On 03/02/15 15:42, Reto Gmür wrote:
Should BNode be shareable across Graphs? The Abstract Syntax says that they
can be shared across the graphs of the same dataset,
Yes - they can be shared.
The note about shared across the graphs of the same dataset is to
highlight an important point. It is not limiting though.
Other important cases: subgraph, composite graphs (union, intersection etc).
It follows from the definitions but because it's by the lack of any text
making restrictions, the dataset is explicitly picked up. The
definitions build up to graphs from below ... rdf term -> triples ->
graphs -> datasets. There is no back linkage, no context, in the data model.
(data model = abstract syntax but "abstract syntax" leads to a bit of
confusion, where as people seem more comfortable with "data model"
because there are real syntaxes (Turtle, et al).
What we don't need:
- A mean for application to re-create identical BNodes (implementations may
however do so), if we have a pointer to the BNode that's fine, otherwise we
get existing BNode by accessing the triples in the graph.
A graph is not limited to one JVM (in time or space).
So an implementation may need to create a specific bNode (a database
would do this, so would transfer
It depends a bit on the "A" in API -- whether "application" means "user
application" or includes system-machinery things like parsers but the
boundary is nit easy to fix.
It might include algorithms spread over many machines, they need to
transfer abstract syntax around.
Andy
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
