
._handle
_request
_noblock
()
228 self
.__shutdown
_request
= False
229 self
.__is
_shut
_down
.set()
232 """Stops the serve_forever loop.
234 Blocks until the loop has finished. This must be called while
235 serve_forever() is running in another thread, or it will
238 self
.__shutdown
_request
= True
239 self
.__is
_shut
_down
.wait()
241 # The distinction between handling, getting, processing and
242 # finishing a request is fairly arbitrary. Remember:
244 # - handle_request() is the top-level call. It calls
245 # select, get_request(), verify_request() and process_request()
246 # - get_request() is different for stream or datagram sockets
247 # - process_request() is the place that may fork a new process
248 # or create a new thread to finish the request
249 # - finish_request() instantiates the request handler class;
250 # this constructor will handle the request all by itself
252 def handle_request(self
):
253 """Handle one request, possibly blocking.
255 Respects self.timeout.
257 # Support people who used socket.settimeout() to escape
258 # handle_request before self.timeout was available.
259 timeout
= self
.socket
.gettimeout()
261 timeout
= self
.timeout
262 elif self
.timeout
is not None:
263 timeout
= min(timeout
, self
.timeout
)
264 fd_sets
= select
.select([self
], [], [], timeout
)
266 self
.handle_timeout()
268 self
._handle
_request
_noblock
()
270 def _handle_request_noblock(self
):
271 """Handle one request, without blocking.
273 I assume that select.select has returned that the socket is
274 readable before this function was called, so there should be
275 no risk of blocking in get_request().
278 request
, client_address
= self
.get_request()
281 if self
.verify_request(request
, client_address
):
283 self
.process_request(request
, client_address
)
285 self
.handle_error(request
, client_address
)
286 self
.close_request(request
)
288 def handle_timeout(self
):
289 """Called if no new request arrives within self.timeout.
291 Overridden by ForkingMixIn.
295 def verify_request(self
, request
, client_address
):
296 """Verify the request. May be overridden.
298 Return True if we should proceed with this request.
303 def process_request(self
, request
, client_address
):
304 """Call finish_request.
306 Overridden by ForkingMixIn and ThreadingMixIn.
309 self
.finish_request(request
, client_address
)
310 self
.close_request(request
)
312 def server_close(self
):
313 """Called to clean-up the server.
320 def finish_request(self
, request
, client_address
):
321 """Finish one request by instantiating RequestHandlerClass."""
322 self
.RequestHandlerClass(request
, client_address
, self
)
324 def close_request(self
, request
):
325 """Called to clean up an individual request."""
328 def handle_error(self
, request
, client_address
):
329 """Handle an error gracefully. May be overridden.
331 The default is to print a traceback and continue.
335 print('Exception happened during processing of request from', end
=' ')
336 print(client_address
)
338 traceback
.print_exc() # XXX But this goes to stderr!
342 class TCPServer(BaseServer
):
344 """Base class for various socket-based server classes.
346 Defaults to synchronous IP stream (i.e., TCP).
348 Methods for the caller:
350 - __init__(server_address, RequestHandlerClass, bind_and_activate=True)
351 - serve_forever(poll_interval=0.5)
353 - handle_request() # if you don't use serve_forever()
354 - fileno() -> int # for select()
356 Methods that may be overridden:
360 - get_request() -> request, client_address
362 - verify_request(request, client_address)
363 - process_request(request, client_address)
364 - close_request(request)
367 Methods for derived classes:
369 - finish_request(request, client_address)
371 Class variables that may be overridden by derived classes or
377 - request_queue_size (only for stream sockets)
378 - allow_reuse_address
383 - RequestHandlerClass
388 address_family
= socket
.AF_INET
390 socket_type
= socket
.SOCK_STREAM
392 request_queue_size
= 5
394 allow_reuse_address
= False
396 def __init__(self
, server_address
, RequestHandlerClass
, bind_and_activate
=True):
397 """Constructor. May be extended, do not override."""
398 BaseServer
.__init
__(self
, server_address
, RequestHandlerClass
)
399 self
.socket
= socket
.socket(self
.address_family
,
401 if bind_and_activate
:
403 self
.server_activate()
405 def server_bind(self
):
406 """Called by constructor to bind the socket.
411 if self
.allow_reuse_address
:
412 self
.socket
.setsockopt(socket
.SOL_SOCKET
, socket
.SO_REUSEADDR
, 1)
413 self
.socket
.bind(self
.server_address
)
414 self
.server_address
= self
.socket
.getsockname()
416 def server_activate(self
):
417 """Called by constructor to activate the server.
422 self
.socket
.listen(self
.request_queue_size
)
424 def server_close(self
):
425 """Called to clean-up the server.
433 """Return socket file number.
435 Interface required by select().
438 return self
.socket
.fileno()
440 def get_request(self
):
441 """Get the request and client address from the socket.
446 return self
.socket
.accept()
448 def close_request(self
, request
):
449 """Called to clean up an individual request."""
453 class UDPServer(TCPServer
):
455 """UDP server class."""
457 allow_reuse_address
= False
459 socket_type
= socket
.SOCK_DGRAM
461 max_packet_size
= 8192
463 def get_request(self
):
464 data
, client_addr
= self
.socket
.recvfrom(self
.max_packet_size
)
465 return (data
, self
.socket
), client_addr
467 def server_activate(self
):
468 # No need to call listen() for UDP.
471 def close_request(self
, request
):
472 # No need to close anything.
477 """Mix-in class to handle each request in a new process."""
480 active_children
= None
483 def collect_children(self
):
484 """Internal routine to wait for children that have exited."""
485 if self
.active_children
is None: return
486 while len(self
.active_children
) >= self
.max_children
:
487 # XXX: This will wait for any child process, not just ones
488 # spawned by this library. This could confuse other
489 # libraries that expect to be able to wait for their own
492 pid
, status
= os
.waitpid(0, 0)
495 if pid
not in self
.active_children
: continue
496 self
.active_children
.remove(pid
)
498 # XXX: This loop runs more system calls than it ought
499 # to. There should be a way to put the active_children into a
500 # process group and then use os.waitpid(-pgid) to wait for any
501 # of that set, but I couldn't find a way to allocate pgids
502 # that couldn't collide.
503 for child
in self
.active_children
:
505 pid
, status
= os
.waitpid(child
, os
.WNOHANG
)
510 self
.active_children
.remove(pid
)
511 except ValueError as e
:
512 raise ValueError('%s. x=%d and list=%r' % (e
.message
, pid
,
513 self
.active_children
))
515 def handle_timeout(self
):
516 """Wait for zombies after self.timeout seconds of inactivity.
518 May be extended, do not override.
520 self
.collect_children()
522 def process_request(self
, request
, client_address
):
523 """Fork a new subprocess to process the request."""
524 self
.collect_children()
528 if self
.active_children
is None:
529 self
.active_children
= []
530 self
.active_children
.append(pid
)
531 self
.close_request(request
)
535 # This must never return, hence os._exit()!
537 self
.finish_request(request
, client_address
)
541 self
.handle_error(request
, client_address
)
546 class ThreadingMixIn
:
547 """Mix-in class to handle each request in a new thread."""
549 # Decides how threads will act upon termination of the
551 daemon_threads
= False
553 def process_request_thread(self
, request
, client_address
):
554 """Same as in BaseServer but as a thread.
556 In addition, exception handling is done here.
560 self
.finish_request(request
, client_address
)
561 self
.close_request(request
)
563 self
.handle_error(request
, client_address
)
564 self
.close_request(request
)
566 def process_request(self
, request
, client_address
):
567 """Start a new thread to process the request."""
568 t
= threading
.Thread(target
= self
.process_request_thread
,
569 args
= (request
, client_address
))
570 if self
.daemon_threads
:
575 class ForkingUDPServer(ForkingMixIn
, UDPServer
): pass
576 class ForkingTCPServer(ForkingMixIn
, TCPServer
): pass
578 class ThreadingUDPServer(ThreadingMixIn
, UDPServer
): pass
579 class ThreadingTCPServer(ThreadingMixIn
, TCPServer
): pass
581 if hasattr(socket
, 'AF_UNIX'):
583 class UnixStreamServer(TCPServer
):
584 address_family
= socket
.AF_UNIX
586 class UnixDatagramServer(UDPServer
):
587 address_family
= socket
.AF_UNIX
589 class ThreadingUnixStreamServer(ThreadingMixIn
, UnixStreamServer
): pass
591 class ThreadingUnixDatagramServer(ThreadingMixIn
, UnixDatagramServer
): pass
593 class BaseRequestHandler
:
595 """Base class for request handler classes.
597 This class is instantiated for each request to be handled. The
598 constructor sets the instance variables request, client_address
599 and server, and then calls the handle() method. To implement a
600 specific service, all you need to do is to derive a class which
601 defines a handle() method.
603 The handle() method can find the request as self.request, the
604 client address as self.client_address, and the server (in case it
605 needs access to per-server information) as self.server. Since a
606 separate instance is created for each request, the handle() method
607 can define arbitrary other instance variariables.
611 def __init__(self
, request
, client_address
, server
):
612 self
.request
= request
613 self
.client_address
= client_address
629 # The following two classes make it possible to use the same service
630 # class for stream or datagram servers.
631 # Each class sets up these instance variables:
632 # - rfile: a file object from which receives the request is read
633 # - wfile: a file object to which the reply is written
634 # When the handle() method returns, wfile is flushed properly
637 class StreamRequestHandler(BaseRequestHandler
):
639 """Define self.rfile and self.wfile for stream sockets."""
641 # Default buffer sizes for rfile, wfile.
642 # We default rfile to buffered because otherwise it could be
643 # really slow for large data (a getc() call per byte); we make
644 # wfile unbuffered because (a) often after a write() we want to
645 # read and we need to flush the line; (b) big writes to unbuffered
646 # files are typically optimized by stdio even when big reads
652 self
.connection
= self
.request
653 self
.rfile
= self
.connection
.makefile('rb', self
.rbufsize
)
654 self
.wfile
= self
.connection
.makefile('wb', self
.wbufsize
)
657 if not self
.wfile
.closed
:
663 class DatagramRequestHandler(BaseRequestHandler
):
665 # XXX Regrettably, I cannot get this working on Linux;
666 # s.recvfrom() doesn't return a meaningful client address.
668 """Define self.rfile and self.wfile for datagram sockets."""
671 from io
import BytesIO
672 self
.packet
, self
.socket
= self
.request
673 self
.rfile
= BytesIO(self
.packet
)
674 self
.wfile
= BytesIO()
677 self
.socket
.sendto(self
.wfile
.getvalue(), self
.client_address
)