lsq rdfize and lsq benchmark use different ID schema

[Aklakan]

The initial idea was to reuse information contained in the static RDFization in the benchmark process.
I.e. instead of trying to parse an ID out from an IRI it should be assembled from attributes.

However, for some reason rdfization and running benchmarking does not yield matching graph IRIs

lsq rdfize:

<http://lsq.aksw.org/q-0ez272gh_Z4IFrwv_dbCEpkmz5ZnL5VaDwmja75SFPE>

lsq benchmark run:

lsqr:lsqQuery-0DgPrLmzT2lANsdflr3XIUpdNZgT09YRsOnbZ9gA7CE

Obviously, also the prefix declarations are not yet consistently used.

It may well be that the only differenec is in q- vs lsqQuery.
In other query logs there is the issue of hex-encoding of checksums vs base64 encoding but this might have been fixed already. Because the latter yields shorter IRIs this one should be used.

[Aklakan]

This issue has been addressed but again a test suite should be added.
All active LSQ code should now use the annotation-based skolemization system; but there are still some leftovers of the utterly complex programmatic approach.