feat: allow identifiers to contain utf-8 characters

[calebdw]

Checklist

• All tests pass in CI
• There are enough tests for the new fix/feature
• Grammar rules have not been renamed unless absolutely necessary (0 rules renamed)
• The conflicts section hasn't grown too much (0 new conflicts)
• The parser size hasn't grown too much (master: 2727, PR: 2727)
  (check the value of STATE_COUNT in src/parser.c)

See #142 (comment) for more context: essentially the true definition of a legal identifier is consistent with what is currently listed on the docs: ^[a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*$

However, from my (limited) understanding there is a bug of sorts when PHP is parsing multibyte characters and instead compares the individual bytes which in practice allows characters outside the given regex to be parsed as identifiers.

This PR updates the grammar and scanner to allow the full utf-8 range as valid identifiers (\u0080-\uffff). PHP may even parse characters higher than this range (I haven't tested it), but given that utf-8 is already technically illegal I don't see the need to try and support anything higher. Although it would be real easy to update the scanner to use the full uint32 returned by the lexer.

Closes #142, closes #171

[amaanq]

LGTM