From 2ed3118c83c90ac77674daff8da60529a7a8295d Mon Sep 17 00:00:00 2001
From: Christian <christian@blaahund.dk>
Date: Mon, 2 Mar 2026 09:01:43 +0100
Subject: [PATCH] fix: exclude OWN_CVR from AI vendor extraction v2.2.25

---
 VERSION                                  |  2 +-
 app/billing/backend/supplier_invoices.py | 33 +++++++++++++++++++++++-
 app/services/ollama_service.py           | 12 ++++++---
 3 files changed, 42 insertions(+), 5 deletions(-)

diff --git a/VERSION b/VERSION
index 8389c48..05f68fc 100644
--- a/VERSION
+++ b/VERSION
@@ -1 +1 @@
-2.2.24
+2.2.25
diff --git a/app/billing/backend/supplier_invoices.py b/app/billing/backend/supplier_invoices.py
index cb5239c..e1418bb 100644
--- a/app/billing/backend/supplier_invoices.py
+++ b/app/billing/backend/supplier_invoices.py
@@ -2326,7 +2326,38 @@ async def reprocess_uploaded_file(file_id: int):
             
             extracted_fields = llm_result
             confidence = llm_result.get('confidence', 0.75)
-            
+
+            # Post-process: clear own CVR if AI mistakenly returned it
+            extracted_cvr = llm_result.get('vendor_cvr')
+            own_cvr = getattr(settings, 'OWN_CVR', '')
+            if extracted_cvr and own_cvr and str(extracted_cvr).replace('DK', '').strip() == str(own_cvr).strip():
+                logger.warning(f"⚠️ AI returned own CVR ({own_cvr}) as vendor_cvr - clearing it")
+                llm_result['vendor_cvr'] = None
+                extracted_cvr = None
+
+            # Try to find vendor in DB by extracted CVR or name (overrides detected_vendor_id)
+            if extracted_cvr:
+                cvr_clean = str(extracted_cvr).replace('DK', '').strip()
+                vendor_row = execute_query_single(
+                    "SELECT id FROM vendors WHERE cvr_number = %s AND is_active = true",
+                    (cvr_clean,))
+                if vendor_row:
+                    vendor_id = vendor_row['id']
+                    logger.info(f"✅ Matched vendor by CVR {cvr_clean}: vendor_id={vendor_id}")
+                    execute_update(
+                        "UPDATE incoming_files SET detected_vendor_id = %s WHERE file_id = %s",
+                        (vendor_id, file_id))
+            if not vendor_id and llm_result.get('vendor_name'):
+                vendor_row = execute_query_single(
+                    "SELECT id FROM vendors WHERE name ILIKE %s AND is_active = true ORDER BY id LIMIT 1",
+                    (f"%{llm_result['vendor_name']}%",))
+                if vendor_row:
+                    vendor_id = vendor_row['id']
+                    logger.info(f"✅ Matched vendor by name '{llm_result['vendor_name']}': vendor_id={vendor_id}")
+                    execute_update(
+                        "UPDATE incoming_files SET detected_vendor_id = %s WHERE file_id = %s",
+                        (vendor_id, file_id))
+
             # Store AI extracted data in extractions table
             extraction_id = execute_insert(
                 """INSERT INTO extractions
diff --git a/app/services/ollama_service.py b/app/services/ollama_service.py
index df84467..65beb74 100644
--- a/app/services/ollama_service.py
+++ b/app/services/ollama_service.py
@@ -28,14 +28,20 @@ class OllamaService:
     
     def _build_system_prompt(self) -> str:
         """Build Danish system prompt for invoice extraction with CVR"""
-        return """Du er en ekspert i at læse og udtrække strukturerede data fra danske fakturaer, kreditnotaer og leverandørdokumenter.
+        own_cvr = getattr(settings, 'OWN_CVR', '29522790')
+        own_cvr_rule = (
+            f"4b. KRITISK - LEVERANDØR CVR: CVR {own_cvr} er VORES eget CVR (køberen/modtageren). "
+            f"Sæt ALDRIG vendor_cvr til {own_cvr}! Leverandørens CVR er CVR-nummeret der hører til "
+            f"firmaet som har SENDT fakturaen (ikke modtageren).\n"
+        )
+        return ("""Du er en ekspert i at læse og udtrække strukturerede data fra danske fakturaer, kreditnotaer og leverandørdokumenter.
 
 VIGTIGE REGLER:
 1. Returner KUN gyldig JSON - ingen forklaring eller ekstra tekst
 2. Hvis et felt ikke findes, sæt det til null
 3. Beregn confidence baseret på hvor sikker du er på hvert felt (0.0-1.0)
 4. Datoer skal være i format YYYY-MM-DD
-5. DANSKE PRISFORMATER: 
+""" + own_cvr_rule + """5. DANSKE PRISFORMATER: 
    - Tusind-separator kan være . (punkt) eller mellemrum: "5.965,18" eller "5 965,18"
    - Decimal-separator er , (komma): "1.234,56 kr"
    - I JSON output skal du bruge . (punkt) som decimal: 1234.56
@@ -126,7 +132,7 @@ Output: {
     "confidence": 0.95
   }],
   "confidence": 0.95
-}"""
+}""")
     
     async def extract_from_text(self, text: str) -> Dict:
         """