> From: Erik Reuter <[EMAIL PROTECTED]>
> On Tue, Sep 24, 2002 at 09:40:08PM -0500, The Fool wrote:
> > > From: Erik Reuter <[EMAIL PROTECTED]>
> > > You're joking, right? You CANNOT KNOW THAT THEY ARE DUPLICATES.
They
> > > look like the start of a new message.
> >
> > Only in the case where the first message has no body, ie no message
> > header like statements, in which case the second message would show
up as
> > the body of the first message.
>
> No. Below is an mbox file containing what was meant to be 2 messages,
> where one message quotes, in entirety, a third message. The correct
> parsing would come up with 2 emails, with the 1st email ending with
> "...message #1." But without some standard (such as mangling From to
> >From), what the parser will come up with is 3 messages, with the 1st
> email ending with "...body text."
Doesent matter if the sender munges from to >from, not all senders do,
and from isn't neccessarily the first message header. (Getting message
headers mixed up between messages is bad). In any case your sender
didn't munge the from below either, so it would still come up wrong. Two
'kludges' do not make a right.
> -----------------
> From X Y
> Header1: akdjf
> Header2: kjdfj
> Subject: This is message #1
>
> This is message #1's body text.
>
> From A B
> Header1: adkjfk
> Subject: This is a message nested into another message
>
> This is some body text of the nested message.
> This is more nested message body text.
>
> This is the last line of body text of message #1.
>
> From Q F
> Header1: kjdklfj
> Header2: kjdfkjf
> Subject: This is message #2
>
> This is the body text of message #2
> ---------------
>
>
> >I just said it wasn't 100% and a kludge.
>
> And you claimed that it was possible to do something that is
> impossible. You made a number of incorrect statements. And nonsensical
> claims (could write a better parser in 1 minute). In order to write
Slight bit of hyperbole. I also claimed creating 'kludges' was an
extremely bad idea, and that I would not in fact write a 'kludgey'
program.
> a parser, you need to know all the available message syntaxes that
> your parser might encounter, and you obviously lack the background and
Those sytaxes are changing and are different from different senders.
Status: U
Return-Path: <[EMAIL PROTECTED]>
Received: from www.mccmedia.com ([206.204.15.162])
by hazard.mail.atl.earthlink.net (Earthlink Mail Service) with ESMTP id
17U2tT7XH3Nl3qG0
for <[EMAIL PROTECTED]>; Tue, 24 Sep 2002 23:04:52 -0400 (EDT)
Received: from www.mccmedia.com (IDENT:mailman@localhost [127.0.0.1])
by www.mccmedia.com (8.11.6/8.11.6) with ESMTP id g8P3Dqs21195;
Tue, 24 Sep 2002 20:13:52 -0700
Received: from erikreuter
(207-172-190-161.c4-0.hlb-ubr1a.hlb-ubr.nj.cable.rcn.com
[207.172.190.161])
by www.mccmedia.com (8.11.6/8.11.6) with ESMTP id g8P3Djs21187
for <[EMAIL PROTECTED]>; Tue, 24 Sep 2002 20:13:46 -0700
Received: from ereuter by erikreuter with local (Exim 3.35 #1 (Debian))
id 17u2TX-0008Fz-00
for <[EMAIL PROTECTED]>; Tue, 24 Sep 2002 23:04:31 -0400
Date: Tue, 24 Sep 2002 23:04:31 -0400
> knowledge of the possibilities. You could be the best programmer in the
> world, but if your program doesn't solve the right problem, then it
> isn't worth much. You need to listen (read) and think before you type
if
> you want to be productive.
>
> > You can't tell what the first message header will be, nor the last.
> > Which is why you always design things that specify size, always. In
any
> > case, this kludge would produce something that would show both
messages,
> > whether they ended up nested or not.
>
> It would split one message, incorrectly, into two. And the size is not
> specified in the original mbox format, so you can't use the size and
> maintain backwards compatibility.
I don't concern myself with backwards compatibility. I concern myself
with quality. Writing a kludgey program in response to someone else's
kludged up format is not a really good idea.
> > > > Also you would assume that when you started getting duplicate
headers
> > > > that it would be a possible message start.
> > >
> > > How can you ignore AND assume it is a message start? Have you been
> > > drinking?
> >
> > Message BODY start.
>
> There is no standard delimiter for message body start nor for message
> body end.
Which is why it should be based on size.
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l
