Previously, when trying to encode the form fields for a request, the

fields were always being encoded as UTF-8, even when they were a binary

type (e.g., unicode text encoded as UTF-8). This lead to encoding

errors and has been rectified.
A new unit test has been added to ensure this heaviour does not

regress.

Ran unit tests. The new unit test fails without the rest of this patch

applied.
Successfully used the RBClient to send requests with fields

containing string values and unicode values.
Successfully uploaded a review request that had UTF-8 characters in the

diff, summary, and description.