HTMLStringToDOM.js

/*
 * HTMLParser for Greasemonkey by snaka(http://d.hatena.ne.jp/snaka72/)
 *
 * Original coode written by swdyh(http://d.hatena.ne.jp/swdyh/)
 * This code includes part of AutoPagerize(http://userscripts.org/scripts/show/8551)
 *
 * Released under the GPL license
 * http://www.gnu.org/copyleft/gpl.html
 */

var HTMLStringToDOM = (function() {

  function createHTMLDocumentByString(str) {
      if (document.documentElement.nodeName != 'HTML') {
          return new DOMParser().parseFromString(str, 'application/xhtml+xml')
      }
      var html = strip_html_tag(str)
      var htmlDoc
      try {
          // We have to handle exceptions since Opera 9.6 throws
          // a NOT_SUPPORTED_ERR exception for |document.cloneNode(false)|
          // against the DOM 3 Core spec.
          htmlDoc = document.cloneNode(false)
          htmlDoc.appendChild(htmlDoc.importNode(document.documentElement, false))
      }
      catch(e) {
          htmlDoc = document.implementation.createDocument(null, 'html', null)
      }
      var fragment = createDocumentFragmentByString(html)
      try {
          fragment = htmlDoc.adoptNode(fragment)
      }
      catch(e) {
          fragment = htmlDoc.importNode(fragment, true)
      }
      htmlDoc.documentElement.appendChild(fragment)
      return htmlDoc
  }

  function createDocumentFragmentByString(str) {
      var range = document.createRange()
      range.setStartAfter(document.body)
      return range.createContextualFragment(str)
  }

  function strip_html_tag(str) {
      var chunks = str.split(/(<html(?:[ \t\r\n][^>]*)?>)/)
      if (chunks.length >= 3) {
          chunks.splice(0, 2)
      }
      str = chunks.join('')
      chunks = str.split(/(<\/html[ \t\r\n]*>)/)
      if (chunks.length >= 3) {
          chunks.splice(chunks.length - 2)
      }
      return chunks.join('')
  }

  return createHTMLDocumentByString;
})();