Allow zero-euro line item amounts

tests/test_extractor.py

@@ -4,8 +4,10 @@ Tests are written following TDD: FAILING TESTS FIRST (RED phase),
 then implementation makes them pass (GREEN phase).
 """
 
-import pytest
 import base64
+import binascii
+
+import pytest
 
 from pypdf import PdfReader, PdfWriter
 
@@ -49,7 +51,7 @@ class TestFileSizeValidation:
 
     def test_file_size_limit_exactly_10mb(self):
         """Test PDF exactly at 10MB limit passes size check but fails PDF parsing."""
-        from src.extractor import extract_zugferd, ExtractionError
+        from src.extractor import ExtractionError, extract_zugferd
 
         # 10MB = 10 * 1024 * 1024 bytes
         large_pdf = b"X" * (10 * 1024 * 1024)
@@ -63,7 +65,7 @@ class TestFileSizeValidation:
 
     def test_file_size_limit_10mb_plus_one_byte(self):
         """Test PDF one byte over 10MB limit is rejected."""
-        from src.extractor import extract_zugferd, ExtractionError
+        from src.extractor import ExtractionError, extract_zugferd
 
         # 10MB + 1 byte
         too_large = b"X" * (10 * 1024 * 1024 + 1)
@@ -75,7 +77,7 @@ class TestFileSizeValidation:
 
     def test_file_size_under_10mb_accepted(self):
         """Test PDF under 10MB is accepted for processing."""
-        from src.extractor import extract_zugferd, ExtractionError
+        from src.extractor import ExtractionError, extract_zugferd
 
         # Small PDF (9MB)
         small_pdf = b"X" * (9 * 1024 * 1024)
@@ -166,8 +168,8 @@ class TestEN16931Extraction:
         assert first_item.description is not None and len(first_item.description) > 0
         assert first_item.quantity > 0
         assert first_item.unit is not None and len(first_item.unit) > 0
-        assert first_item.unit_price > 0
-        assert first_item.line_total > 0
+        assert first_item.unit_price is not None and first_item.unit_price > 0
+        assert first_item.line_total is not None and first_item.line_total > 0
 
         # Totals
         assert xml_data.totals.line_total_sum > 0
@@ -181,7 +183,7 @@ class TestErrorHandling:
 
     def test_corrupt_pdf_raises_error(self):
         """Test corrupt PDF raises ExtractionError with correct code."""
-        from src.extractor import extract_zugferd, ExtractionError
+        from src.extractor import ExtractionError, extract_zugferd
 
         # Invalid PDF data
         corrupt_pdf = b"NOT A PDF FILE AT ALL"
@@ -194,14 +196,14 @@ class TestErrorHandling:
 
     def test_empty_pdf_raises_error(self):
         """Test empty PDF raises ExtractionError."""
-        from src.extractor import extract_zugferd, ExtractionError
+        from src.extractor import ExtractionError, extract_zugferd
 
         with pytest.raises(ExtractionError):
             extract_zugferd(b"")
 
     def test_invalid_base64(self):
         """Test invalid base64 raises ExtractionError."""
-        from src.extractor import extract_zugferd, ExtractionError
+        from src.extractor import ExtractionError, extract_zugferd
 
         # This would be called by API layer, but we can test the concept
         # Invalid PDF that's not valid base64-encoded
@@ -210,7 +212,7 @@ class TestErrorHandling:
             # If API layer decodes invalid base64, it gets error
             decoded = base64.b64decode(invalid_base64, validate=True)
             extract_zugferd(decoded)
-        except (base64.binascii.Error, ValueError):
+        except (binascii.Error, ValueError):
             # base64 error is expected
             pass
         except ExtractionError as e:
@@ -232,8 +234,6 @@ class TestPDFTextExtraction:
 
         assert result.pdf_text is not None
         assert len(result.pdf_text) > 0
-        # Should contain some common German invoice terms
-        text_lower = result.pdf_text.lower()
         # PDF text may contain invoice-related terms in German or English