fix(b9): clean entity names in multi-entity-impressum (GT IMPRESSUM-001)

Der Multi-Entity-Check fängt Elli's USt-IdNr-Lücke (VW Group Charging
GmbH hat keine, Elli Mobility GmbH hat eine), aber Entity-Namen waren
mit Header-Noise verunreinigt:

  'Impressum\n\nVolkswagen Group Charging GmbH'
  'eco\n\nElli Mobility GmbH'

Behoben:
  - _ENTITY_PAT lässt nur Space im Namen zu (kein \s/\n mehr)
  - _clean_entity_name() trimmt Header-Worte (Impressum, Anbieter, ...)
    und nimmt nur die letzte Zeile vor Legal-Form-Suffix
  - 11 neue Tests, davon einer mit Elli-like Impressum als
    Charakterisierungs-Test

Damit ist die finale Finding-Ausgabe für Audit-Reports lesbar
('Fehlt bei: Volkswagen Group Charging GmbH') statt verunreinigt.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

backend-compliance/compliance/services/impressum_multi_entity_check.py

@@ -23,11 +23,17 @@ import re
 logger = logging.getLogger(__name__)
 
 _ENTITY_PAT = re.compile(
-    r"([A-ZÄÖÜ][\w\-\&\s]{1,50}?\s+(?:GmbH|AG|UG|KG|SE|"
+    r"([A-ZÄÖÜ][\w\-\& ]{1,50}?\s+(?:GmbH|AG|UG|KG|SE|"
     r"e\.V\.|GbR|OHG|Limited|Ltd|LLC))",
     re.IGNORECASE,
 )
 
+_NAME_NOISE_PAT = re.compile(
+    r"^(?:Impressum|Anbieter|Anbieterkennzeichnung|Diensteanbieter|"
+    r"Verantwortlich(?:er)?|Kontakt|Adresse|@\S+|.+@.+)\s*[:|\-]?\s*",
+    re.IGNORECASE,
+)
+
 _USTID_PAT = re.compile(r"\b(?:USt-?Id(?:Nr)?\.?|VAT(?:-?Id)?)\s*[:.\s]\s*"
                         r"(DE\d{8,10}|[A-Z]{2}\d{6,12})", re.IGNORECASE)
 _HR_PAT = re.compile(r"\b(?:HR[BA]|Handelsregister|Registergericht)"
@@ -36,6 +42,17 @@ _GF_PAT = re.compile(r"(?:Geschäftsführer|Vertretungsberechtigt|"
                      r"vertreten\s+durch)\s*[:.\s]+", re.IGNORECASE)
 
 
+def _clean_entity_name(raw: str) -> str:
+    """Strip leading header noise + collapse whitespace."""
+    name = raw.strip()
+    # If the match spans multiple lines (regex captured a header before
+    # the actual company name), keep only the last line.
+    if "\n" in name:
+        name = name.rsplit("\n", 1)[-1].strip()
+    name = _NAME_NOISE_PAT.sub("", name).strip()
+    return re.sub(r"\s+", " ", name)
+
+
 def _slice_entities(text: str) -> list[tuple[str, str]]:
     """Return [(entity_name, text_slice)] for each detected entity."""
     matches = list(_ENTITY_PAT.finditer(text))
@@ -45,7 +62,7 @@ def _slice_entities(text: str) -> list[tuple[str, str]]:
     for i, m in enumerate(matches):
         start = m.start()
         end = matches[i + 1].start() if i + 1 < len(matches) else len(text)
-        slices.append((m.group(1).strip(), text[start:end]))
+        slices.append((_clean_entity_name(m.group(1)), text[start:end]))
     return slices
 
 

backend-compliance/tests/test_impressum_multi_entity_check.py

@@ -0,0 +1,109 @@
+"""Tests for B9 Multi-Entity-Impressum-Check.
+
+GT-Anker: Elli IMPRESSUM-001 — bei mehreren Entitäten USt-IdNr nur
+bei einer von zwei vorhanden → Finding muss per-Entity erkennen.
+"""
+
+from compliance.services.impressum_multi_entity_check import (
+    _clean_entity_name,
+    _slice_entities,
+    check_multi_entity_impressum,
+)
+
+
+_ELLI_LIKE = """
+Impressum
+
+Volkswagen Group Charging GmbH
+Karl-Liebknecht-Str. 32
+10178 Berlin
+Amtsgericht Charlottenburg HRB 208967 B
+Geschäftsführer: Giovanni Palazzo, Mark Möller
+Telefon: 00800 3554 1111
+E-Mail: info@elli.eco
+
+Elli Mobility GmbH
+Karl-Liebknecht-Str. 32
+10178 Berlin
+Amtsgericht Charlottenburg HRB 274616 B
+USt-IdNr.: DE814424009
+Geschäftsführer: Joschi Jennermann
+Telefon: 00800 00002030
+E-Mail: ellimobility@elli.eco
+"""
+
+
+class TestCleanEntityName:
+    def test_strips_header_prefix(self):
+        assert _clean_entity_name(
+            "Impressum\n\nVolkswagen Group Charging GmbH"
+        ) == "Volkswagen Group Charging GmbH"
+
+    def test_strips_email_artifact(self):
+        # mid-text matches may capture an email TLD before the next entity
+        assert _clean_entity_name(
+            "eco\n\nElli Mobility GmbH"
+        ) == "Elli Mobility GmbH"
+
+    def test_collapses_whitespace(self):
+        assert _clean_entity_name(
+            "Acme   Holding   GmbH"
+        ) == "Acme Holding GmbH"
+
+    def test_clean_name_passes_through(self):
+        assert _clean_entity_name("Elli Mobility GmbH") == "Elli Mobility GmbH"
+
+
+class TestSliceEntities:
+    def test_no_entities_empty(self):
+        assert _slice_entities("Random text without companies.") == []
+
+    def test_single_entity_no_slice(self):
+        # Single-entity impressum is the normal case — multi-entity-only.
+        assert _slice_entities("Acme GmbH\nMusterstr 1\n") == []
+
+    def test_two_entities_clean_names(self):
+        slices = _slice_entities(_ELLI_LIKE)
+        names = [n for n, _ in slices]
+        assert names == [
+            "Volkswagen Group Charging GmbH",
+            "Elli Mobility GmbH",
+        ]
+
+
+class TestCheck:
+    def test_elli_pattern_finds_missing_ust_id(self):
+        findings = check_multi_entity_impressum(
+            {"doc_texts": {"impressum": _ELLI_LIKE}}
+        )
+        ust_findings = [
+            f for f in findings if f["check_id"] == "IMPRESSUM-MULTI-UST_ID"
+        ]
+        assert len(ust_findings) == 1
+        f = ust_findings[0]
+        assert f["entities_missing"] == ["Volkswagen Group Charging GmbH"]
+        assert f["entities_present"] == ["Elli Mobility GmbH"]
+        assert f["severity"] == "MEDIUM"
+        assert "TMG" in f["norm"]
+
+    def test_no_impressum_no_findings(self):
+        assert check_multi_entity_impressum({"doc_texts": {}}) == []
+
+    def test_single_entity_no_findings(self):
+        text = "Acme GmbH\nMusterstr 1\nUSt-IdNr DE123456789"
+        assert check_multi_entity_impressum(
+            {"doc_texts": {"impressum": text}}
+        ) == []
+
+    def test_both_entities_have_ust_id_no_finding(self):
+        text = (
+            "Acme GmbH\nUSt-IdNr DE111111111\n\n"
+            "Foo Holding GmbH\nUSt-IdNr DE222222222\n"
+        )
+        ust_findings = [
+            f for f in check_multi_entity_impressum(
+                {"doc_texts": {"impressum": text}}
+            )
+            if f["check_id"] == "IMPRESSUM-MULTI-UST_ID"
+        ]
+        assert ust_findings == []