From: Joseph Polanik 
James C. Sewell wrote:
>I said:
>>>The only difficulty is that according to the RFC, which is the
>>>"standard ruleset" for mail, the headers may appear in any order.
I'm not questioning your description of the RFC. It's just that, not being
a Borg, I don't directly interface with the electronic packets zipping down
the T-1/T-3 lines. Like everyone else, I use a mail client (or a web
browser) as an interface.
>B. A "superior" mail client will STILL receive headers in RANDOM order from
>whatever mailers the senders use, unless the mail client reorders headers
>which is outside what most mail programs need to do.
The input from the 'Net to the mail client is irrelevant to this
discussion. What's important is the output of that client to its user.
 
Now, the fact is that mail clients can be divided into two classes: 
Class 1 (User Friendly): A certain header line is consistently the first
line of each msg.
Class 2 (Substandard): Those that do not consistently place a certain line
first.
Note that not all the Class 1 mail clients will agree on which line should
go first. This mail reader may put the 'Subject:' line first. That one may
put the 'Date:' first. Eudora Pro puts 'Return-Path:' first.
That's not important. 
What *is* important is that writing a macro to strip unwanted header lines
from a .txt file of messages saved from a Class 1 browser is not especially
difficult.
>The statement I made is that if one writes a macro that will handle
>everyone's mail as was suggested, then it will be very difficult to write
>since the rules governing email state the headers may be in any order.
The statement I made was that one simple macro could handle the output of
any Class 1 mail client or browser. 
>But it would be much better to make a macro that can handle ALL legal mail
>messages before posting your masterpiece to the 'world at large'. ... I
simply state >that:
>  1. A great macro would handle ALL LEGAL combinations.
>  2. Such a macro is non-trivial to write.
You are right to suggest that an attempt to write a stripper type of macro
for Class 1 browsers would be a formidable task. It's a task that is beyond
me. However, until some professional programmer comes along to solve the
problem for those who insist on using mail clients inferior to some that
are available for free, the rest of us can get on with our lives.
>To make a macro specific to a particular mail reader is limited at best.  
If you are suggesting that this is what I have done you are mistaken. I am
claiming to have written a macro that could be used by anyone with *any*
Class 1 mail client. 
[Send bug reports to: Joseph Polanik ] [;-)
Actually, Stripper is more versatile than that, even. I use it to delete
unwanted header lines from Usenet articles saved to a .txt file by Netscape
Navigator 3.0. NN always starts the message with a characteristic line, so
Stripper knows where to do her thing. My latest version takes out some
extra garbage from header lines (Re: or RE[2]: etc.), recognizes some
unwanted footer lines (repetitious instructions for subscribing and
unsubscribing) and will optionally extract and collate the mailing
addresses remaining in the stripped files.
I am sure there is much more that can be done and I'd like to share ideas
with those who use Stripper or adapt it to their needs. However, I am not
up to the challenge of supporting Class 2 mail clients.  
Live Long and Prosper  \\//
                      \ 
Joseph Polanik
---
---------------