compile_keywords.js

// keywords that should have no default relevance value
const COMMON_KEYWORDS = [
  'of',
  'and',
  'for',
  'in',
  'not',
  'or',
  'if',
  'then',
  'parent', // common variable name
  'list', // common variable name
  'value' // common variable name
];

const DEFAULT_KEYWORD_SCOPE = "keyword";

/**
 * Given raw keywords from a language definition, compile them.
 *
 * @param {string | Record<string,string|string[]> | Array<string>} rawKeywords
 * @param {boolean} caseInsensitive
 */
export function compileKeywords(rawKeywords, caseInsensitive, scopeName = DEFAULT_KEYWORD_SCOPE) {
  /** @type {import("highlight.js/private").KeywordDict} */
  const compiledKeywords = Object.create(null);

  // input can be a string of keywords, an array of keywords, or a object with
  // named keys representing scopeName (which can then point to a string or array)
  if (typeof rawKeywords === 'string') {
    compileList(scopeName, rawKeywords.split(" "));
  } else if (Array.isArray(rawKeywords)) {
    compileList(scopeName, rawKeywords);
  } else {
    Object.keys(rawKeywords).forEach(function(scopeName) {
      // collapse all our objects back into the parent object
      Object.assign(
        compiledKeywords,
        compileKeywords(rawKeywords[scopeName], caseInsensitive, scopeName)
      );
    });
  }
  return compiledKeywords;

  // ---

  /**
   * Compiles an individual list of keywords
   *
   * Ex: "for if when while|5"
   *
   * @param {string} scopeName
   * @param {Array<string>} keywordList
   */
  function compileList(scopeName, keywordList) {
    if (caseInsensitive) {
      keywordList = keywordList.map(x => x.toLowerCase());
    }
    keywordList.forEach(function(keyword) {
      const pair = keyword.split('|');
      compiledKeywords[pair[0]] = [scopeName, scoreForKeyword(pair[0], pair[1])];
    });
  }
}

/**
 * Returns the proper score for a given keyword
 *
 * Also takes into account comment keywords, which will be scored 0 UNLESS
 * another score has been manually assigned.
 * @param {string} keyword
 * @param {string} [providedScore]
 */
function scoreForKeyword(keyword, providedScore) {
  // manual scores always win over common keywords
  // so you can force a score of 1 if you really insist
  if (providedScore) {
    return Number(providedScore);
  }

  return commonKeyword(keyword) ? 0 : 1;
}

/**
 * Determines if a given keyword is common or not
 *
 * @param {string} keyword */
function commonKeyword(keyword) {
  return COMMON_KEYWORDS.includes(keyword.toLowerCase());
}