stringifier is encoding inline <script> tags

[jonathanong]

var parse5 = require('parse5')
var serializer = new parse5.Serializer
serializer.serialize({
  "nodeName": "#document-fragment",
  "quirksMode": false,
  "childNodes": [
    {
      "nodeName": "script",
      "tagName": "script",
      "attrs": [],
      "namespaceURI": "http://www.w3.org/1999/xhtml",
      "childNodes": [
        {
          "nodeName": "#text",
          "value": "function test(t,n){return t&&n}"
        }
      ]
    }
  ]
})

returns:

<script>function test(t,n){return t&amp;&amp;n}</script>

how do i avoid encoding the contents of the script tag?

[jonathanong]

i realized that the text itself probably shouldn't even be a child node but instead the value or the script node or something, but that's how its getting parsed. i'm just confused right now haha

[inikulin]

@jonathanong node should have parentNode field to be serialized correctly. In your case text node's parentNode should reference to it's parent <script> element.

[jonathanong]

weird issue and solution, but thanks for the quick response!

[inikulin]

@jonathanong you are welcome

weird issue and solution

Not actually, serialization algorithm relies on things like context element namespace, tag name and so on. So, serializer expects that at least minimal tree format is preserved. However, in this particular case serializer can avoid call to parentNode property and I think I'll rewrite it for the next release. Just keep in mind that parse5 expects that it always works with the well formed tree.