where the method could block indefinitely if called just before the
event loop started running.  This also fixes the occasional freezes
witnessed in test_httpservers.

added the shutdown_request() which can perform shutdown before calling close.  This is needed for the ForkingMixIn because different close semantics are required for child and parent process.  shutdown_request(), for TCP servers, calls socket.shutdown() and then calls close_request().  Therefore, this is not an backwards incompatible change, since subclasses that continue to override close_request() continue to work.

close_request() (which can send a socket.shutdown()) must be called by the child process in a forking server.  The parent must merely close the socket handle.

some platforms may raise ENOTCONN if the stack has disconnected the socket on behalf of the peer.

Cumulative patch to http and xmlrpc

Move the newly introduced disable_nagle_algorithm flag into the StreamRequestHandler, where it is more appropriate.

Add a feature to disable the Nagle algorithm on sockets in TCPServer

place and fix all references to it.  Closes #2926.

I'm doing this case-only rename in a two-step process via a different directory
to hopefully avoid problems with case-insensitive filesystems.

Deprecated old name.

SocketServers. The core of the patch was written by Pedro Werneck, but any bugs
are mine. I've also rearranged the code for timeouts in order to avoid
interfering with the shutdown poll.

didn't create, in most cases. When there are max_children handlers running, it
will still wait for any child process, not just handler processes.

when trying to remove a pid that in not in the activechildren list.

.handle_timeout() method when no requests are received within the timeout period.

Patch #1599845: Add an option to disable the implicit calls to server_bind() and server_activate() in the constructors for TCPServer, SimpleXMLRPCServer and DocXMLRPCServer.

Bug #1531963: Make SocketServer.TCPServer's server_address always be equal to calling getsockname() on the server's socket.
Will backport.

Backport candidate.

robust. This makes socketserver's close() method callable repeatedly
without error - similar to other file-like objects.

In DatagramRequestHandler.setup(), the wfile initialization should be
StringIO.StringIO(), not StringIO.StringIO(slf.packet).

Bugfix candidate (all the way back to Python 1.5.2 :-).

ThreadingMixIn/TCPServer forgets close (Max Neunhöffer).

This ensures that handle_error() and close_request() are called when
an error occurs in the thread.

(I am not applying the second chunk of the patch, which moved the
finish() call into the finally clause in BaseRequestHandler's __init__
method; that would be a semantic change that I cannot accept at this
point - the data would be sent even if the handler raised an
exception.)

Solved with a helper method that calls finish_request() and then
close_request().  The code is by Max Neunhöffer.

loss for no reason.

Fix various serious problems:

- The ThreadingTCPServer class and its derived classes were completely
  broken because the main thread would close the request before the
  handler thread had time to look at it.  This was introduced by
  Ping's close_request() patch.  The fix moves the close_request()
  calls to after the handler has run to completion in the BaseServer
  class and the ForkingMixIn class; when using the ThreadingMixIn,
  closing the request is the handler's responsibility.

- The ForkingUDPServer class has always been been broken because the
  socket was closed in the child before calling the handler.  I fixed
  this by simply not calling server_close() in the child at all.

- I cannot get the UnixDatagramServer class to work at all.  The
  recvfrom() call doesn't return a meaningful client address.  I added
  a comment to this effect.  Maybe it works on other Unix versions.

- The __all__ variable was missing ThreadingMixIn and ForkingMixIn.

- Bumped __version__ to "0.4".

- Added a note about the test suite (to be checked in shortly).

can close the request connection when it's done handling it.

added test script and expected output file as well
this closes patch 103297.
__all__ attributes will be added to other modules without first submitting
a patch, just adding the necessary line to the test script to verify
more-or-less correct implementation.

for SocketServer.py (inherited by TCPServer)

Luke wrote:

  The socketserver code, with a little bit of tweaking, can be made
  sufficiently general to service "requests" of any kind, not just by sockets.
  The BaseServer class was created, for example, to poll a table in a MYSQL
  database every 2 seconds. each entry in the table can be allocated a
  Handler which deals with the entry.

  With this patch, using BaseServer and ThreadedServer classes, the creation
  of the server that reads and handles MySQL table entries instead of a
  socket was utterly trivial: about 50 lines of python code.

  You may consider this code to be utterly useless [why would anyone else
  want to do anything like this???] - you are entitled to your opinion. if you
  think so, then think of this: have you considered how to cleanly add SSL to
  the TCPSocketServer? What about using shared memory as the
  communications mechanism for a server, instead of sockets? What about
  communication using files?

  The SocketServer code is extremely good every useful. it's just that as it
  stands, it is tied to sockets, which is not as useful.

I heartily approve of this idea.

and wfile class variables (that the instance can also override).
Change the default for rfile to buffered, because that seems to make a
big difference in performance on some platforms.

An anti-patch is needed to revert the effect in CGIHTTPServer.py which
I'll check in momentarily.