Initial SQL-only 1C adapter baseline

scripts/search_1c_rag.py

@@ -0,0 +1,65 @@
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+
+from common import read_json, search_lexical_index
+from rag_profiles import resolve_rag_profile
+
+ROOT = Path(__file__).resolve().parents[1]
+DEFAULT_INDEX = ROOT / "plugins" / "1c" / "datasets" / "prepared" / "rag_index.json"
+
+
+def print_result(result: dict, index: int) -> None:
+    document = result["document"]
+    content = (document.get("content") or "").strip().replace("\n", " ")
+    if len(content) > 500:
+        content = content[:497].rstrip() + "..."
+
+    print(f"{index}. score={result['score']:.4f}")
+    print(f"   source={document.get('source_path')} chunk={document.get('chunk_index')}")
+    print(f"   title={document.get('title')}")
+    if (document.get("metadata") or {}).get("heading"):
+        print(f"   heading={(document.get('metadata') or {}).get('heading')}")
+    print(f"   content={content}")
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(description="Search the local 1C RAG lexical index.")
+    parser.add_argument("query")
+    parser.add_argument("--index", type=Path, default=DEFAULT_INDEX)
+    parser.add_argument("--profile", default="auto")
+    parser.add_argument("--limit", type=int)
+    parser.add_argument("--candidate-limit", type=int)
+    parser.add_argument("--dedupe-by-document", action="store_true")
+    parser.add_argument("--min-score", type=float)
+    parser.add_argument("--source-type", action="append", dest="source_types")
+    parser.add_argument("--json", action="store_true")
+    args = parser.parse_args()
+
+    index = read_json(args.index)
+    profile = resolve_rag_profile(args.profile, args.query)
+    source_types = args.source_types if args.source_types is not None else profile["source_types"]
+    results = search_lexical_index(
+        index,
+        query=args.query,
+        limit=args.limit or int(profile["limit"]),
+        candidate_limit=args.candidate_limit or int(profile["candidate_limit"]),
+        dedupe_by_document=args.dedupe_by_document or bool(profile["dedupe_by_document"]),
+        min_score=float(args.min_score if args.min_score is not None else profile["min_score"]),
+        source_types=source_types,
+    )
+
+    if args.json:
+        print(json.dumps(results, ensure_ascii=False, indent=2))
+    elif results:
+        for position, result in enumerate(results, start=1):
+            print_result(result, position)
+    else:
+        print("No matches.")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())