- From: Henrik Frystyk Nielsen <frystyk@w3.org>
- Date: Tue, 24 Oct 1995 23:36:10 -0400
- To: garnaat@wrc.xerox.com (mitch garnaat cacao garnaat@wbst128)
- Cc: www-lib@w3.org, lib-friends@w3.org
> I've recently started using the Libwww to incorporate web access into > a modeling environment that we have developed internally. I really > need to have support for multipart/mixed MIME types and it doesn't > seem to be a part of the latest Libwww distribution. Interesting coincidence - I was just thinking about how I could implement it for the next release! This will officially get out to W3C members December 1st (with an alpha release leading up to it) and go into public domain one month after. It will consist of two stream - one for generating and one for parsing. The generating stream would be a bit like the structured stream with the following methods: typedef struct _MultipartClass { /* ... Normal stream stuff... */ BOOL (*begin_part) (HTMultipart * me, const char * boundary); BOOL (*end_part) (HTMultipart * me); BOOL (*prolog) (HTMultipart * me, const char * prolog) BOOL (*epilog) (HTMultipart * me, const char * epilog) } MultipartClass; "boundary" can be NULL in which case it is generated automatically. The parser will work like the current MIME parser as it will call the stream stack for each part that it finds. Anybody want to help doing this ;-? > I came across a message in the mail archive from Sept 94 that > discussed a new MIME parser that was being developed for the library. > It specifically stated that multipart was not going to be handled > initially but sort of implied that it would come along eventually. > What's the current status? -- Henrik Frystyk Nielsen, <frystyk@w3.org> World-Wide Web Consortium, MIT/LCS NE43-356 545 Technology Square, Cambridge MA 02139, USA
Received on Tuesday, 24 October 1995 23:36:16 UTC