One document, two content parts ?

For the build of a twinning site (French/German cities), I would like have for each documents, a part of the content in French and the other in German.
The layout of the site will be in German or in French, depending of the nationality of the web user, but all the documents will be presented in the two languages on the same page.
How could I add a sort of template to make easier the input of the content (in French and in German) ?

Is it possible to add a second text-area in the document manager for instance ? Which pgms should be modified ?
Is there somewhere an example of this kind of customisation ?

Thanks for your answer and your suggestion.

Isn’t this a perfect use for a template variable?

Yes, probably. But if I use TVs instead of site_content, I will loose the search results details provided by AjaxSearch. Unless I find a solution to improve AS
Nevertheless, thanks for your answer, TVs is probably the better way for this kind of customization.

I see.

IMHO, we need an external search indexing tool. I’ve partially integrated a site with Solr (a Java-based search/indexing tool) and the concept works wonderfully. But it’s anything but trivial to architect a search indexing tool properly, though I had planned to try a simple implementation as part of the efforts for 1.0, which will consequently also change the way site_content works, allowing documents to have predefined sets of content/metadata attached (i.e. multiple content areas, custom metadata fields, etc.) without having the relationships be dependent on a Template.

Lucene might be interesting, as it already has some good PHP interfaces

Quote from: olafmol at Aug 13, 2008, 07:55 AM

Lucene might be interesting, as it already has some good PHP interfaces

Yes, Lucene has various interfaces and even a native PHP server implementation as part of the Zend Framework.  However, the Zend impl is still a bit immature, and Lucene is but a framework for constructing robust search solutions.  Solr is built on Lucene and includes lots of tools for doing search term anaylsis, extraction and indexing from various kinds of source content (including Word docs and PDF’s), and IMHO, holds the best promise for a "reference" advanced search solution integrated with Revolution.

I have a few projects coming up that will need this solution, and hopefully I can come up with a way to turn it into a generic component package as part of the work.