[pywin32-bugs] [ pywin32-Patches-1937527 ] win32file.ConnectEx patch

[SourceForge.net <no...@so...>]

Patches item #1937527, was opened at 2008-04-08 19:32
Message generated for change (Comment added) made by mhammond
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=551956&aid=1937527&group_id=78018

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
Status: Closed
Resolution: Fixed
Priority: 5
Private: No
Submitted By: ionel (ionel_mc)
Assigned to: Nobody/Anonymous (nobody)
Summary: win32file.ConnectEx patch

Initial Comment:
A patch for ConnectEx.

called as follows:
win32file.ConnectEx(sock, (host, port), overlappedobj [, bufferobj])

here's an example:
import socket, win32file, pywintypes
s = socket.socket()
ol = pywintypes.OVERLAPPED()
s.bind(('0.0.0.0', 0)) # connectex requires the socket be bound beforehand

print win32file.ConnectEx(s, ("google.com", 80), ol, "GET / HTTP/1.1\r\n\r\n")
print win32file.GetOverlappedResult(s.fileno(), ol, 1)
ol = pywintypes.OVERLAPPED()
buff = win32file.AllocateReadBuffer(20480)
print win32file.WSARecv(s, buff, ol, 0)
print win32file.GetOverlappedResult(s.fileno(), ol, 1)        
print buff


I've tested this on win xp.
I use getaddrinfo to process the (host, port) tuple.

Please review.


----------------------------------------------------------------------

>Comment By: Mark Hammond (mhammond)
Date: 2009-02-07 12:58

Message:
Roger has this in his py3k branch so its now merged to the trunk - thanks
for your patience.

----------------------------------------------------------------------

Comment By: Mark Hammond (mhammond)
Date: 2009-02-07 12:57

Message:
Roger has this in his py3k branch so its now merged to the trunk - thanks
for your patience.

----------------------------------------------------------------------

Comment By: ionel (ionel_mc)
Date: 2008-05-09 10:36

Message:
Logged In: YES 
user_id=1189761
Originator: YES

To be completely honest I don't really know how to use the tuples returned
from the socket.getaddrinfo - since they are very far from the C addrinfo
struct I need to use in the ConnectEx call, also, i don't think the python
function call overhead to get that addrinfo and make whatever conversions
back to the struct is worth it.

I would add some sort of demo but i need some suggestions (it seems I lack
imagination right now).

I don't cover the corner cases (like bad params) in the test_connectex.




----------------------------------------------------------------------

Comment By: ionel (ionel_mc)
Date: 2008-05-09 10:25

Message:
Logged In: YES 
user_id=1189761
Originator: YES

File Added: test_connectex.py

----------------------------------------------------------------------

Comment By: Mark Hammond (mhammond)
Date: 2008-05-05 20:33

Message:
Logged In: YES 
user_id=14198
Originator: NO

This looks pretty good.  I'd probably prefer for getaddrinfo() to be
called by the user of this function, and the necessary info passed in - is
there a reason not to do that?  Also, would you be so kind as to add
something to the win32\test directory (and maybe even the win32\Demos
directory :) that covers this?

Thanks!

----------------------------------------------------------------------

Comment By: ionel (ionel_mc)
Date: 2008-04-24 18:22

Message:
Logged In: YES 
user_id=1189761
Originator: YES

wops, forgot something
File Added: win32file.i-4.patch

----------------------------------------------------------------------

Comment By: ionel (ionel_mc)
Date: 2008-04-24 18:22

Message:
Logged In: YES 
user_id=1189761
Originator: YES

wops, forgot something
File Added: win32file.i-4.patch

----------------------------------------------------------------------

Comment By: ionel (ionel_mc)
Date: 2008-04-24 17:57

Message:
Logged In: YES 
user_id=1189761
Originator: YES

ok i've made some changes:
i've moved the wsaioctl in the function body 
  - the msdn docs didn't say if the pointer i'm getting is always the
same
  - wsaioctl requires a wsainit before, i feel win32file is the wrong
place to do that and i don't want to force users to import the socket
module (for wsainit) before win32file
still looking for some feedback on this one though.

i've changed the buffer handling parts to use the 'buffer api'

File Added: win32file.i-3.patch

----------------------------------------------------------------------

Comment By: ionel (ionel_mc)
Date: 2008-04-24 17:57

Message:
Logged In: YES 
user_id=1189761
Originator: YES

ok i've made some changes:
i've moved the wsaioctl in the function body 
  - the msdn docs didn't say if the pointer i'm getting is always the
same
  - wsaioctl requires a wsainit before, i feel win32file is the wrong
place to do that and i don't want to force users to import the socket
module (for wsainit) before win32file
still looking for some feedback on this one though.

i've changed the buffer handling parts to use the 'buffer api'

File Added: win32file.i-3.patch

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=551956&aid=1937527&group_id=78018