Okta decode error

[adorton-adobe]

The Okta connector makes a "decode" call that is only valid in Python 2.7. In py27, the strings present in Okta user objects are unicode strings, which can be decoded. However, in py3, they are plain strings (which are unicode by default). The str object does not have a decode function.

In py3, the call to decode causes an uncaught exception.

https://github.com/adobe-apiplatform/user-sync.py/blob/v2/user_sync/connector/directory_okta.py#L382

To resolve the issue, we should pass the string through six.text_type to normalize the string type.