mime processing

[David Relson]

At 11:55 AM 12/29/02, Gyepi SAM wrote:

>Hi David,
>
>I have committed a couple of changes to the mime code.

Good!

>In the change to yyinput, I removed the code to eat attachments.
>I'll put it back, but it needs more work and further changes.
>We need to note when we encounter mime boundaries, both start and end,
>so we know how to process the contents.  I'll do this later.

What do you think we should do with attachments?  Documents might be of 
interest to bogofitler.  Images would decode as binary.

I think it might also be useful to have separate text and html 
parsers.  That would let us remove all the html tags from the text 
parser.  A simple html parser could eat everything between angle brackets.

>BTW, the name yyinput is a bad choice since flex uses that when compiled
>with a C++ compiler.

That was ESR's doing.  I'm innocent.  Since the name is internal to 
lexer.l, we can easily rename it.  If you have a good name in mind, use 
it.  My ideas are lexer_getline() and lgets(), neither of which is great.


>OK, time to go out and play with the family!

Yeah, we're off to the movies this afternoon.  My wife and daughter are 
going to see Harry Potter and my sons and I are due for LOTR.