Strongly recommend the D language to support new feature of “compile time char type identification”

[Paul Findlay]

> Strongly recommend the D language to support new feature of "compile
> time char type identification"
Something like templates? 

The tango project (http://dsource.org/projects/tango/) has a lot of code
that works with all three UTF encodings because it is templated.

My only wish is that phobos or tango came with a function that "scrubbed"
supposedly utf-8 strings of invalid values.

 - Paul