All the data and structures are maintained in only one file, which I uploaded on http://parendili.org/doc/ithildin.db. Note that you will need some sort of SQLite interface in order to access it: you can use the command line tool supplied by sqlite.org or another one (there's a few out there, some also free). I found that the firefox plugin is very handy: it provides a graphical user interface so that you can browse through the data.
(if you just want the sql to re-create all tables, I uploaded that here: http://parendili.org/doc/ithildin.db.dump )
In order to demonstrate how the model works, here's a query that retrieves the entries, plus the metadata that I entered (it's only very basic, because I added these by hand):
select l.name, e1.gloss, e2.gloss, dc.name, md.text from entry e1
join translation t
on e1.id = t.sind_entry_id
join entry e2
on e2.id = t.mod_entry_id
join tr_md trmd
on trmd.translation_id = t.id
join metadata md
on md.metadata_id = trmd.metadata_id
join dataclass dc
on dc.dataclass_id = md.dataclass_id
join language l
on l.lang_id = md.language_id
and l.lang_id = e2.language_id
and l.lang_id = dc.language_id
and l.lang_id = l.desc_lang_id
this gives the following result:
if you would restrict the query to one language only, say "German", you can add the clause " where l.name = 'Deutsch' ":
or if you want to only retrieve the short names for the metadata, you can replace this part
select (...) dc.name, md.text from (...)
with this
select (...) dc.short_name, md.short_text from (...)
Just to make sure .. it's not necessary to deal with any of this technical stuff in order to use it. It's just to illustrate how it works.
