Fix encoding choosing for body content.Set charset for alternative views...

[AlexSte]

2 changes -

1. To correctly process body of emails with non ASCII encoding we need to use encoding that was set in server. if not explicitly set - use ASCII.
2. sometime is usefull to know which encoding is using in alternative views.(can be different from body encoding). so setting charset param to alternative view object will solve this issue.

[Desteny]

Thanks for this,
I can confirm the Bug, and i confirm your Pull request as a Solution.

I have tested your fix with several cCents in different Usecases, and i found a missing Part:

Your Solution doesn't work for "FetchOptions.Normal" cause of IMapClient.GetMessageData has a internal GetData call without passing the correct encoding!

I tryed to fix this in the GetMessageData, but i'm not able to split the Response into different Bodyparts with different encoding for each one.

So i made a different Workaround, changing FetchOptions.Normal to use "GetMessage".
This solves the Problem but it requires aditional server roundtrips (fetching header and body structure)

I hope this Feature/Fix will be added soon to NuGet cause its realy important for UTF8 Mails, and maybe with proper Solution for my Problem:

Here my Patch (no pull request cause of only one line changed)

diff --git a/ImapClient.cs b/ImapClient.cs
index 033317e..2e75a87 100644
--- a/ImapClient.cs
+++ b/ImapClient.cs
@@ -1103,7 +1103,9 @@ public class ImapClient : IImapClient
                    return GetMessage(uid, p => { return p.Type == ContentType.Text; },
                        seen, mailbox);
                default:
-                   return MessageBuilder.FromMIME822(GetMessageData(uid, seen, mailbox));
+                   return GetMessage(uid, p => { return true; },
+                       seen, mailbox);
+                   //return MessageBuilder.FromMIME822(GetMessageData(uid, seen, mailbox));
            }
        }

[jstedfast]

As an alternative, you guys might consider using MailKit instead as it properly deals with charsets.