Re: CFNetwork?
Re: CFNetwork?
- Subject: Re: CFNetwork?
- From: Béla Kiss <email@hidden>
- Date: Thu, 27 Dec 2001 23:46:46 -0800
You're right. Constructing the CFHTTPMessageRef your way seems to make
CFHTTPMessageIsHeaderComplete() happy. However (and there is always a
however), if one uses that constructed CFHTTPMessageRef for calling
readStream = CFReadStreamCreateForHTTPRequest(kCFAllocatorDefault, request);
openReadStreamSucceeded = CFReadStreamOpen(readStream);
causes CFReadStreamOpen() to crash inside _CFStreamGetCallbackPtr.
If one constructs the CFHTTPMessageRef the long way, this crash doesn't
happen (but nothing much seems to happen - maybe because
CFHTTPMessageIsHeaderComplete() isn't happy).
Grrr...
bk
on 2001.12.27 13.31, Dustin Mierau at email@hidden wrote:
>
Well alrighty then. Yea I just looked at the rfc, you're right...
>
>
So this works:
>
>
CFHTTPMessageRef request;
>
NSString* requestString;
>
>
request = CFHTTPMessageCreateEmpty( kCFAllocatorDefault, TRUE );
>
requestString = [NSString stringWithString:@"GET / HTTP/1.1\nHost:
>
www.apple.com\nUser-Agent: Mozilla/4.0\n\n"];
>
>
CFHTTPMessageAppendBytes( request, [requestString cString],
>
[requestString cStringLength] );
>
>
if( CFHTTPMessageIsHeaderComplete( request ) )
>
NSLog( @"Header is complete" );
>
else
>
NSLog( @"Header is NOT complete" );
>
>
CFRelease( request );
>
>
Output is:
>
>
"Header is complete"
>
>
I tried it the other way, and sure enough, I get "Header is NOT
>
complete"..odd.
--
Bila Kiss telno. 604.264.6113
Arborealis Research email. email@hidden
3322 W 36th Avenue
Vancouver, BC, CANADA V6N 2R9