I don't quite understand what you mean by "map table" in questions 1 & 2, but so far as the following goes:

Another Problem:  Encoding Example: "Entw�rfe"

The Mullberry Client sends "Entw�rfe" encoded: Entw&APw-rfe
Following the Rules, I have changed an Base64 Encoder to handle the mod. UTF7:
but the results of Encoding shouldn't really change a lot. My result of Encoding was this Entw&/A-rfe


So what's right now for the Character "�" following mod. UTF7?

  Character -->   �
  mullberry -->   &APw-
  mine        -->   &/A-

Why could mine be right?
I have checked some sources in the web. There I found a site where I could online Encode text into Base64 and for the Character "�"
I got the same Code like have with my encoding: /A.


Why could mine be wrong?
I've got another Base64 Encoder (but without source code) and I get the encoding like in Mullberry.


So what's the right Encoding? Why is that the right? What could probably be my fault?

UTF-7 (IMAP-modified or not) is a conversion of UTF-16, so each character you pass in is a two-byte unicode value (most significant byte first). So, the base-64 conversion is of the byte stream { 0x00, 0xfc }, giving "APw=", and therefore "&APw-" in modified UTF-7.


You are incorrectly base-64 converting { 0xfc }, which isn't UTF-16 (it's half a character). That "/A==", as the base64 conversion. Even if it were the correct input, the modified UTF-7 in this case would be "&,A-", I believe (/ becomes ,).

--Grant



Reply via email to