Hi all,
I'm using 'Mutt 1.6 (be82a2fde82f) (2016-04-01)' compiled from the
mercurial repository.
Sometimes I receive binary attachments (PDF mostly) that have the
content type information wrongly set. When saving these attachments,
mutt tries to decode them and produces invalid files as a result.
For instance, I received recently a PDF attachment with the following
metadata:
text/x-unknown-content-type; charset=us-ascii; name="file.pdf"
When pressing 's' in the attachment menu mutt ends up calling
mutt_decode_attachment (attach.c, line 771), which corrupts the content.
I was hit first by this problem back in 2012 (bug #3568). It was fixed
then (changeset b9f9e3147eb4, I think), but I'm experiencing it again.
best,
aitor
