RE: Web Service that requires XML document in HTTP Header

[Nick Townsend <Nick.Townsend@xxxxxxxxxxxxxxx>]

Scott,

Just when I thought/hoped this project had been dropped it has
suddenly been resurrected.  I would like to submit my changes
to the HTTPAPI to allow additional long headers to be added to
a request for your approval, and hopefully for inclusion in a
future release.  The code changes are limited to two source
members, HTTPAPI_H and HTTPAPIR4.  What is the best way for me
to get the code to you?

Regards,

Nick
__________________________________
Nick Townsend
Technical Leader
Endsleigh Insurance Services Limited
Telephone: +44 (0)1242 866866 ext. 6426

-----Original Message-----
From: ftpapi-bounces@xxxxxxxxxxxxxxxxxxxxxx [mailto:ftpapi-bounces@xxxxxxxxxxxxxxxxxxxxxx] On Behalf Of Scott Klement
Sent: 13 February 2008 19:54
To: HTTPAPI and FTPAPI Projects
Subject: Re: Web Service that requires XML document in HTTP Header

Hi Nick,

How big do you need your additional header to be?

Just in case it's not clear -- it's possible to call the SendReq()
procedure twice, three times, etc, until you've sent your whole request.
  Right now, the code sends it all at once, but that's not actually
required, it can send it piece-by-piece if you want it to.... the data
all gets appended together in the socket, anyway.

The server keeps receiving data (no matter how many times you call
SendReq to send it) until it receives 4 consecutive bytes containing
x'0d0a0d0a'.  (x'0d0a' is the ASCII representation of CRLF).    So feel
free to call SendReq multiple times to send the whole string, as long as
you only send the extra CRLF on the last call to SendReq().


Nick Townsend wrote:
> Looking at the API code it seems that the actual restriction on the
> size of the header field is actually less than 32,767 bytes because it
> has to fit into field wwReqChain along with all the other headers, and
> wwReqChain itself has a maximum length of 32,767 bytes (presumably the
> HTTPAPI code is limited to 32,767 rather than 65,535 for backwards
> compatibility with earlier compiler releases). I'm going to have to
> work out whether this is a feasible solution for us because of this
> restriction.
-----------------------------------------------------------------------
This is the FTPAPI mailing list.  To unsubscribe, please go to:
http://www.scottklement.com/mailman/listinfo/ftpapi
-----------------------------------------------------------------------
-----------------------------------------------------------------------
This is the FTPAPI mailing list.  To unsubscribe, please go to:
http://www.scottklement.com/mailman/listinfo/ftpapi
-----------------------------------------------------------------------