ucd-generate: add joining-type sub-command

[wezm]

This PR makes use of the ability to parse ArabicShaping.txt to implement joining type table generation. To do this it also needs to do general category lookups, so I extracted a expand_into_categories method from the general_category command in order for it to be able to do this.

Sample output:

// DO NOT EDIT THIS FILE. IT WAS AUTOMATICALLY GENERATED BY:
//
//  ucd-generate joining-type --rust-enum /home/wmoore/Downloads/ucd-12.1
//
// ucd-generate is available on crates.io.

#[derive(Clone, Copy, Debug, Hash, Eq, PartialEq)]
pub enum JoiningType {
  DualJoining, JoinCausing, LeftJoining, NonJoining, RightJoining,
  Transparent,
}

pub const JOINING_TYPE: &'static [(u32, u32, JoiningType)] = &[
  (0, 172, JoiningType::NonJoining), (173, 173, JoiningType::Transparent),
  (174, 767, JoiningType::NonJoining), (768, 879, JoiningType::Transparent),
  (880, 1154, JoiningType::NonJoining),
  (1155, 1161, JoiningType::Transparent),
  (1162, 1424, JoiningType::NonJoining),
  (1425, 1469, JoiningType::Transparent),
  (1470, 1470, JoiningType::NonJoining),
  (1471, 1471, JoiningType::Transparent),
  (1472, 1472, JoiningType::NonJoining),
  (1473, 1474, JoiningType::Transparent),
  ⋮
  (918000, 1114111, JoiningType::NonJoining),
];

[BurntSushi]

This all LGTM with one minor nit, unless I'm mistaken!

[BurntSushi]

I would be inclined to panic here, no? That is, this error should never be reached. The only way it could be reached, I think, is with a programmer error.

[wezm]

Yes you're right. An invalid value needs to make it though canonicalization successfully too before hitting this branch. I've changed it.

[BurntSushi]

One more nit!

[BurntSushi]

It looks like there is an unused import here?

Also, I would just squash this back into the original commit. :-)

[wezm]

Ahh dang it, sorry about that. Fixed now.

[BurntSushi]

All good, thanks!

[BurntSushi]

This PR is on crates.io in ucd-generate 0.2.6.