Yes, you just need to install the Mimetools bundle from CPAN. We use Lotus Notes at work, and from time to time it barfs up on a MIME encoded message and is unable to pull it apart. In that case I save the file out as text and then run the following code on it, which dumps out the different sections into seperate files.

#! /usr/bin/perl -w

use strict;
use MIME::Parser;

my $parser = new MIME::Parser;
my $entity = $parser->parse(\*STDIN) or die "parse failed\n";
$entity->dump_skeleton;
[download]

That code is straight out of the documentation. For cleaning up the HTML, you might want to consider crafting something manually with HTML::Parser (a good learning exercise).

At first I planned on doing exactly what you suggest, but this cheap hack was "good enough," so I let that idea slide.

g r i n d e r

In addition to grinder I would like to advice you to have a further look at MIME::Entity.

Using its methods mime_type resp. effective_type on each MIME part you may find the "text/plain" part and process it.

You may use these methods on MIME::Parser objects.

############################################################
sub split_entity {
############################################################
  local $entity = shift; # needs a MIME::Entity object
  my $num_parts = $entity->parts; # how many mime parts?

  if ($num_parts) { # we have a multipart mime message

    foreach (1..$num_parts) {
      split_entity( $entity->parts($_ - 1) );
    }

  } else { # we have a single mime message/part

    # text
    if ($entity->effective_type =~ /^text\/(?!(html|enriched))/) {
      # do something
    } else {
      # do something different
    }
  }
}
[download]

alex pleiner <alex@zeitform.de>
zeitform Internet Dienste

thank you for the replies

I guess thats a 'no' on the existing scripts, I'm a little surprised there isn't already some perl module to deal with these things a little more intelligently..

the problem is, email includes some fun surprises like nested messages (with full headers) that most existing email readers are smart enough to understand, and process/display correctly. This isn't really a big deal, since we're not dealing with something very complicated here.. but this is a personal project, so I was hoping there was something already out there..

thanks again

Don't be that pessimistic. The MIME-Tools and the above sub do a great job for me separating text and non-text MIME components. As long as the mailer that sent the mail has ever heard about the relevant RFCs it is easy to handle.

If you don't believe, have a look at this webmailer (in German, but guessable) using the above sub.

For the fun surprises, notice the recursive call of that sub.

alex pleiner <alex@zeitform.de>
zeitform Internet Dienste