On Tue, Oct 22, 2013 at 10:43 PM, John Cowan <xxxxxx@mercury.ccil.org> wrote:
Alex Shinn scripsit:

> I think it's handy to include the short property names because many
> regexp libraries allow for matching these with something like \p{Lu}.

They allow it, but it's not really the Right Thing.  Lu is an artifact
of the general classification system, which doesn't allow something to
be both a symbol and a letter.  Alphabetic is far better, and it's what
should be exposed.

I've fixed these (keeping the Pp names for reference),
and hope to have a new draft out soon, sorry for the long
delay.

I'm not sure I like having so many multiple names for patterns, however.
Pick one and stick to it, I think.

There is only one name, I'm just including additional
documentation that is helpful for users of the more
common regexp libraries.

There are however, multiple names in the sense we
support both "alpha" and "alphabetic," etc.  This
comes from SCSH and seems reasonable, so I'm not
inclined to change it now.

-- 
Alex