[rdf4j-dev] expensive size() calls for fedx/sparql proxy

[Matthew Nguyen <nguyenm9@xxxxxxx>]

Hey folks,

Playing around with FedX and Sparql Proxy.  When you activate one of the repos, it tries and fetches the size which seems to fetch all the records to calc the count.  This seems like an overly expensive operation.  This is somewhat compounded b/c my sparql endpoints are hosted vi OnTop (which virtualizes a SQL table to 3store).  Is the rationale that it will timeout on the workbench side and thus drop the connection and thus the remote sparql endpoint should cancel its query?  

Both fedx & sparql use this approach.

https://github.com/eclipse/rdf4j/blob/main/core/repository/sparql/src/main/java/org/eclipse/rdf4j/repository/sparql/SPARQLConnection.java#L290
https://github.com/eclipse/rdf4j/blob/main/tools/federation/src/main/java/org/eclipse/rdf4j/federated/FedXConnection.java#L372

matt