mail: fix _encode to be more correct on Python 3
This code appears to be on the wrong side of the law in Python 2, at
least some of the time. In Python 3, it's definitely wrong in places,
but fortunately that's easy to fix.
Differential Revision: https://phab.mercurial-scm.org/D3953
--- a/mercurial/mail.py Mon Jul 16 14:14:27 2018 -0400
+++ b/mercurial/mail.py Mon Jul 16 17:48:03 2018 -0400
@@ -252,10 +252,27 @@
order. Tries both encoding and fallbackencoding for input. Only as
last resort send as is in fake ascii.
Caveat: Do not use for mail parts containing patches!'''
+ sendcharsets = charsets or _charsets(ui)
+ if not isinstance(s, bytes):
+ # We have unicode data, which we need to try and encode to
+ # some reasonable-ish encoding. Try the encodings the user
+ # wants, and fall back to garbage-in-ascii.
+ for ocs in sendcharsets:
+ try:
+ return s.encode(pycompat.sysstr(ocs)), ocs
+ except UnicodeEncodeError:
+ pass
+ except LookupError:
+ ui.warn(_('ignoring invalid sendcharset: %s\n') % ocs)
+ else:
+ # Everything failed, ascii-armor what we've got and send it.
+ return s.encode('ascii', 'backslashreplace')
+ # We have a bytes of unknown encoding. We'll try and guess a valid
+ # encoding, falling back to pretending we had ascii even though we
+ # know that's wrong.
try:
s.decode('ascii')
except UnicodeDecodeError:
- sendcharsets = charsets or _charsets(ui)
for ics in (encoding.encoding, encoding.fallbackencoding):
try:
u = s.decode(ics)
@@ -263,7 +280,7 @@
continue
for ocs in sendcharsets:
try:
- return u.encode(ocs), ocs
+ return u.encode(pycompat.sysstr(ocs)), ocs
except UnicodeEncodeError:
pass
except LookupError: