From 4058824f154d43ee5692bdd8ebd4bdb17335bc60 Mon Sep 17 00:00:00 2001
From: TrisNol <tristan.nolde@yahoo.de>
Date: Tue, 17 Oct 2023 17:56:26 +0200
Subject: [PATCH] fix(data-extraction): Resolve regex issue in detecting
 auditors

---
 .../utils/data_extraction/bundesanzeiger.py                | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/src/aki_prj23_transparenzregister/utils/data_extraction/bundesanzeiger.py b/src/aki_prj23_transparenzregister/utils/data_extraction/bundesanzeiger.py
index 0a4a4e9..f59c889 100644
--- a/src/aki_prj23_transparenzregister/utils/data_extraction/bundesanzeiger.py
+++ b/src/aki_prj23_transparenzregister/utils/data_extraction/bundesanzeiger.py
@@ -102,12 +102,9 @@ class Bundesanzeiger:
             list[Auditor]: List of Auditors found in the given report
         """
         auditor_company = self.extract_auditor_company(report)
-        auditor_regex = r"[a-z A-Z,.'-]+, Wirtschaftsprüfer"
+        auditor_regex = r"([a-z A-ZÄäÜüÖö,.'-]+), Wirtschaftsprüfer(in)?"
         hits = re.findall(auditor_regex, report)
-        return [
-            Auditor(hit.replace(", Wirtschaftsprüfer", "").lstrip(), auditor_company)
-            for hit in hits
-        ]
+        return [Auditor(hit[0].strip(), auditor_company) for hit in hits]
 
     def __extract_kpis__(self, report: str) -> dict:
         """Source: https://github.com/bundesAPI/deutschland/pull/87/files#diff-f5b9db5384cf523fcc677056065041e7793bfc4da9cf74c4eebd6fab732739bd.