RAW/UCS string codes?

[dpkristensen]

Should allow proprietary text formats by defining a "RAW" string code. Ex: RAW-8, etc... Terminators and Padding characters may be specified if needed, meaning the specification writer only has to define a compatible length determination method.

This would have the effect for all intents and purposes of having 1:1 correspondence with Unicode code points when it is passed to the application; which means it starts to sound a lot like UCS (ISO/IEC 10646) in practical terms.

I think UCS should be supported for sure (UCS-8, etc...); the RAW encoding would potentially need some way to integrate into the parser. So the source encoding probably must be available to the application layer in some form. RAW encodings should not have any textual representation in a GUI, other than as a hex-coded byte stream of the appropriate width (ex: 2-byte hex for RAW-8).

One benefit of RAW encoding support is to use RAW-8 as a fall-back for unknown encodings.