Content Variants
Content variants are a tool to distinguish the databus:Part(s) of a databus:Version. Parts of a Dataset (i.e. Version) may describe files of different formats or compression types. Sometimes however, they differ in other aspects, e.g. the language or a data specific subtopic. These special distinguations can be reflected with content variants to allow a more meaningful selection of files.
In fact, features such as the faceted browsing interface or the Databus Collections rely on a proper setup of content variants.
The main rule for content variant setup is the following:
All databus:Part(s) of a databus:Version have to be distinguishable by either format, compression type or at least one content variant
This ensures that each file in the databus:Version can be selected individually by querying for its unique tuple of format, compression type and content variants.
A content variant is a key-value pair with the key being a sub-property of databus:contentVariant
and the value being a (preferrably short) string that can be chosen freely. Content variants could describe either a property of the file or its content.
Examples:
The above example shows two databus:Part(s) of a databus:Version. The two parts are distinguishable by format (nt
and ttl
). Hence, no content variant is required.
The above example shows two databus:Parts of a databus:Version. Both parts have a format of ttl
and a compression type of none
. In order to improve the distinguishability of the two parts, an additional content variant has to be used. The publisher of the databus:Version chose the property dcv:size
as the content variant dimension and assigned each part a different value (small
and big
).
Last updated