html parser doesn't handle cdata

[ForbesLindesay]

<![CDATA[
This should be CDATA...
]]>

results in

[ { data: '[CDATA[\nThis should be CDATA...\n]]',
    type: 'comment' } ]

when it should result in

[ { data: '\nThis should be CDATA...\n',
    type: 'cdata' } ]

using version 3.1.5

[ForbesLindesay]

Even more seriously:

<script><![CDATA[
</script>
]]></script>

results in

[ { type: 'script',
    name: 'script',
    attribs: {},
    children: [ { data: '<![CDATA[\n', type: 'text' } ],
    prev: null,
    next: null,
    parent: null },
  { data: '\n]]>', type: 'text' } ]

when it should result in

[ { type: 'script',
    name: 'script',
    attribs: {},
    children: [ { data: '\n</script>\n', type: 'cdata' } ],
    prev: null,
    next: null,
    parent: null } ]

[fb55]

This behavior was changed in order of fixing jsdom/jsdom#618 and #32.

At least the behavior for special tags could arguably be changed, your example points out nicely what issues may arise.

The HTML behavior is in line with the spec; if you want CDATA, enable the XML mode.

[georgephillips]

Here is a pretty common example which breaks

<script>window.jQuery || document.write('<script src="js/vendor/jquery-1.10.2.min.js"><\/script>')</script>

[stevenvachon]

CDATA sections are normalized with normalizeWhitespace:true. In my opinion, they should be preserved.

[fb55]

This issue was actually solved in #74.

@stevenvachon If you don't want your whitespace to be normalized, don't set that option.