Last modified: 2013-03-11 02:37:58 UTC
It should be possible to programmatically extract languages from a Wiktionary entry. For example, I want to be able to programmatically extract "English, Dutch, French, Italian, Latin, Norwegian, Romanian, Spanish, Swedish, Walloon" from <https://en.wiktionary.org/wiki/taxi>.
This will require a PHP implementation, which can then be added to MediaWiki's API. The PHP implementation will also resolve bug 41441.
I talked to Daniel of Wikidata last week and this is something that will neither easily nor soon be doable by using Wikidata either...