view release on metacpan or  search on metacpan

files/syntax_highlight/prettify.js  view on Meta::CPAN

    * successive chunks of text, and will maintain its own internal state to
    * keep track of how tabs are expanded.
    * @return {function (string) : string} a function that takes
    *   plain text and return the text with tabs expanded.
    * @private
    */
  function makeTabExpander(tabWidth) {
    var SPACES = '                ';
    var charInLine = 0;

    return function (plainText) {
      // walk over each character looking for tabs and newlines.
      // On tabs, expand them.  On newlines, reset charInLine.
      // Otherwise increment charInLine
      var out = null;
      var pos = 0;
      for (var i = 0, n = plainText.length; i < n; ++i) {
        var ch = plainText.charAt(i);

        switch (ch) {
          case '\t':
            if (!out) { out = []; }
            out.push(plainText.substring(pos, i));
            // calculate how much space we need in front of this part
            // nSpaces is the amount of padding -- the number of spaces needed
            // to move us to the next column, where columns occur at factors of
            // tabWidth.
            var nSpaces = tabWidth - (charInLine % tabWidth);
            charInLine += nSpaces;
            for (; nSpaces >= 0; nSpaces -= SPACES.length) {
              out.push(SPACES.substring(0, nSpaces));
            }
            pos = i + 1;
            break;
          case '\n':
            charInLine = 0;
            break;
          default:
            ++charInLine;
        }
      }
      if (!out) { return plainText; }
      out.push(plainText.substring(pos));
      return out.join('');
    };
  }

  var pr_chunkPattern = new RegExp(
      '[^<]+'  // A run of characters other than '<'
      + '|<\!--[\\s\\S]*?--\>'  // an HTML comment
      + '|<!\\[CDATA\\[[\\s\\S]*?\\]\\]>'  // a CDATA section
      // a probable tag that should not be highlighted
      + '|<\/?[a-zA-Z](?:[^>\"\']|\'[^\']*\'|\"[^\"]*\")*>'
      + '|<',  // A '<' that does not begin a larger chunk
      'g');
  var pr_commentPrefix = /^<\!--/;
  var pr_cdataPrefix = /^<!\[CDATA\[/;
  var pr_brPrefix = /^<br\b/i;
  var pr_tagNameRe = /^<(\/?)([a-zA-Z][a-zA-Z0-9]*)/;

  /** split markup into chunks of html tags (style null) and
    * plain text (style {@link #PR_PLAIN}), converting tags which are
    * significant for tokenization (<br>) into their textual equivalent.
    *
    * @param {string} s html where whitespace is considered significant.
    * @return {Object} source code and extracted tags.
    * @private
    */
  function extractTags(s) {
    // since the pattern has the 'g' modifier and defines no capturing groups,
    // this will return a list of all chunks which we then classify and wrap as
    // PR_Tokens
    var matches = s.match(pr_chunkPattern);
    var sourceBuf = [];
    var sourceBufLen = 0;
    var extractedTags = [];
    if (matches) {
      for (var i = 0, n = matches.length; i < n; ++i) {
        var match = matches[i];
        if (match.length > 1 && match.charAt(0) === '<') {
          if (pr_commentPrefix.test(match)) { continue; }
          if (pr_cdataPrefix.test(match)) {
            // strip CDATA prefix and suffix.  Don't unescape since it's CDATA
            sourceBuf.push(match.substring(9, match.length - 3));
            sourceBufLen += match.length - 12;
          } else if (pr_brPrefix.test(match)) {
            // <br> tags are lexically significant so convert them to text.
            // This is undone later.
            sourceBuf.push('\n');
            ++sourceBufLen;
          } else {
            if (match.indexOf(PR_NOCODE) >= 0 && isNoCodeTag(match)) {
              // A <span class="nocode"> will start a section that should be
              // ignored.  Continue walking the list until we see a matching end
              // tag.
              var name = match.match(pr_tagNameRe)[2];
              var depth = 1;
              var j;
              end_tag_loop:
              for (j = i + 1; j < n; ++j) {
                var name2 = matches[j].match(pr_tagNameRe);
                if (name2 && name2[2] === name) {
                  if (name2[1] === '/') {
                    if (--depth === 0) { break end_tag_loop; }
                  } else {
                    ++depth;
                  }
                }
              }
              if (j < n) {
                extractedTags.push(
                    sourceBufLen, matches.slice(i, j + 1).join(''));
                i = j;
              } else {  // Ignore unclosed sections.
                extractedTags.push(sourceBufLen, match);
              }
            } else {
              extractedTags.push(sourceBufLen, match);
            }
          }
        } else {