@atlaskit/adf-schema/schema-generators/__tests__/helpers/_utils.ts

Shared package that contains the ADF-schema (json) and ProseMirror node/mark specs

export const filterAndSortMarks = (marks: string, excluded = []) => {
  return marks
    .split(' ')
    .filter((v) => !excluded.includes(v))
    .sort()
    .join(' ');
};

/**
 * filter out the unsupported nodes in content expression
 * @param skipLists list includes node name need to be excluded
 * @returns content without node name defined in skipList
 */
const unSupportedExp =
  (skipLists: string[] = []) =>
  (str: string) => {
    // this is used to treat cases like 'unsupportedBlock*'
    if (str.endsWith('*')) {
      return !skipLists.includes(str.substring(0, str.length - 1));
    }
    if (str.endsWith('+')) {
      return !skipLists.includes(str.substring(0, str.length - 1));
    }
    return !skipLists.includes(str);
  };

/**
 * Sort the nodes in (), and also exclude the node if matched in skipList
 * @param input content expression, e.g. "(paragrah | heading)"
 * @param skipLists a list defined node name can be removed from content expression.
 * @param enableWrap flag to add back ()
 * @returns
 */
const sortAndFilterOrsExp = (
  input: string,
  skipLists = [],
  enableWrap = false,
) => {
  const filteredOr = input.split('|').filter(unSupportedExp(skipLists));
  if (filteredOr.length === 1) {
    return `${filteredOr[0]}`;
  }
  if (enableWrap) {
    return `(${filteredOr.sort().join('|')})`;
  }
  return `${filteredOr.sort().join('|')}`;
};

/**
 * Find the string inside () of provided string
 * @param inputString input string likely have () in it
 * @returns
 */
function findStringInParentheses(inputString: string) {
  const regex = /\((.*?)\)/g;
  const matches = inputString.match(regex);
  if (matches) {
    return matches.map((match) => match.replace('(', '').replace(')', ''));
  } else {
    return null;
  }
}

/**
 * Format the content expression
 * @param exp content expression string, e.g. (aTest, bTest)+
 * @param operatorStart could be {, +, *
 * @param skipLists
 * @returns sorted and removed node defined in skipList
 */
function formatExp(exp: string, operatorStart = '', skipLists = []) {
  const orEnd = exp.indexOf(operatorStart);
  const suffix = exp.substring(orEnd);
  const matches = findStringInParentheses(exp);
  if (matches) {
    const temp = sortAndFilterOrsExp(matches[0], skipLists);
    if (temp.includes('|')) {
      return `(${temp})${suffix}`;
    }
    return `${temp}${suffix}`;
  }
  return sortAndFilterOrsExp(exp, skipLists);
}

const expFormatter =
  (skipLists = []) =>
  (content: string) => {
    if (content.includes('|')) {
      if (content.endsWith('}')) {
        return formatExp(content, '{', skipLists);
      }
      if (content.endsWith('+')) {
        return formatExp(content, '+', skipLists);
      }
      if (content.endsWith('*')) {
        return formatExp(content, '*', skipLists);
      }
      if (!content.includes('+') && !content.includes('*')) {
        if (content.startsWith('(') && content.endsWith(')')) {
          const temp = content.substring(1, content.length - 1);
          return `(${sortAndFilterOrsExp(temp, skipLists)})`;
        }
        return sortAndFilterOrsExp(content, skipLists);
      }
    }
    return sortAndFilterOrsExp(content, skipLists);
  };

/**
 * Sort the node inside of () in contents string,
 * remove spaces around '|', because there were inconsistence in legacy PM schema
 * remove not supported nodes defined in skipList
 * @param content contents string, e.g. (aTest | bTest)+ (cTest | dTest)*
 * @param skipLists a list contains node name can be removed from contents string
 * @returns sorted content string, without node defined in skipList
 */
export const formatContent = (content: string, skipLists = []) => {
  const result = content.replace(/\s*\|\s*/g, '|');
  if (result.includes(' ')) {
    return result.split(' ').map(expFormatter(skipLists)).join(' ').trim();
  }
  return expFormatter(skipLists)(result).trim();
};