Encoding of kde-commits mails

[Harri Porten]

Hi!

This might appear like a rather academic issue to some but I'm using a 
mail server written by rather pedantic people which turns a cosmetic issue 
into a fatal one for me :)

The post-commit.pl script that produces the log mails send to kde-commits 
adds the headers

  push(@head, "Content-Type: text/plain; charset=UTF-8\n");
  push(@head, "Content-Transfer-Encoding: 8bit\n");

marking the message as utf-8 encoded but does not ensure that the body 
really uses this encoding. Well, there's of course the check that 
produces the

  [UTF-8 ENCODING PROBLEMS]

warnings. They provide a useful hint but the mail sent out to all 
subscribers still isn't RFC compliant. A current sample that my mail 
server complained about is this one:

   http://lists.kde.org/?l=kde-commits&m=114337899122478&w=2

I therefore suggest to use a flag like Encode::FB_PERLQQ

  http://perl.active-venture.com/lib/Encode.html

for decode_utf8() that will replace malformed characters by \xHH. The 
result will then be converted by encode_utf8() for insertion into the 
mail. I volunteer to prepare a patch - hopefully assisted by Thiago who 
added the already existing validity check :)

Harri.

P.S. I've followed the recent thread about copyright on this list but I 
still believe that using the copyright symbol will cause us more 
(technical) trouble (like this one) than (legal) good.