import {values} from 'micromark-util-symbol'

/**
 * Normalize an identifier (as found in references, definitions).
 *
 * Collapses markdown whitespace, trim, and then lower- and uppercase.
 *
 * Some characters are considered “uppercase”, such as U+03F4 (`ϴ`), but if their
 * lowercase counterpart (U+03B8 (`θ`)) is uppercased will result in a different
 * uppercase character (U+0398 (`Θ`)).
 * So, to get a canonical form, we perform both lower- and uppercase.
 *
 * Using uppercase last makes sure keys will never interact with default
 * prototypal values (such as `constructor`): nothing in the prototype of
 * `Object` is uppercase.
 *
 * @param {string} value
 *   Identifier to normalize.
 * @returns {string}
 *   Normalized identifier.
 */
export function normalizeIdentifier(value) {
  return (
    value
      // Collapse markdown whitespace.
      .replace(/[\t\n\r ]+/g, values.space)
      // Trim.
      .replace(/^ | $/g, '')
      // Some characters are considered “uppercase”, but if their lowercase
      // counterpart is uppercased will result in a different uppercase
      // character.
      // Hence, to get that form, we perform both lower- and uppercase.
      // Upper case makes sure keys will not interact with default prototypal
      // methods: no method is uppercase.
      .toLowerCase()
      .toUpperCase()
  )
}