More unicode support for character parsing: `anyUniCharOf` and `uniCharCodeWhere`

[Pamplemousse]

anyCharOf and charCodeWhere aren't adequate when parsing strings containing unicode characters:

import { anyCharOf, charCodeWhere, string } from "parjs";
import { qthen } from "parjs/combinators";

const s = "🥔";

const p1 = anyCharOf(s);
const p2 = charCodeWhere((x) => x === s.codePointAt(0) || { reason: "Bar" });

console.log([
  p1.parse(s),
  p2.parse(s),
]);

give the following result

[
  ParjsFailure {
    trace: {
      userState: ParserUserState {},
      position: 1,
      reason: 'parsers did not consume all input',
      input: '🥔',
      location: [Getter],
      stackTrace: [],
      kind: 'Soft'
    }
  },
  ParjsFailure {
    trace: {
      userState: ParserUserState {},
      position: 0,
      reason: 'Bar',
      input: '🥔',
      location: [Getter],
      stackTrace: [Array],
      kind: 'Soft'
    }
  }
]

Not unexpected, but I would like to be able to do something like this.

[GregRos]

This happens to run into an inherent limitation of Parjs and JavaScript's way of representing characters. But I definitely think it should be supported. It will probably wait for a major version release.

I've wrote up a proposal here: #100