- From: William Bardwell <bardwell@curl.com>
- Date: Tue, 24 Jun 2003 18:49:04 -0400
- To: www-lib@w3.org
I am trying to use libwww (latest version) to build a general purpose
HTTP API, and I am having trouble with some very basic things.
I want to get both the data and the headers for all requests, even
ones that are considered failures, and regardless of if I have
turned on the cache.
I am finding that the HTAnchor has the headers only for successful
requests with the Cache turned on, and HTResponse never seems to have all
of the headers. I am finding that if I get the output/debug streams in
RAW format, then I do see the headers on the stream (and I can parse
them myself), but the data is sometimes in chunked mode, and I really
don't want to re-implement chunk decoding. Is there some way to get
both the headers and the decoded data at the same time?
(Perhaps I need to register a converter that can see the raw stream, but
then pass everything that it sees on to a stream stack that WWW_SOURCE
would normally construct? But I can't figure out what exactly I do to
do that, and that seems like a lot of work for something fairly simple.
(I guess that I register a converter for some www/x-my-header-capture
that captures the headers, and then have that use
HTStreamStack(
"www/x-my-header-capture",
WWW_SOURCE,
HTRequest_outputStream(request),
request,
NO)
that it passes the data on to, but that seems like then the stream
is getting the wrong format (since it asked for my custom format),
instead I want my converter to somehow be always part of the
WWW_RAW -> WWW_SOURCE stack...)
William Bardwell
wbardwel@curl.com
Received on Tuesday, 24 June 2003 18:49:06 UTC