Re: [Nmh-workers] Thoughts: header/address parsing

[Ken Hornstein]

>> Looking at it ... the grammer is not huge.  I think it will be easier
>> than the RFC-5322 address parser.
>
>It's not just the bison part, it's all the other code that
>builds the internal Content, etc., structures.  That'll have to
>be ported.  (I assume that would be the approach.)

Well, all of this pretty much assumes that the code to generate a Content
structure is going to have to be redone from scratch.  In general, my
thinking was:

- Slurp in headers via our header parsing routine, build Content structure.
- Parse the address headers, as needed.
- Parse the MIME headers, as needed.
- Assuming a multipart message, find the next content boundary, read in
  content headers via our header parsing routines.  Parse the MIME headers
  as needed.

You get the idea.

--Ken