Benjamin_Boenisch/breakpilot-core
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

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

Browse Source 
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>
This commit is contained in:
Benjamin Admin
2026-06-21 18:31:45 +02:00
parent de542633e2
commit 0c5f1fd7a4
3 changed files with 71 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
rag-service/tests/test_legal_metadata.py
+24
View File
@@ -113,3 +113,27 @@ def test_build_legal_fields_ruling_uses_aktenzeichen_not_paragraph():
assert out["article"] == ""
assert out["article_label"] == "BAG, 1 ABR 22/21"
assert out["regulation_code"] == "BAG_1_ABR_22_21"
def test_normalize_regulation_short():
# title-caste Akronyme -> korrekt GROSS (Token-basiert)
assert L.normalize_regulation_short("Dsgvo") == "DSGVO"
assert L.normalize_regulation_short("Gpsr") == "GPSR"
assert L.normalize_regulation_short("Ifrs") == "IFRS"
assert L.normalize_regulation_short("EU Mdr") == "EU MDR"
assert L.normalize_regulation_short("OWASP Api TOP10") == "OWASP API TOP10"
assert (
L.normalize_regulation_short("Osha Otm Section-4-safety-hazards Chapter-1")
== "OSHA OTM Section-4-safety-hazards Chapter-1"
)
# legitimes Mixed-Case bleibt unberuehrt (NICHT in der Akronym-Liste)
assert L.normalize_regulation_short("GeschGehG") == "GeschGehG"
assert L.normalize_regulation_short("MuSchG") == "MuSchG"
assert L.normalize_regulation_short("GoBD") == "GoBD"
assert L.normalize_regulation_short("MiCA") == "MiCA"
assert L.normalize_regulation_short("eIDAS") == "eIDAS"
assert L.normalize_regulation_short("EuGH C-252/21 (Meta)") == "EuGH C-252/21 (Meta)"
# Ganz-Wert-Overrides + Leerfaelle
assert L.normalize_regulation_short("Dataact") == "Data Act"
assert L.normalize_regulation_short("") == ""
assert L.normalize_regulation_short(None) == ""