Nordlöw:
I was specifically interested in something that exercises (random samples) potentially _all_ code points for string, wstring and dstring (all code units that is).
That's harder. Generating all uints and then testing if it's a Unicode dchar seems possible.
Bye, bearophile
