[Groff] \[uXXXX] and ASCII glyph names

[Werner LEMBERG]

Bruno,


your changes to make the Unicode->groff name mapping work for
non-composite entities also has interesting consequences.

For example, the glyph entity \[u0078] (which people expect to be a
lowercase `x' character) is equal to \[x] -- but this is the glyph
with the name `\x', not `x', which is surprising for many users.

What do you think of adding glyph name `\X', (`X' a single letter,
taken from the printable area of ASCII 0x00-0x7F), making them equal
to glyph name `X'?  Then the user sees an `x' if she types `\[u0078]'.

The only problematic case is \[u002D] which can't be mapped to \[-]
but which should probably be made equal to U+2010.

Comments?


    Werner