Found some code that strips out the attachments and writes them as files. I tried it on 2 test files, one with 3 parts, another with about 15 parts. In both cases the script wrote out the attachments correctly. Here is that code with a 'print' added ..

use Email::MIME;
use Email::MIME::Attachment::Stripper;
use File::Slurp qw(slurp write_file);

my $infile  = '/home/******/Mail/.family.directory/1280915907.6583.I9x
+0z:2,S';
my $m = Email::MIME->new( scalar slurp $infile );
my $s = Email::MIME::Attachment::Stripper->new( $m, 'force_filename' =
+> 1 );

print $s->message;  #displays "Email::MIME=HASH(0xe0fca8)"

foreach my $attachment ( $s->attachments ) {
write_file( $attachment->{filename},
{ buf_ref => \$attachment->{payload} } )
or die "Can't write $attachment->{filename}: $!\n";
}
[download]

It displays - Email::MIME=HASH(0xe0fca8) How do I get the contents of part one into a variable so that I can them extract the email addresses please ?

Seems the attachment stripper either doesn't work, or I have it coded incorrectly. Am currently running a script without the attachment stripper. It has been running for an hour or so, and those messages appear all the time - "Complex regular subexpression recursion limit (32766) exceeded at /usr/share/perl5/Email/Address.pm line 108." This most likely has something to do with a regex on the attachments. Hence the need to process email files without attachments.

Was looking through some small scripts I have here that look only for "From:", "To:", "Cc" and "Bcc". That led to using Email::Simple ..

#!/usr/bin/env perl
#
use strict;
use warnings;
use File::Find;
use Email::Simple;
use File::Slurp qw( read_file );

my $directory = '/home/******/Mail/.family.directory/Browne, Bill & Ma
+rtha';
my $outfile = 'output2.txt';

my @found_files;
find( sub { push @found_files, $File::Find::name }, $directory );

foreach(@found_files) {
  my $file = "$_";
      
  if (-f $file) {
    print $_,"\n";
    my $intext = File::Slurp::read_file( $file );
    
    my $mail            = Email::Simple->new($intext);
    my $from_header     = $mail->header("From");
    my $to_header       = $mail->header("To");
    my $date_header     = $mail->header("Date");
    my $cc_header       = $mail->header("CC");
    my $bcc_header      = $mail->header("BCC");

    my @emails = "";
    push @emails, ($from_header, $to_header);

    if( length $cc_header ) {
      push @emails, $cc_header;
    }
    if( length $bcc_header ) {
      push @emails, $bcc_header;
    }
  
    File::Slurp::write_file( $outfile, {append => 1 }, join("\n", @ema
+ils ) );

  }
}
[download]

This took about 2 seconds to process all the 592 emails, and successfully output the names and emails to a file. Just a few observations:

My use of an array needs improving

I'm unsure if the $mail->header("CC"); will also read line/s with "Cc" or "cc". The same is true for BCC.

Where there are a lot of emails, I need to format them so that every "," is replaced so that it becomes a seperate entry in the array. At present it is one large string with email names/address, seperated with a comma. (Will need to be careful where a "," is in the name though). How do I do that ?

The last solution is not exactly what I need. It misses all the (text or html) attachments that have email address in them. This code is more like what I need ..

#!/usr/bin/perl

# Print the types of messages in the folder.  Multi-part messages will
# be shown with all their parts.
#
# This code can be used and modified without restriction.
# Mark Overmeer, <********@overmeer.net>, 9 nov 2001

use warnings;
use strict;
use lib '..', '.';

use Mail::Box::Manager 2.00;
use Email::Address;
use File::Slurp qw( write_file );

my $outfile = 'multipart4_out.txt';

sub show_type($;$);

#
# Get the command line arguments.
#

die "Usage: $0 folderfile\n"
    unless @ARGV==1;

my $filename = shift @ARGV;

#
# Open the folder
#

my $mgr    = Mail::Box::Manager->new;

my $folder = $mgr->open
   ( $filename
   , extract    => 'LAZY'   # never take the body unless needed
   );                       #  which saves memory and time.

die "Cannot open $filename: $!\n"
    unless defined $folder;

#
# List all messages in this folder.
#

my @messages = $folder->messages;
print "Mail folder $filename contains ", scalar @messages, " messages:
+\n";

my $counter  = 1;
foreach my $message (@messages)
{   printf "%3d. ", $counter++;
    print $message->get('Subject') || '<no subject>', "\n";

    show_type $message;
    get_header_values($message);
}

sub get_header_values($)
{
    my $msg    = shift;
    # return all Mail::Message objects
    my @from    = $msg->from;
    my $sender  = $msg->sender;
    my $subject = $msg->subject;
    my $msgid   = $msg->messageId;
    my @to      = $msg->to;
    my @cc      = $msg->cc;
    my @bcc     = $msg->bcc;
    my @dest    = $msg->destinations;
    my $replyto   = $msg->get('Reply-To');
    # need code here to use @from, @to, @cc , @bcc and $replyto, and p
+ush out those emails (name/email address) to an array,
    #      then output to $outfile, like below
    #File::Slurp::write_file( $outfile, {append => 1 }, join("\n", @em
+ails) );
}

sub show_type($;$)
{   my $msg    = shift;
    my $indent = (shift || '') . '    ';   # increase indentation

    print $indent, " type=", $msg->get('Content-Type'), ', '
      , $msg->size, " bytes\n";

    if($msg->isMultipart)
    {   foreach my $part ($msg->parts)
    { 
        my $content_type = $part->contentType;

        # the only parts we are interested in are text and html - bypa
+ss all others
        if (defined($content_type) && $content_type eq 'text/plain')
        or (defined($content_type) && $content_type eq 'text/html')
        {
        my @emails = Email::Address->parse($part->body);    #this isn'
+t working properly
        #File::Slurp::write_file( $outfile, {append => 1 }, join("\n",
+ @emails) );
        }
        show_type $part, $indent;
        }
    }
}

#
# Finish
#

$folder->close;
[download]

But there are some errors when running it ..

"my" variable $part masks earlier declaration in same statement at multipart4.pl line 97. main::get_header_values() called too early to check prototype at multipart4.pl line 57. syntax error at multipart4.pl line 92, near ") or" syntax error at multipart4.pl line 100, near "}"

I don't think the "OR" is constructed properly. Why is there a msg on the sub get_header_values ?