Factor Query enrichment chunking into a single helper

[ubyndr]

• Added _batched_enrichment_results so all enrichment-style methods (minimal_slim_enrichment, full_slim_enrichment, contextual_slim_enrichment, ancestor_enrichment, and the graph mirroring) share the same chunk/query loop. Each caller now just provides a lambda to build its SPARQL string.
• This removes duplicated code, keeps chunk sizes consistent, and gives us a single place to adjust batching behavior later.