diff options
| author | Dan Winship <danw@src.gnome.org> | 2001-05-31 06:05:36 +0800 |
|---|---|---|
| committer | Dan Winship <danw@src.gnome.org> | 2001-05-31 06:05:36 +0800 |
| commit | fc68923b1110d26c8d5185adaa701dd9a34991fa (patch) | |
| tree | f8ab367629c78118161574768fb7e4af80680a24 /shell | |
| parent | 2bb495b6eb8e1e7e2c22691a5b3891879baa5f76 (diff) | |
| download | gsoc2013-evolution-fc68923b1110d26c8d5185adaa701dd9a34991fa.tar.gz gsoc2013-evolution-fc68923b1110d26c8d5185adaa701dd9a34991fa.tar.zst gsoc2013-evolution-fc68923b1110d26c8d5185adaa701dd9a34991fa.zip | |
Use the config-specified default character set (which in turn defaults to
* e-msg-composer.c (build_message): Use the config-specified
default character set (which in turn defaults to the locale
character set) as the default for encoding non-US-ASCII messages,
assuming it works. Fall back to camel_charset_best() if it fails.
Also, pick the Content-Transfer-Encoding based on how the data
looks in the destination charset, not in UTF8.
While I was here, I also refactored a bunch and tried to reduce
redundant code. There are still too many places that set the
transfer-encoding on the plaintext part though... Also, I think
some more cleanliness could happen if the sign/encrypt interfaces
returned CamelMultiparts instead of CamelMimeParts.
svn path=/trunk/; revision=10058
Diffstat (limited to 'shell')
0 files changed, 0 insertions, 0 deletions