So, for you fellow coders out there, here is how MemeMiner works

Two calls are made to the wikipedia API. The goal from the start was to use Ajax-like technology, that would keep the user always on MemeMiner page, without the need for refreshing. JSON was the sensible technology to use for the API calls, and XmlHttpRequest (aka XHR, aka Ajax) for the communication with the php server side methods. JSON consists of adding a little script to the head of the page, that is a Javascript call to Wikipedia's API. The parameters on the first call allow for the retrieval of all the links of the different languages available on the requested term. So, for a certain language there will be an entry, and for that entry links for the equivalent article in other languages. This of course, is the core of MemeMiner.

After retrieving the JSON response from Wikipedia, the need arises to go through it and fetch the relevant info. In this case, the list of language links. If, from the entries on this list, the destination language is found, it means an equivalent expression was found.

By using this equivalent expression, the Wikipedia article that corresponds to this entry - in the destination language - is requested to the API. This is the second API request, again using JSON. It's not the whole article that is fetched, but the description, composed, amongst other things, of the first Wikipedia 'paragraph' that is a synthesis of the entry.

The last step is to clean the retrieved text. This is needed due to the fact that the response is encoded in Wikitext (thus using those fancy markup tags). Mediawiki2html, a handy php regex utilty, was used. My thanks goes to the developer, Johannes Buchner. The call to the php is done Ajax way. Take a look at the code!

There you go, the real peek behing the curtains. Use, share and tweak the code to your hearts content. But remember to link back to me, and keep the ShareAlike, Copyleft license alive!