Did you write the MIME parser yourself or are you using an existing one? If MIME parsing is so tricky, have you considered using freely available implementation (such as Mozilla's libmime) at least in the interim?
Henry
On Thu, 18 Sep 2003 22:35:12 +0200 (CEST), Andreas Aardal Hanssen <[EMAIL PROTECTED]> wrote:
On Thu, 18 Sep 2003, Andreas Aardal Hanssen wrote:- There are errors with documents with multiparts insid multiparts. The error may be in parseFull() or they may be in getPart() - I'm investigating this now. The errors are that some multiparts are simply not parsed, that is they are skipped..
More on this: A concrete bug is that the end of a multipart is always the end of any enclosing multiparts, as the "foundendofpart" boolean in mime-parsefull.cc line 435 is set to true, making the whole function return 1, making its enclosing multipart parsing loop (around line 394) exit instead of continuing to parse more parts..
Parsing MIME is tricky, but I think the solution to this bug is quite simple. And I think the MIME parser _will_ look great once all the bugs get pulled out ;).
Andy
-- Henry Baragar Principal, Technical Architecture 416-453-5626 Instantiated Software Inc. http://www.instantiated.ca
