Fallback decoding of subject (and other headers?)
I sometimes get emails where 'Charset="GB2312"' is specified, so that the body is decoded and displayed properly, but the GB2312-encoded subject is not. I suppose this is because the subject itself is part of the header and the 'Charset="GB2312"' is not applied to it.
For these emails, apparently Gmail decodes the subject as intended but Geary does not. I'm using Geary 0.12.4.
Can we have a heuristic like the following? If the subject does not decode properly (e.g. considered to be ASCII and yet have characters >= 128), try decoding with the encoding specified in the Charset. If that works, use that. Otherwise continue to treat the subject as ASCII.
Edited by Michael Gratton