Rules for the Declass Schema
-
All collections should be held in their own database named
declassification_[collection]
, where[collection]
is the name of the specific document collection. For example, currently we havedeclassification_frus
for the FRUS collection. -
Because all collections are made up of documents, all collection databases should have a table called
docs
which contains fields relevant to that document type. The different fields in variousdocs
tables will vary from collection to collection, because different document collections have different structures. That said, there are some things that are common across document collections:a) A
body
field -- this is what we determine to be the main textual content of a given document b) Anid
field as a unique primary key c) Date information, most of the time, in some form. d) Asubject
,title
, or some other quick descriptive string. Documents can have multiple, for example both asubject
and atitle