Re: mime QP and B headers in body

[Kjetil Torgrim Homme <kjetilho@haey.ifi.uio.no>]

[Dan Jacobson]:
>
>   > I make a daily spam report
>   > * Subject: =?BIG5?B?ttmhS...
>   > * Subject:=?big5?Q?=A7A=B...
>   > * Subject: Want a BIG Pen...
>   > * Subject: =?GB2312?B?1tC...
>   > * Subject: =?ISO-8859-1?B...
>   
>    If you normalized all the charsets to, say, iso-2022-* or utf-8, that
>   would probably make it easier to read.
>   
>   i could make a big program to make them all utf-8 which would
>   involve a lot of sed and awk and iconv which is very picky and i
>   dare not complain about lest i have another encounter with drepper

please quote properly!

it's not that hard to do:

#! /usr/bin/perl -w

use MIME::Base64;
use MIME::QuotedPrint;

sub recode {
    my ($charset) = shift;
    my $tmp = "/tmp/reco.$$";
    open (TMP, ">$tmp") || die;
    print TMP @_;
    close (TMP);
    my $ans = "";
    print "recode $charset..UTF8 < $tmp\n";
    open (RECODE, "recode <$tmp $charset..UTF8|") || die;
    while (<RECODE>) {
        $ans .= $_;
    }
    unlink ($tmp);
    $ans;
}

while (<>) {
    if (/=\?(\S+?)\?B\?(\S+)\?=/i) {
        my ($charset, $word) = ($1, $2);
        print recode ($charset, decode_base64 ($word)), "\n";
    } elsif (/=\?(\S+?)\?Q\?(\S+)\?=/i) {
        my ($charset, $word) = ($1, $2);
	$word =~ tr/_/ /;
        print recode ($charset, decode_qp ($word)), "\n";
    } else {
	print;
    }
}

oh, duh, this only decodes the first quoted-word on each line.
consider it an exercise for the reader.

-- 
Kjetil T.			|  read and make up your own mind
				|  http://www.cactus48.com/truth.html