sphinxc0re / rust-base24 Goto Github PK

Since the protocol specifies no padding, it makes little sense to encode a &[u8] and letting the encoder fail on non-divisible-by-4 array lengths.
If the user wants to encode a &[u8] then the user could use byteorder or some custom code to reinterpret the slice as u32.

This will make encoding infallible.

In the java version, I throw exceptions when the input has the wrong length or if invalid characters are present in the string to be decoded. But exceptions are normal error handling in Java but not in rust.

I suggest that encode and decode functions return Result types instead, for easier error handling in a rust way.