Initial travel knowledge graph release

scripts/cleanup_huaxi_demo_duplicates.py

@@ -0,0 +1,80 @@
+#!/usr/bin/env python3
+"""Remove Huaxi demo duplicate nodes after their knowledge is on canonical POIs."""
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+ROOT = Path("/Users/xuexue/new2")
+if str(ROOT) not in sys.path:
+    sys.path.insert(0, str(ROOT))
+
+from falkordb import FalkorDB  # noqa: E402
+
+from app.config import settings  # noqa: E402
+
+HUAXI = "amap:B035300A51"
+TIANHETAN = "amap:B035300A2B"
+QINGYAN = "amap:B035300ESE"
+
+
+def main() -> None:
+    g = FalkorDB(host=settings.falkordb_host, port=settings.falkordb_port).select_graph(
+        settings.falkordb_graph
+    )
+
+    # Keep one canonical Huaxi Park: the AMap sight POI with coordinates.
+    # The kg_schema_v1 facts were already mirrored to this node.
+    g.query("MATCH (n {id:'ent_huaxi_park'}) DETACH DELETE n")
+
+    # Redirect nearby attractions to existing AMap scenic POIs, then remove temp nodes.
+    g.query(
+        """
+        MATCH (h:Place {element_id:$huaxi})
+        MATCH (t:Place {element_id:$tianhetan})
+        MERGE (h)-[r:NEARBY_ATTRACTION]->(t)
+        SET r.confidence=0.82,
+            r.evidence_quote='逆流而行8公里即可抵达天河潭',
+            r.source='kg_schema_v1+entity_alignment',
+            r.alignment_note='target aligned from ent_tianhetan to existing AMap 天河潭旅游度假区'
+        """,
+        {"huaxi": HUAXI, "tianhetan": TIANHETAN},
+    )
+    g.query(
+        """
+        MATCH (h:Place {element_id:$huaxi})
+        MATCH (q:Place {element_id:$qingyan})
+        MERGE (h)-[r:NEARBY_ATTRACTION]->(q)
+        SET r.confidence=0.84,
+            r.evidence_quote='花溪以南12公里处的青岩镇',
+            r.source='kg_schema_v1+entity_alignment',
+            r.alignment_note='target aligned from ent_qingyan to existing AMap 青岩古镇'
+        """,
+        {"huaxi": HUAXI, "qingyan": QINGYAN},
+    )
+    g.query("MATCH (n {id:'ent_tianhetan'}) DETACH DELETE n")
+    g.query("MATCH (n {id:'ent_qingyan'}) DETACH DELETE n")
+
+    # Remove confusing extraction id from canonical display properties.
+    g.query(
+        """
+        MATCH (p:Place {element_id:$huaxi})
+        SET p.source_extraction='kg_schema_v1'
+        REMOVE p.kg_id
+        """,
+        {"huaxi": HUAXI},
+    )
+
+    res = g.query(
+        """
+        MATCH (n)
+        WHERE n.id IN ['ent_huaxi_park','ent_tianhetan','ent_qingyan']
+        RETURN count(n)
+        """
+    )
+    remaining = res.result_set[0][0] if res.result_set else None
+    print({"removed_demo_duplicates": True, "remaining_target_temp_nodes": remaining})
+
+
+if __name__ == "__main__":
+    main()