> Because the digest runner is creating a MIMEMultipart instance on line 175 of
> src/mailman/runners/digest.py. Should we change that instead?
Good catch. Yes, I think we should change that. What I think I'd like to see is a test that creates a mime digest, and uses that to trigger the attribute errors, rather than creating a MIMEMultipart inside the tests themselves. Does that makes sense?
Then the fix would be to change MIMEDigester._make_message() to create the proper subclass. The question then is what that method should return. My first thought is a class that inherits both from MIMEMultipart and mailman.email.message.Message.
> Because the digest runner is creating a MIMEMultipart instance on line 175 of runners/ digest. py. Should we change that instead?
> src/mailman/
Good catch. Yes, I think we should change that. What I think I'd like to see is a test that creates a mime digest, and uses that to trigger the attribute errors, rather than creating a MIMEMultipart inside the tests themselves. Does that makes sense?
Then the fix would be to change MIMEDigester. _make_message( ) to create the proper subclass. The question then is what that method should return. My first thought is a class that inherits both from MIMEMultipart and mailman. email.message. Message.
What do you think?