Add describe endpoint features

[MathiasVDA]

A couple of default queries such as listing a count of classes and predicates, .. to better understand an endpoint and help to write queries

Should open in a smart side panel or overlay. Something that can be made sticky but moved out of the way while query writing.

Results should be remembered even when new queries are executed.

[MathiasVDA]

1. Endpoint & Dataset Overview

• List all named graphs and their sizes to see how data is partitioned.
• Count total triples and triples per graph for a quick sense of scale.
• Find the most frequent namespaces to understand which vocabularies are in use.
• Check for custom functions or extensions exposed by the engine.

2. Vocabulary & Schema Discovery

• List all classes with instance counts to see the main entity types.
• List all properties with usage counts to learn the data’s “shape.”
• Extract SKOS concept schemes and counts of concepts per scheme.
• Identify shapes/constraints (SHACL/Shex) if present in the endpoint.

3. Instance-Level Exploration

• Sample instances per major class (top N per class) to inspect real records.
• List the most connected nodes (highest degree) to find hubs.

4. Labels, Languages & Literals

• Count labels/titles by language to see localization coverage.
• List properties that carry language-tagged literals (and their language distribution).

5. Interlinking & External Links

• List external namespaces referenced (e.g., Wikidata, GeoNames, DBpedia).
• Find entities with outbound links

6. Time & Geo Dimensions

• List temporal properties (creation dates, event times, validity intervals).
• List geospatial properties (WKT, lat/long) and CRS usage.
• Get a bounding box / centroid of all geospatial entities.

[VladimirAlexiev]

There is

• SPARQL Service Description for describing features of a repo.
  Returned when no ?query= is passed
• VOID (and VOID-ext) for describing datasets and various partitions.
  Attached at /.well-known/void
• both are RDF, and the SD can point to the VOID

If YASGUI will interrogate repos for such info (which is expensive), it may as well make SD and VOID and cache them.
Abs even suggest to the user to pass to sever owners for inclusion!

List all named graphs

I've RDFized Crunchbase to 14M graphs: one per csv row, to facilitate Updates.
So not "all" please :-)

[MathiasVDA]

@VladimirAlexiev

So not "all" please :-)

You should build triple stores that can handle these queries :)

Thanks for the tips about SD and VOID. I'll read up on those before I start this task.

[VladimirAlexiev]

Oh, GDB can list 14M graphs. With pagination.
Btw, GDB does not support SD.