Hi Martin,
Thanks François - no problem. Do you think it can be achieved with a hook? I might try to code s/th like that.
In fact the "results" plugin (pi2) was always meant to be absolutely basic and fulfill one precise goal: display a list of pages (it works only for pages) that matches exactly the list where a given keyword was found when building up the tag cloud. Honestly I didn't find this feature very exciting, but it was wished for by the original client. I always thought I would rather plug tag cloud results into a search engine or something.
One thing to remember is that the links in the tag cloud are fully customizable, so it's perfectly possible to have a link that uses only the keyword and not the page numbers. So you can change the link easily with TypoScript in order not to display the page numbers.
Then you are faced with having to display a content element which receives a keyword and should display a list of pages that make sense (ideally the same that the keyword was found on when building the tag cloud). The idea that was floated at the time was to add a cache table, where the pages for each keyword would be stored. This would work but introduce an additional level of complexity, since you have to take care of all the usual mechanisms that affect cache, like logged in users (if access-restricted content was used in the tag cloud), language, etc. Furthermore if you have several tag clouds, you would have to know which one is pointing to which page in order to retrieve the right cache. One possibility would then be for the pi2 to make a reference to the pi1, so that you know which tag cloud is at the original of the link. And of course, cache must be cleared at the right point.
The other possibility is to recalculate the tag cloud and get the pages for the given work. This costs extra performance but does not require to handle cache. It would also not require any change in vge_tagcloud, since this is something you could package in your own extension and just call the vge_tagcloud API.
Another possibility would be to use extensions that can display generic lists of stuff (like the Tesseract framework I developed, or pt_list or browser, which I don't know), which could take the keyword as input and which should allow you to recreate the query that was used as the base of the tag cloud and thus retrieve the right list of pages (albeit that would depend on the complexity of your original tag cloud creation).
This should give you a few directions you might want to follow. Keep me posted about which solution you eventually adopt.
Cheers -- Francois Suter Cobweb Development Sarl - http://www.cobweb.ch _______________________________________________ TYPO3-english mailing list [email protected] http://lists.typo3.org/cgi-bin/mailman/listinfo/typo3-english
