Last modified: 2013-04-04 22:02:47 UTC
Due to the structure of Wiktionary, it's unlikely that the opensearch description (as currently implemented) will ever give a useful result It would be nice to customise this for Wiktionary - particularly we would need the content of the first <ol> (or # list) to be used in place of the first paragraph. It would also be preferably not to delete cross-project links. Relevant links: User trying to get information from opensearch: http://www.mwusers.com/forums/showthread.php?14641-Questions-about-the-MediaWiki-API-Component http://en.wiktionary.org/w/api.php?format=xml&action=opensearch&search=test&limit=1 <Description xml:space="preserve">< < < *, pp. </Description> http://en.wiktionary.org/w/api.php?format=xml&action=opensearch&search=chalcidology&limit=1 <Description xml:space="preserve"># The study of .</Description>
I'd probably recommend two things: 1) Merge the OpenSearch XML summary extraction into core 2) Give it some nice extension hooks to allow overrides Shouldn't be too hard then to add a custom extraction for Wiktionary which does better.
This is basically the same problem as bug 20411, though Wiktionary pages are particularly egregious in not actually including much (or any) inline body text.
Since the core bug is bug 20411, moving this one to site requests.
