!!!Risten2 meeting

Filtering is working now.
Next issues:

* when you filter, the article isn't updated (__done__)
* dictionary list isn't filtered against the hit list - do this all in the
  browser, even though it means that the dictionary hit list might ''grow'' in
  certain cases (when a dictionary gets its first hit within the first 50 that
  previously was below the first 50)
* i18n
** the id elements containing dictionary names must be renamed (__done__)
* the search field should be automatically focused when opening the page
* the hit list is strictly ascii/unicode value sorted
** we want to use the sorted index feature in eXist
* the hit list should be uniqued
** search for ''ko'', to the bottom you get twice ''kort'', it should be one.
   Be aware that the uniqued hits still need to have references to both/all
   dictionares where the hit is found, for filtering to work properly.
* add a second tab for dictionary/term collection metadata (authours, sources, etc)
* add a third tab with Risten2 metadata (technologies used, link to REST API reference, maintainers, etc)
* define a dictionary xml structure used by Risten2 - this structure should be
  fixed (to avoid a lot of messy JS code for the article display)
* set up conversion of existing dictionaries and term collections to this format
* start to enhance the article display

Regarding i18n:
Change the following
{{{<id xml:lang="nob">Mekanikktermer, 1999</id>}}}
to
{{{<name xml:lang="nob">Mekanikktermer, 1999</name>}}}