[rt-devel] RT 3.0.3rc2

[Autrijus Tang]

On Thu, Jun 19, 2003 at 12:11:48PM +0400, Dmitry Sivachenko wrote:
> 1) When I am sending http://mitya.pp.ru/rt-bad-email1.dat, the following
> appears in the maillog:

That mail has the extremely evil header:

    From: =?UNKNOWN?Q?=C8=ED=F2=E5=F0=ED=E5=F2_=EC=E0=F0=EA=E5=F2=E8=ED=E3?=

and RT simply does not know how to handle this encoding.

I have committed the patch below into my branch which takes care of
this case by applying the same _GuessCharset routine to MIME words.

As for your second problem, I can't duplicate it -- but maybe you'd
like to check your @EmailOutputEncoding?

Thanks,
/Autrijus/

--- //depot/RT/rt/lib/RT/I18N.pm	2003/06/19 08:03:42
+++ //depot/RT/rt/lib/RT/I18N.pm	2003/06/19 08:34:19
@@ -273,7 +273,11 @@
 	# now we have got a decoded subject, try to convert into
 	# utf-8 encoding
 	unless ($charset =~ m/utf-8/i) {
-	    Encode::from_to($enc_str, $charset, "utf8");
+	    eval { Encode::from_to($enc_str, $charset, "utf8") };
+	    if ($@) {
+		$charset = _GuessCharset( $enc_str );
+		Encode::from_to($enc_str, $charset, "utf8");
+	    }
 	}
 
 	$str .= $prefix . $enc_str . $trailing;
@@ -296,17 +300,33 @@
 sub _FindOrGuessCharset {
     my $entity = shift;
     my $head = $entity->head;
-    my $charset;
-    my $fallback = 'iso-8859-1';
 
-    return $charset if ($charset = $head->mime_attr("content-type.charset"));
+    if ($head->mime_attr("content-type.charset")) {
+	return $head->mime_attr("content-type.charset");
+    }
 
     my $body = $entity->bodyhandle or return;
-    return 'utf-8' if Encode::is_utf8( $body->as_string );
+    return _GuessCharset( $head->as_string . $body->as_string );
+}
+
+# }}}
+
+
+# {{{ _GuessCharset
+
+=head2 _GuessCharset STRING
+
+use Encode::Guess to try to figure it out the string's encoding.
+
+=cut
+
+sub _GuessCharset {
+    my $fallback = 'iso-8859-1';
+    my $charset;
 
     if ( @RT::EmailInputEncodings and eval { require Encode::Guess; 1 } ) {
 	Encode::Guess->set_suspects(@RT::EmailInputEncodings);
-	my $decoder = Encode::Guess->guess( $head->as_string . $body->as_string );
+	my $decoder = Encode::Guess->guess( $_[0] );
 
 	if ( ref $decoder ) {
 	    $charset = $decoder->name;
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://pallas.eruditorum.org/pipermail/rt-devel/attachments/20030619/29aecb59/attachment.pgp