Merge: Migration from ascii to code_point

As complement to #1262 and #1718, this PR solves part of both issues.

One does the migration from `ascii` to `code_point` since Unicode is a bit more than ASCII, let's try to use the right vocabulary when dealing with it.

The `ascii` service is however kept alive, but its contract is changed from Int to Byte to be used directly on a `Char` without the need to have the old `'c'.ascii.to_b`, now a simple `'c'.ascii` will do the exact same.

A further PR will be submitted later to prevent the use of `ascii` on non-ASCII literal characters.

Pull-Request: #1726
Reviewed-by: Alexis Laferrière <alexis.laf@xymus.net>
Reviewed-by: Jean Privat <jean@pryen.org>