feat(rag): regulation_short Casing-Normalisierung am Ingest-Rand

Der Re-Ingest leitet regulation_short z.T. via title()-Casing aus Dateinamen ab
('dsgvo'->'Dsgvo', 'osha otm'->'Osha Otm') -> falsche Akronyme im Payload UND im
article_label ('Art. 37 Dsgvo'). NEU: normalize_regulation_short() in legal_metadata,
token-basiert mit kuratiertem Akronym-Set -> nur gelistete Akronyme werden gross,
legitimes Mixed-Case (GeschGehG, MuSchG, GoBD, MiCA, eIDAS, EuGH) bleibt unberuehrt.
Angewandt am Ingest-Rand in documents.py (greift fuer Payload-Feld + display_name).
+13 Tests gruen. Bestandsdaten brauchen separaten einmaligen Qdrant-Patch.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

rag-service/api/documents.py

@@ -10,7 +10,12 @@ from embedding_client import embedding_client
 from html_utils import decode_html_bytes, looks_like_html, strip_html
 from minio_client_wrapper import minio_wrapper
 from qdrant_client_wrapper import qdrant_wrapper
-from legal_metadata import build_legal_fields, compute_chunk_hash, deterministic_point_id
+from legal_metadata import (
+    build_legal_fields,
+    compute_chunk_hash,
+    deterministic_point_id,
+    normalize_regulation_short,
+)
 
 logger = logging.getLogger("rag-service.api.documents")
 
@@ -155,6 +160,13 @@ async def upload_document(
         except json.JSONDecodeError:
             logger.warning("Invalid metadata_json, ignoring")
 
+    # Casing-Normalisierung am Ingest-Rand: title-caste Akronyme korrigieren
+    # ('Dsgvo'->'DSGVO'), damit Payload-Feld UND article_label sauber sind.
+    if extra_metadata.get("regulation_short"):
+        extra_metadata["regulation_short"] = normalize_regulation_short(
+            extra_metadata["regulation_short"]
+        )
+
     # --- Build payloads (rag_reingest_spec.md §2/§3: zitierfaehige Legal-Metadaten) ---
     reg_code = (
         extra_metadata.get("regulation_code")