Implement UAX#44-LM2

[emilyyyylime]

Closes #44.

At long last I got around to implementing this. The code changes are a little all over the place, but all the tests pass and I've even added a new one to check the behaviour wrt loose name matching.

I've opted to implement this by following the UAX's example procedure:

An implementation of this loose matching rule can obtain the correct results when comparing two strings by doing the following three operations, in order:

1. remove all medial hyphens (except the medial hyphen in the name for U+1180)
2. remove all whitespace and underscore characters
3. apply toLowercase() to both strings
   After applying these three operations, if the two strings compare binary equal, then they are considered to match.

and generating the name->codepoint PHF with the mapped ("normalised") names as keys.

This isn't a breaking change; in particular, any name that mapped to a character before will continue mapping to that same character.
However, it still might be worth it to mention this behaviour in the documentation for the crate or the character function.

[progval]

Thanks.

I'm not familiar with any of this, so this will take me a while to review. Other reviews are welcome

[emilyyyylime]

I don't really know what to do about the build on non-nightly test it appears one of the build.rs dependencies is using an unstable feature? Might need to specify the crate version more precisely

[emilyyyylime]

Apparently getopts depends on Unicode-width? #42 should fix this in that case

[emilyyyylime]

wait nevermind this version doesn't even work

[emilyyyylime]

Okay I believe now this is ready for approval. Merging #42 should make CI pass too

[emilyyyylime]

I've also realised we don't really need MAX_NAME_LENGTH (before normalisation) anymore so I changed some code related to it

[emilyyyylime]

Have you had time to look at the new version? Don't wanna pressure you, just making sure you didn't forget

[progval]

Sorry, I did forget.

One last thing: could you do this?

However, it still might be worth it to mention this behaviour in the documentation for the crate or the character function.

[emilyyyylime]

I went a little extra polishing out some other docs around the crate ^^

[progval]

Published in v2.0.0

I ended up ruling this a breaking change, due to the code generated by unicode_names2_generator after this change wouldn't build with existing versions of unicode_names2, and vice versa