view release on metacpan or  search on metacpan

lib/Mail/Abuse/Incident/Normalize.pm  view on Meta::CPAN

	return;
    }

    my $text;
    $rep->header($e->head);
    $rep->body(\$text);

    my $decoded_body = ($self->decode_parts($rep, 'any', $e) || '');
    $rep->body(\$decoded_body) if $decoded_body;

    $rep->normalized(ref $self);
    $self->html_parser(undef);
    $parser->filer->purge;

    return unless $rep->body and $ {$rep->body};

				# Phase 2 - This might be actually a
				# reply of forward. Remove any
				# indications from the message body

    $ {$rep->body} =~ s/
	^(([\t ]|[^\w\s])*	# 0 or more non-space, non-word chars
	  [^\w\s]+		# followed by one or more non-space, non-words
	  )+			# repeated one or more times,
	 [\t ]*			# followed by zero or more spaces.
	//xmsg;

				# Phase 3 - Try to recognize a
				# timezone in the body. If this fails,
				# try at the Date: header. Finally,
				# use UTC as the guess 

    # These timezones were taken w/o permission from Time::Zone.
    # Thanks to the authors anyway :)

    my %Zones =
	(
	 "GMT"	=> '+0000',	# Greenwich Mean
	 "UT"	=> '+0000',	# Universal (Coordinated)
	 "UTC"	=> '+0000',
	 "WET"	=> '+0000',	# Western European
	 "WAT"	=> '-0100',	# West Africa
	 "AT"	=> '-0200',	# Azores
	 "FNT" 	=> '-0200',	# Brazil Time (Extreme East)
	 "BRT"	=> '-0300',	# Brazil Time (East Standard)
	 "MNT"	=> '-0400',	# Brazil Time (West Standard)
	 "EWT"	=> '-0400',	# U.S. Eastern War Time
	 "AST"	=> '-0400',	# Atlantic Standard
	 "VET"	=> '-0400',	# Venezuela Standard Time
	 "EST"	=> '-0500',	# Eastern Standard
	 "ACT"	=> '-0500',	# Brazil Time (Extreme West - Acre)
	 "CST"	=> '-0600',	# Central Standard
	 "MST"	=> '-0700',	# Mountain Standard
	 "PST"	=> '-0800',	# Pacific Standard
	 "YST"	=> '-0900',	# Yukon Standard
	 "HST"	=> '-1000',	# Hawaii Standard
	 "CAT"	=> '-1000',	# Central Alaska
	 "AHST"	=> '-1000',	# Alaska-Hawaii Standard
	 "NT"	=> '-1100',	# Nome
	 "IDLW"	=> '-1200',	# International Date Line West
	 "CET"	=> '+0100',	# Central European
	 "MEZ"	=> '+0100',	# Central European (German)
	 "ECT"	=> '+0100',	# Central European (French)
	 "MET"	=> '+0100',	# Middle European
	 "MEWT"	=> '+0100',	# Middle European Winter
	 "SWT"	=> '+0100',	# Swedish Winter
	 "SET"	=> '+0100',	# Seychelles
	 "FWT"	=> '+0100',	# French Winter
	 "EET"	=> '+0200',	# Eastern Europe, USSR Zone 1
	 "UKR"	=> '+0200',	# Ukraine
	 "BT"	=> '+0300',	# Baghdad, USSR Zone 2
	 "IT"	=> '+0330',	# Iran
	 "ZP4"	=> '+0400',	# USSR Zone 3
	 "ZP5"	=> '+0500',	# USSR Zone 4
	 "IST"	=> '+0530',	# Indian Standard 
	 "ZP6"	=> '+0600',	# USSR Zone 5
	 "WST"	=> '+0800',	# West Australian Standard
	 "HKT"	=> '+0800',	# Hong Kong
	 "CCT"	=> '+0800',	# China Coast, USSR Zone 7
	 "JST"	=> '+0900',	# Japan Standard, USSR Zone 8
	 "KST"	=> '+0900',	# Korean Standard
	 "CAST"	=> '+0930',	# Central Australian Standard 
	 "EAST"	=> '+1000',	# Eastern Australian Standard
	 "GST"	=> '+1000',	# Guam Standard, USSR Zone 9
	 "NZT"	=> '+1200',	# New Zealand
	 "NZST"	=> '+1200',	# New Zealand Standard
	 "IDLE"	=> '+1200',	# International Date Line East
	 );

    # In the following matches, we'll attempt to use the inverted
    # message to avoid mistaking a timezone in a forwarded header for
    # our target timestamp...

    my $rev = join "\n", reverse split /\n/, $ {$rep->body};

				# Step 3a - Try to locate a numeric
				# timezone in the format [+-]\d\d:?\d\d

#    warn "Body: ${$rep->body}\n*********************\n";
#    warn "Rev: $rev\n******************\n";

    if ($rev =~ m!([-+]\d\d:?\d\d)!ms) 
    {
	$rep->tz($1);
    }

				# Step 3b - If failed, try to locate the
				# longest timezone posssible

    unless ($rep->tz)
    {
	my $zone = '';

	for my $tz ( sort keys %Zones )
	{
	    next if length($tz) < length($zone);
	    next unless ($rev =~ m/\W ${tz} \W /msx
			 or $rev =~ m/^ ${tz} \W /msx
			 or $rev =~ m/\W ${tz} $ /msx
			 or $rev =~ m/^ ${tz} $ /msx);
#	    warn "*** Zone $tz matched\n";