source: node_modules/highlight.js/lib/languages/xml.js

/**
 * @param {string} value
 * @returns {RegExp}
 * */

/**
 * @param {RegExp | string } re
 * @returns {string}
 */
function source(re) {
  if (!re) return null;
  if (typeof re === "string") return re;

  return re.source;
}

/**
 * @param {RegExp | string } re
 * @returns {string}
 */
function lookahead(re) {
  return concat('(?=', re, ')');
}

/**
 * @param {RegExp | string } re
 * @returns {string}
 */
function optional(re) {
  return concat('(', re, ')?');
}

/**
 * @param {...(RegExp | string) } args
 * @returns {string}
 */
function concat(...args) {
  const joined = args.map((x) => source(x)).join("");
  return joined;
}

/**
 * Any of the passed expresssions may match
 *
 * Creates a huge this | this | that | that match
 * @param {(RegExp | string)[] } args
 * @returns {string}
 */
function either(...args) {
  const joined = '(' + args.map((x) => source(x)).join("|") + ")";
  return joined;
}

/*
Language: HTML, XML
Website: https://www.w3.org/XML/
Category: common
Audit: 2020
*/

/** @type LanguageFn */
function xml(hljs) {
  // Element names can contain letters, digits, hyphens, underscores, and periods
  const TAG_NAME_RE = concat(/[A-Z_]/, optional(/[A-Z0-9_.-]*:/), /[A-Z0-9_.-]*/);
  const XML_IDENT_RE = /[A-Za-z0-9._:-]+/;
  const XML_ENTITIES = {
    className: 'symbol',
    begin: /&[a-z]+;|&#[0-9]+;|&#x[a-f0-9]+;/
  };
  const XML_META_KEYWORDS = {
    begin: /\s/,
    contains: [
      {
        className: 'meta-keyword',
        begin: /#?[a-z_][a-z1-9_-]+/,
        illegal: /\n/
      }
    ]
  };
  const XML_META_PAR_KEYWORDS = hljs.inherit(XML_META_KEYWORDS, {
    begin: /\(/,
    end: /\)/
  });
  const APOS_META_STRING_MODE = hljs.inherit(hljs.APOS_STRING_MODE, {
    className: 'meta-string'
  });
  const QUOTE_META_STRING_MODE = hljs.inherit(hljs.QUOTE_STRING_MODE, {
    className: 'meta-string'
  });
  const TAG_INTERNALS = {
    endsWithParent: true,
    illegal: /</,
    relevance: 0,
    contains: [
      {
        className: 'attr',
        begin: XML_IDENT_RE,
        relevance: 0
      },
      {
        begin: /=\s*/,
        relevance: 0,
        contains: [
          {
            className: 'string',
            endsParent: true,
            variants: [
              {
                begin: /"/,
                end: /"/,
                contains: [ XML_ENTITIES ]
              },
              {
                begin: /'/,
                end: /'/,
                contains: [ XML_ENTITIES ]
              },
              {
                begin: /[^\s"'=<>`]+/
              }
            ]
          }
        ]
      }
    ]
  };
  return {
    name: 'HTML, XML',
    aliases: [
      'html',
      'xhtml',
      'rss',
      'atom',
      'xjb',
      'xsd',
      'xsl',
      'plist',
      'wsf',
      'svg'
    ],
    case_insensitive: true,
    contains: [
      {
        className: 'meta',
        begin: /<![a-z]/,
        end: />/,
        relevance: 10,
        contains: [
          XML_META_KEYWORDS,
          QUOTE_META_STRING_MODE,
          APOS_META_STRING_MODE,
          XML_META_PAR_KEYWORDS,
          {
            begin: /\[/,
            end: /\]/,
            contains: [
              {
                className: 'meta',
                begin: /<![a-z]/,
                end: />/,
                contains: [
                  XML_META_KEYWORDS,
                  XML_META_PAR_KEYWORDS,
                  QUOTE_META_STRING_MODE,
                  APOS_META_STRING_MODE
                ]
              }
            ]
          }
        ]
      },
      hljs.COMMENT(
        /<!--/,
        /-->/,
        {
          relevance: 10
        }
      ),
      {
        begin: /<!\[CDATA\[/,
        end: /\]\]>/,
        relevance: 10
      },
      XML_ENTITIES,
      {
        className: 'meta',
        begin: /<\?xml/,
        end: /\?>/,
        relevance: 10
      },
      {
        className: 'tag',
        /*
        The lookahead pattern (?=...) ensures that 'begin' only matches
        '<style' as a single word, followed by a whitespace or an
        ending braket. The '$' is needed for the lexeme to be recognized
        by hljs.subMode() that tests lexemes outside the stream.
        */
        begin: /<style(?=\s|>)/,
        end: />/,
        keywords: {
          name: 'style'
        },
        contains: [ TAG_INTERNALS ],
        starts: {
          end: /<\/style>/,
          returnEnd: true,
          subLanguage: [
            'css',
            'xml'
          ]
        }
      },
      {
        className: 'tag',
        // See the comment in the <style tag about the lookahead pattern
        begin: /<script(?=\s|>)/,
        end: />/,
        keywords: {
          name: 'script'
        },
        contains: [ TAG_INTERNALS ],
        starts: {
          end: /<\/script>/,
          returnEnd: true,
          subLanguage: [
            'javascript',
            'handlebars',
            'xml'
          ]
        }
      },
      // we need this for now for jSX
      {
        className: 'tag',
        begin: /<>|<\/>/
      },
      // open tag
      {
        className: 'tag',
        begin: concat(
          /</,
          lookahead(concat(
            TAG_NAME_RE,
            // <tag/>
            // <tag>
            // <tag ...
            either(/\/>/, />/, /\s/)
          ))
        ),
        end: /\/?>/,
        contains: [
          {
            className: 'name',
            begin: TAG_NAME_RE,
            relevance: 0,
            starts: TAG_INTERNALS
          }
        ]
      },
      // close tag
      {
        className: 'tag',
        begin: concat(
          /<\//,
          lookahead(concat(
            TAG_NAME_RE, />/
          ))
        ),
        contains: [
          {
            className: 'name',
            begin: TAG_NAME_RE,
            relevance: 0
          },
          {
            begin: />/,
            relevance: 0,
            endsParent: true
          }
        ]
      }
    ]
  };
}

module.exports = xml;