feat: complete pre-ai local completion v3

central/scripts/prepare_ai_candidates.py

@@ -0,0 +1,31 @@
+"""Prepare versioned candidate evidence from already computed deterministic candidates."""
+import argparse
+import json
+from pathlib import Path
+
+from central.app.ai.retrieval_v3 import prepare_candidate_set
+
+
+def main(input_path, output_path):
+    source = json.loads(Path(input_path).read_text(encoding="utf-8"))
+    prepared = []
+    for item in source.get("items", []):
+        prepared.append(prepare_candidate_set(
+            type("WorkItem", (), {"task_type": item.get("task_type", "INVALID_NCM"), "dedup_key": item["dedup_key"], "payload": item.get("payload", {})})(),
+            [tuple(row) for row in item.get("candidates", [])],
+            dataset_versions=source.get("dataset_versions"),
+            gtin_evidence=item.get("gtin_evidence"),
+            kb_neighbors=item.get("kb_neighbors"),
+            hierarchical_context=item.get("hierarchical_context"),
+        ))
+    report = {"artifact": "AI_CANDIDATE_PREPARATION_V3", "items": prepared, "semantic_decisions": 0}
+    Path(output_path).write_text(json.dumps(report, indent=2, ensure_ascii=False) + "\n", encoding="utf-8")
+    return report
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("input_path")
+    parser.add_argument("output_path")
+    args = parser.parse_args()
+    print(json.dumps(main(args.input_path, args.output_path), indent=2, ensure_ascii=False))