Aklakan commented on PR #3511:
URL: https://github.com/apache/jena/pull/3511#issuecomment-4119998395
> Aside from Virtuoso, which is an outlier and we have a approach for that
already, which systems had you in mind here?
We are experimenting with and using various systems across a mix of research
and industrial projects. Besides Fuseki and Virtuoso, there are also setups
with Qlever, GraphDB and Tentris.
But that's actually beside the point - to recap:
* The main reason for the adapter system is as base infrastructure for the
execution tracking of #2701. The idea is to have a wrapper builders that
wrap/enhance `QueryExec` and `UpdateExec` with the tracking.
* `DatasetGraphSparql` is the test bed for whether the infrastructure is
also sufficiently generic to expand to SPARQL queries against a DatasetGraph
that is backed by an `RDFLink`. This idea is not really new: There was the
unfinished `GraphSPARQL` class which is now implemented as `DatasetGraphSparql`.
* Yesterday's change to separate the query generation from
`DatasetGraphSparql` is actually again motivated by Virtuoso/DBpedia, which
does not support `GRAPH ?g { }`. With the recent change, it is possible to
override `DsgSparqlExecutor.listGraphNodes` to use `SELECT DISTINCT ?g { GRAPH
?g { ?s ?p ?o } }`. And my custom override can stay outside of Jena.
> There are 2 "// XXX Unwrap Graph view over a DatasetGraph?"
I think unwrapping here, in the general case, would be wrong.
Some wrappers are query change execution.
I removed those comments. Yes, the unwrapping does not belong directly into
`SparqlAdapter`. It can be handled with a `QueryExecBuilderProvider` that e.g.
interjects `DatasetGraphOne/GraphView/DatasetGraphSparql` (likewise for
updates). But that's for the future.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
