Re: [patch] treatment of wrongly encoded 8-bit messages

[Albrecht Dreß <albrecht dress arcor de>]

Am 16.02.03 19:39 schrieb(en) Albrecht Dreß:
> deeper into this... My first guess is that libmutt sees some of the mime 
> headers, assumes a charset (us-ascii?), and replaces the 8-bit chars by 
> "?". The result is that libbalsa_utf8_sanitize (and my patch) will never 
> see the original chars. Stay tuned...

O.k., I got it... libmutt assumes us-ascii for all mime messages without 
the charset parameter (see parse.c). I think we should just omit the 
conversion from us-ascii to utf8, which does not do any harm if the 
messages are encoded correctly, but enables the user-defined fallback 
conversion. The file to change (apply it after my first patch) is again 
libmutt/handler.c:

~~~~~snip here~~~~~
--- handler.c.orig      Sun Feb 16 21:48:20 2003
+++ handler.c   Sun Feb 16 21:38:55 2003
@@ -1716,7 +1716,14 @@
    if (istext && s->flags & M_CHARCONV)
    {
      char *charset = mutt_get_parameter ("charset", b->parameter);
+#ifdef LIBMUTT
+    /* balsa 2: don't convert us-ascii. It should not do any harm with
+       correctly encoded messages, but enables fallback converion to a
+       user-defined codeset for broken messages. */
+    if (charset && Charset && strcmp(charset, "us-ascii"))
+#else
      if (charset && Charset)
+#endif
        cd = mutt_iconv_open (Charset, charset, M_ICONV_HOOK_FROM);
    }

~~~~~snip here~~~~~

Cheers, Albrecht.

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  Albrecht Dreß  -  Johanna-Kirchner-Straße 13  -  D-53123 Bonn (Germany)
        Phone (+49) 228 6199571  -  mailto:albrecht.dress@arcor.de
_________________________________________________________________________