Paperlib MCP

#!/usr/bin/env python3 import sys from pathlib import Path # Add src to path sys.path.insert(0, str(Path(__file__).parent.parent / "src")) from paperlib_mcp.db import query_all def run_diagnostics(): print("--- 1. Top Duplicate Relations (Same Subj, Pred, Obj) ---") sql1 = """ SELECT subj_entity_id, predicate, obj_entity_id, COUNT(*) AS n FROM relations GROUP BY subj_entity_id, predicate, obj_entity_id HAVING COUNT(*) > 1 ORDER BY n DESC LIMIT 20; """ rows1 = query_all(sql1) if not rows1: print("No duplicates found with count > 1.") for row in rows1: print(f"Subj: {row['subj_entity_id']}, Pred: {row['predicate']}, Obj: {row['obj_entity_id']}, Count: {row['n']}") print("\n--- 2. Relation Statistics ---") sql2 = "SELECT COUNT(*) as total FROM relations;" total = query_all(sql2)[0]['total'] sql3 = """ SELECT COUNT(DISTINCT (subj_entity_id, predicate, obj_entity_id)) as unique_keys FROM relations; """ unique = query_all(sql3)[0]['unique_keys'] print(f"Total Relations: {total}") print(f"Unique Relation Keys: {unique}") if total > 0: print(f"Redundancy Ratio: {(total - unique) / total:.2%}") if __name__ == "__main__": run_diagnostics()