forrest-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ross Gardler <rgard...@apache.org>
Subject Re: search engine friendly skins
Date Tue, 21 Dec 2004 00:01:03 GMT
Glen Tulukin wrote:
> 1. Currently document2html does not create *html*, it creates a div which a
> *part* of html. It does not render <html> as the root element.

Yes, but everything within the document conforms to a subset of HTML.

> 2. document2html is not the last point in the processing pipe, it assumes
> that there is the last stage (site2html) that produces the final html and
> thus it can have other elements like meta-data as long as the site2html
> generates valid html.

Yes, but what you may not know is that we are moving from the document 
DTD to XHTML2 in the next release. As a consequence document2html will 
dissappear. Consequently, it is important that the input to site2html 
continues to process html in oder to make this transition easier.

> 3. Having two elements <meta-data> and <div> as result of document2html
> allows to separate meta data (not necessarily html-related data) from a
> <div>. It allows to have additional xslt transformation that can be applied
> to <meta-data> only and not to parse <div>.

Yes, but this is also true if we conform to standard html and use

<html>
   <head>
     <meta ...>
   </head>
   <body>
     <div>
        ...
     </div>
   ...



> Do not think about meta data as merely html/head/title or
> html/meta/keywords -- there might be other data associated with a document.
> Look at advanced jpg properties (date, location, equipment make, camera
> model, author, etc.). It's more than title and keywords. I can put
> camera_model to the <div> as a hidden element and than extract from the div,
> but this will be awkward. To have a separate meta-data element is cleaner.


Yes, I agree and this is why my personal solution would be to have a 
separate meta-data document to enable efficient processing of this 
important data. However, this seems to go beyond your own use case and 
that XHTML meta elemets would be sufficient. I assumed this since you 
start in the document DTD using the the XHTML meta-data element.

If we need a separate record of meta-data in a more flexible format than 
the HTML meta element allows then we should really use a separate 
document. The processing you propose in site2html.xsl will remain the 
same, or at least very similar since we can use te XSL document(...) 
function to import the meta-data document when one exists.

Am I mssing something?

Ross

Mime
View raw message