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

[Lyndon Nerenberg]

On Aug 4, 2014, at 6:05 PM, Ken Hornstein <address@hidden> wrote:

> I can't speak for things like Received, but at least for headers
> containing comma-separated addresses you're supposed to (according to
> the RFC) add a comma and logically combine them together (we mostly do
> okay on that).

If I'm reading Norm's request the right way, he wants something that can easily 
be parsed with things like cut and awk.  I know that's what I would be looking 
for.  So, let's say you were given

  To: NMH Workers <address@hidden (NMH Workers Unite!)

you would get something like

  address@hidden
  phrase\tNMH Workers
  comment\t NMH Workers Unite!

The idea being each structured part of the header would be broken out with an 
identifying tag, and with the content attached in a completely un-encoded 
manner.  So you would have 

  TAG HT CONTENT EOL

(in pseudo-BNFO, with CONTENT being fully decoded utf8 text.

Looking at this I might have answered my own question, at least partially for 
the To: case.  Each recipient in the To: header would print its own 
addr/phrase/comment tuple.  You could separate recipients in the header by 
blank lines to let awk figure out the individual recipients.  But how do you 
deal with 

  mhparse to first-last

??

--lyndon

signature.asc
Description: Message signed with OpenPGP using GPGMail