So far I have decomposed a web page into 3 sections (in order of assembly):
1) meta section (title, keywords, doctype, language, etc)
2) view section (layout template - everything inside BODY tag which defines layout)
3) page section (Content supplied by user, etc)
1. META HTML is constructed from fields (CSV, DB, etc)
2. View is already in native HTML but wrapped by META
3. Page content is interpolated into the finished compiled product (Native HTML).
Depending on page content, certain CSS and JS files may be required...and so by generating the META section on the fly will allow my CMS to reconstruct META section accordingly (ignoring the fact that TITLE is both meta and markup)
Originally I considered have the Native HTML content of View and Page stored as DOM objects, so depending on the META doctypes I could dynamically generate HTML, XHTML (trans, strict, etc) accordingly...however I have decided that would be likely overkill and too time consuming to both implement and generate (even with caching).
If someone desires to have an HTML and XHTML 2.0 web site, they can define two different view templates...
1) HTML compliant
2) XHTML compliant
I can't think of any reason why anyone would want to work backwards though (backwards support I suppose) and support HTML if they already had xhtml 1.1 trans...but anycase, this isn't my question...
In fact...I don't really have a question...I just needed a brain dump and figured I'd troll for ideas, suggestions, comments, etc...
What do you think of the above? Have any comments?
Cheers