Hi again. This is snipped from the RFC:
The "charset" parameter is used with some media types to define the character set (section 3.4) of the data. When no explicit charset parameter is provided by the sender, media subtypes of the "text" type are defined to have a default charset value of "ISO-8859-1" when received via HTTP. Data in character sets other than "ISO-8859-1" or its subsets MUST be labeled with an appropriate charset value. See section 3.4.1 for compatibility problems. The sender might be either the client or the server. The document following the headers MUST be encoded using the ISO-8859-1 charset if nothing else is explicitly stated in the Content-Type header. This means that the recipient MUST interpret the document according to the ISO-8859-1 charset if the sender has not explicitly stated that another charset is used in the Content-Type header. Wether this the charset is guessed while parsing the headers, or if one implicitly assumes ISO-8859-1 everywhere else in the code if no charset is specified in the Content-Type header, does not matter. It is clear from section 3.4.1 that no indicated charset DOES NOT mean that the indicated preference from the recipient has been heeded, but that the document actually is encoded using the ISO-8859-1 charset. Regards Rune --- Rune Sætre <[EMAIL PROTECTED]> NetCom as, Infrastruktur Telefon (mob): 934 34 285 .. On Wed, 16 Feb 2005, Stipe Tolj wrote: > Paul P Komkoff Jr wrote: > > > Replying to Stipe Tolj: > > > >>the behaviour of libcurl and gwlib's http.c? > > > > > > Yes. curl doesn't mangle headers in any way. > > hm... As I noticed, the section reading from the HTTP/1.1 RFC is more like > "this > is what the server should do", IMO. So the question is if it is up to the > server, then obiously out gwlib/http client module does "too much" from the > role > of the server. If we interpret it this way, it's up to the maintainer of the > application served by the HTTP server to ensure that HTTP headers do reflect > correct content-type charset encoding headers. > > I'm not quite sure. What would you mean reading the refenrenced section of > RFC 2616? > > In terms of libcurl support, this would mean, we would have to check > "afterwards" if this symptom is given and recode the charset like we do four > our > own gwlib/http, right? > > Stipe > > mailto:stolj_{at}_wapme.de > ------------------------------------------------------------------- > Wapme Systems AG > > Vogelsanger Weg 80 > 40470 DÃsseldorf, NRW, Germany > > phone: +49.211.74845.0 > fax: +49.211.74845.299 > > mailto:info_{at}_wapme-systems.de > http://www.wapme-systems.de/ > ------------------------------------------------------------------- > >
