!!!Test setup

The code is supposed to run locally on the development machine. The setup should contain at least:

* a fresh check-out/update of the eXist svn code (remember to build!)
* a suitable collection of dictionaries and terminological collections installed



!!!Searching and dictionary/terminology lookups

* non-ASCII chars: search for __bå*__, click on a few select hits to check that entry lookups are working
* language-specific searches:
* dictionary-specific searches:


kjlhlkjh

!!!Editing

The testing tasks re: editing covers at least:

* searching for entries to edit
* update an entry
* delete an entry (not yet implemented)
* add a new entry
* merge two entries (not yet implemented - will it ever be?)


These tasks should be done for all collections.

!!Terminology

We have the following terminology collections to test, with partially separate code and functionality:

* SD-terms
* propnouns
* mekanikk1999


There will be more collections in the future.

!!Dictionaries

We have only one dictionary installed, mainly for testing purposes:

!komi-JR

Additional dictionaries are planned for the future, for real use, and not only testing.

!!Classification schemes

The main task here is to edit the different language labels for each class, and, after the editing session, update the translation tables. It should be done for all existing classification schemes:

* SD-class
* propnounclass (not yet created, but will consist of the semantic classifiction of the proper nouns)
* mek-1999-class


Test tasks:

* edit an existing label
* delete an existing label for a language
* add a new label for a language
* update the translation tables (=export the classification scheme to the Cocoon i18n component)
* run some queries to see that the changes appear in the browser/risten.no interface
* run classification searches, to see that the result is as expected, and that the updated labels show up as they should (NB! Classification searches are presently only available from within the editor!)



!!Metadata

Metadata storage and display isn't implemented yet, but when it is, we need to test changes, additions and removals here as well.

Metadata is supposed to be data about each collection, mainly human readable (like the foreword of a dictionary), but also some intended for machine parsing (like the availabe languages for a given collection, the ID, whether public or private, etc).