[Groff] Using \U for unicode representation

[Bernd Warken]

The \U escape seems to be unused.  So we could take it to represent
unicode characters.

It would make sense to define it as a function with a single argument, a
hexadecimal unicode character code.  For example, "\U'03A4" will either
map the argument to the corresponding character with code 03A4 or print
the official unicode representation "U+03A4" if it is not available.

Moreover there should be a tmac file that defines \[U+xxxx] strings or
character mappings for the existing groff special characters.

As there seems to be some unicode work going on in groff, there might be 
a way to automate this process.

This would be a long-term project, not meant for 1.16.  If you think it
makes sense, it should be set on the TODO list.

Bernd Warken <address@hidden>