Analysing XML file with embedded HTML content

Hello,

We are analysing a large XML file with embedded content and some text which doesn't need to be translated still displays in the editor.

To set up the project, I created a custom XML file type, decided which elements should be translatable or not at the parsing stage. I have enabled 'Embedded content (Legacy)' and selected sdl:cdata in the document structure information box and created tag definition rules.

However, I have encountered a couple of issues.

First of all, text which I defined as not being translatable at the parsing stage is appearing in the editor. Secondly, embedded text appears in the editor, such as <![CDATA[<html>  and text relating to images etc - do I need to set up a tag definition rule for each of these, or is there a catch all rule I can use?

Also, should I be using embedded content (legacy) or embedded content processor? 

Thanks for your help!

Jane