docid 0.1.3__tar.gz → 0.1.4__tar.gz

{docid-0.1.3 → docid-0.1.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docid
-Version: 0.1.3
+Version: 0.1.4
 Summary: Deterministyczny generator identyfikatorów dokumentów z OCR
 Home-page: https://github.com/softreck/docid
 Author: Softreck

{docid-0.1.3 → docid-0.1.4}/docid.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docid
-Version: 0.1.3
+Version: 0.1.4
 Summary: Deterministyczny generator identyfikatorów dokumentów z OCR
 Home-page: https://github.com/softreck/docid
 Author: Softreck

{docid-0.1.3 → docid-0.1.4}/docid.egg-info/SOURCES.txt

@@ -18,4 +18,5 @@ exef_docid/pipeline.py
 exef_docid/extractors/__init__.py
 exef_docid/extractors/base.py
 tests/test_document_id.py
-tests/test_extractors.py
+tests/test_extractors.py
+tests/test_samples_id.py

{docid-0.1.3 → docid-0.1.4}/exef_docid/__init__.py

@@ -1,5 +1,5 @@
 """
-EXEF Document ID Generator
+DOC Document ID Generator
 
 Deterministyczny generator identyfikatorów dokumentów z OCR.
 Generuje zawsze ten sam ID dla tego samego dokumentu,
@@ -10,14 +10,14 @@ Przykład użycia:
 
     # Pełne przetwarzanie
     result = process_document("faktura.pdf")
-    print(result.document_id)      # EXEF-FV-A7B3C9D2E1F04856
+    print(result.document_id)      # DOC-FV-A7B3C9D2E1F04856
     print(result.extraction.issuer_nip)  # 5213017228
 
     # Tylko ID
     doc_id = get_document_id("paragon.jpg")
 
     # Weryfikacja
-    is_same = verify_document_id("skan.png", "EXEF-FV-A7B3C9D2E1F04856")
+    is_same = verify_document_id("skan.png", "DOC-FV-A7B3C9D2E1F04856")
 
 Wymagania:
     pip install paddleocr paddlepaddle pdf2image pillow

{docid-0.1.3 → docid-0.1.4}/exef_docid/cli.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
 """
-CLI dla EXEF Document ID Generator.
+CLI dla DOC Document ID Generator.
 
 Użycie:
     # Przetwórz pojedynczy plik
@@ -13,7 +13,7 @@ Użycie:
     docid batch ./dokumenty/ --output results.json
 
     # Weryfikacja ID
-    docid verify faktura.pdf EXEF-FV-A7B3C9D2E1F04856
+    docid verify faktura.pdf DOC-FV-A7B3C9D2E1F04856
 
     # Tylko OCR (bez generowania ID)
     docid ocr skan.jpg
@@ -267,7 +267,7 @@ def cmd_generate_id(args):
 
 def main():
     parser = argparse.ArgumentParser(
-        description='EXEF Document ID Generator - deterministyczne ID dokumentów z OCR',
+        description='DOC Document ID Generator - deterministyczne ID dokumentów z OCR',
         formatter_class=argparse.RawDescriptionHelpFormatter,
     )
     parser.add_argument('--version', action='version', version='docid 0.1.0')
@@ -279,7 +279,7 @@ def main():
     common.add_argument('--engine', choices=['paddle', 'tesseract'], default='paddle',
                        help='Silnik OCR (domyślnie: paddle)')
     common.add_argument('--lang', default='pl', help='Język dokumentów')
-    common.add_argument('--prefix', default='EXEF', help='Prefiks ID')
+    common.add_argument('--prefix', default='DOC', help='Prefiks ID')
     common.add_argument('--gpu', action='store_true', help='Użyj GPU')
     common.add_argument('-v', '--verbose', action='store_true', help='Więcej szczegółów')
 
@@ -324,7 +324,7 @@ def main():
     p_gen.add_argument('--number', help='Numer dokumentu')
     p_gen.add_argument('--date', help='Data (YYYY-MM-DD)')
     p_gen.add_argument('--amount', help='Kwota brutto')
-    p_gen.add_argument('--prefix', default='EXEF', help='Prefiks ID')
+    p_gen.add_argument('--prefix', default='DOC', help='Prefiks ID')
     p_gen.set_defaults(func=cmd_generate_id)
 
     args = parser.parse_args()

{docid-0.1.3 → docid-0.1.4}/exef_docid/cli_universal.py

@@ -431,7 +431,7 @@ def main():
     """Main CLI entry point"""
     parser = argparse.ArgumentParser(
         prog='docid',
-        description='EXEF Document ID Generator - CLI'
+        description='DOC Document ID Generator - CLI'
     )
     
     parser.add_argument('--version', action='version', version='%(prog)s 0.1.0')

{docid-0.1.3 → docid-0.1.4}/exef_docid/document_id.py

@@ -14,8 +14,8 @@ from decimal import ROUND_HALF_UP, Decimal
 from enum import Enum
 from typing import Optional, Union
 
-# Namespace UUID dla EXEF (RFC 4122 UUID v5)
-EXEF_NAMESPACE = uuid.UUID('a1b2c3d4-e5f6-7890-abcd-ef1234567890')
+# Namespace UUID dla DOC (RFC 4122 UUID v5)
+DOC_NAMESPACE = uuid.UUID('a1b2c3d4-e5f6-7890-abcd-ef1234567890')
 
 
 class DocumentType(Enum):
@@ -207,10 +207,10 @@ class DocumentIDGenerator:
     niezależnie od formatu źródłowego dokumentu.
     """
 
-    def __init__(self, prefix: str = "EXEF"):
+    def __init__(self, prefix: str = "DOC"):
         """
         Args:
-            prefix: Prefiks identyfikatora (domyślnie EXEF)
+            prefix: Prefiks identyfikatora (domyślnie DOC)
         """
         self.prefix = prefix
 
@@ -229,7 +229,7 @@ class DocumentIDGenerator:
 
         >>> gen = DocumentIDGenerator()
         >>> gen.generate_invoice_id("5213017228", "FV/2025/00142", "2025-01-15", 1230.00)
-        'EXEF-FV-A7B3C9D2E1F04856'
+        'DOC-FV-A7B3C9D2E1F04856'
         """
         canonical = CanonicalData(
             document_type=DocumentType.INVOICE,
@@ -265,7 +265,7 @@ class DocumentIDGenerator:
 
         >>> gen = DocumentIDGenerator()
         >>> gen.generate_receipt_id("5213017228", "2025-01-15", 45.99)
-        'EXEF-PAR-...'
+        'DOC-PAR-...'
         """
         parts = [
             NIPValidator.normalize(seller_nip),
@@ -649,7 +649,7 @@ class DocumentIDGenerator:
         Generuje finalny identyfikator z danych kanonicznych.
 
         Format: {PREFIX}-{TYPE}-{HASH16}
-        Przykład: EXEF-FV-A7B3C9D2E1F04856
+        Przykład: DOC-FV-A7B3C9D2E1F04856
         """
         # SHA256 z canonical string
         hash_bytes = hashlib.sha256(canonical.canonical_string.encode('utf-8')).digest()
@@ -662,7 +662,7 @@ class DocumentIDGenerator:
         Weryfikuje czy ID odpowiada danym kanonicznym.
 
         >>> gen = DocumentIDGenerator()
-        >>> gen.verify_id("EXEF-FV-A7B3C9D2E1F04856", "5213017228|FV/2025/00142|2025-01-15|1230.00")
+        >>> gen.verify_id("DOC-FV-A7B3C9D2E1F04856", "5213017228|FV/2025/00142|2025-01-15|1230.00")
         True
         """
         hash_bytes = hashlib.sha256(canonical_string.encode('utf-8')).digest()
@@ -679,8 +679,8 @@ class DocumentIDGenerator:
         """
         Parsuje identyfikator dokumentu.
 
-        >>> DocumentIDGenerator.parse_id("EXEF-FV-A7B3C9D2E1F04856")
-        {'prefix': 'EXEF', 'type': 'FV', 'hash': 'A7B3C9D2E1F04856',
+        >>> DocumentIDGenerator.parse_id("DOC-FV-A7B3C9D2E1F04856")
+        {'prefix': 'DOC', 'type': 'FV', 'hash': 'A7B3C9D2E1F04856',
          'document_type': <DocumentType.INVOICE>}
         """
         parts = document_id.split('-')

{docid-0.1.3 → docid-0.1.4}/exef_docid/ocr_processor.py

@@ -195,12 +195,8 @@ class PaddleOCRProcessor(BaseOCRProcessor):
                 self._ocr = PaddleOCR(
                     use_angle_cls=True,
                     lang=lang,
-                    use_gpu=self.use_gpu,
                     det_model_dir=self._det_model_dir,
                     rec_model_dir=self._rec_model_dir,
-                    # Optymalizacje CPU
-                    enable_mkldnn=True,
-                    cpu_threads=4,
                 )
             except ImportError:
                 raise ImportError(
@@ -216,7 +212,7 @@ class PaddleOCRProcessor(BaseOCRProcessor):
         self._init_ocr()
 
         image_path = str(image_path)
-        result = self._ocr.ocr(image_path, cls=True)
+        result = self._ocr.ocr(image_path)
 
         lines = []
         full_text_parts = []
@@ -439,8 +435,8 @@ class OCRProcessor:
 
     def __init__(
         self,
-        preferred_engine: OCREngine = OCREngine.PADDLE,
-        fallback_engine: OCREngine = OCREngine.TESSERACT,
+        preferred_engine: OCREngine = OCREngine.TESSERACT,
+        fallback_engine: OCREngine = OCREngine.PADDLE,
         lang: str = 'pl',
         use_gpu: bool = False,
     ):

{docid-0.1.3 → docid-0.1.4}/exef_docid/pipeline.py

@@ -97,20 +97,20 @@ class DocumentPipeline:
     Przykład użycia:
         pipeline = DocumentPipeline()
         result = pipeline.process("faktura.pdf")
-        print(result.document_id)  # EXEF-FV-A7B3C9D2E1F04856
+        print(result.document_id)  # DOC-FV-A7B3C9D2E1F04856
     """
 
     def __init__(
         self,
-        ocr_engine: OCREngine = OCREngine.PADDLE,
-        id_prefix: str = "EXEF",
+        ocr_engine: OCREngine = OCREngine.TESSERACT,
+        id_prefix: str = "DOC",
         lang: str = "pl",
         use_gpu: bool = False,
     ):
         """
         Args:
             ocr_engine: Silnik OCR (PADDLE lub TESSERACT)
-            id_prefix: Prefiks identyfikatorów (domyślnie EXEF)
+            id_prefix: Prefiks identyfikatorów (domyślnie DOC)
             lang: Język dokumentów (pl, en)
             use_gpu: Czy używać GPU (domyślnie False dla CPU)
         """
@@ -390,7 +390,7 @@ class DocumentPipeline:
 _default_pipeline: Optional[DocumentPipeline] = None
 
 
-def get_pipeline(ocr_engine: OCREngine = OCREngine.PADDLE) -> DocumentPipeline:
+def get_pipeline(ocr_engine: OCREngine = OCREngine.TESSERACT) -> DocumentPipeline:
     """Zwraca domyślny pipeline (lazy init)."""
     global _default_pipeline
     if _default_pipeline is None or _default_pipeline.ocr.preferred_engine != ocr_engine:
@@ -398,7 +398,7 @@ def get_pipeline(ocr_engine: OCREngine = OCREngine.PADDLE) -> DocumentPipeline:
     return _default_pipeline
 
 
-def process_document(file_path: Union[str, Path], ocr_engine: OCREngine = OCREngine.PADDLE, use_ocr: bool = True) -> ProcessedDocument:
+def process_document(file_path: Union[str, Path], ocr_engine: OCREngine = OCREngine.TESSERACT, use_ocr: bool = True) -> ProcessedDocument:
     """
     Przetwarza dokument i zwraca wynik z ID.
 
@@ -416,7 +416,7 @@ def get_document_id(file_path: Union[str, Path]) -> str:
 
     Przykład:
         doc_id = get_document_id("faktura.pdf")
-        print(doc_id)  # EXEF-FV-A7B3C9D2E1F04856
+        print(doc_id)  # DOC-FV-A7B3C9D2E1F04856
     """
     return get_pipeline().process(file_path).document_id
 
@@ -426,6 +426,6 @@ def verify_document_id(file_path: Union[str, Path], expected_id: str) -> bool:
     Weryfikuje czy dokument ma oczekiwany ID.
 
     Przykład:
-        is_valid = verify_document_id("skan.jpg", "EXEF-FV-A7B3C9D2E1F04856")
+        is_valid = verify_document_id("skan.jpg", "DOC-FV-A7B3C9D2E1F04856")
     """
     return get_pipeline().verify_document(file_path, expected_id)

{docid-0.1.3 → docid-0.1.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "docid"
-version = "0.1.3"
+version = "0.1.4"
 description = "Deterministyczny generator identyfikatorów dokumentów z OCR"
 readme = "README.md"
 license = {text = "MIT"}

docid-0.1.4/tests/test_samples_id.py

@@ -0,0 +1,277 @@
+"""
+Testy ID dla wszystkich próbek (samples).
+
+Sprawdza generowanie ID dla wszystkich plików w katalogu samples,
+w tym spójność ID między różnymi formatami tego samego dokumentu.
+"""
+
+import os
+from pathlib import Path
+
+import pytest
+
+from exef_docid.document_id import DocumentIDGenerator, DocumentType
+from exef_docid.pipeline import DocumentPipeline, process_document
+from exef_docid.ocr_processor import OCREngine
+
+
+SAMPLES_DIR = Path(__file__).parent.parent / "samples"
+
+
+class TestSampleIDGeneration:
+    """Testy generowania ID dla wszystkich próbek."""
+
+    @pytest.fixture
+    def pipeline(self):
+        """Fixture tworzący pipeline z Tesseract OCR."""
+        return DocumentPipeline(ocr_engine=OCREngine.TESSERACT)
+
+    def get_sample_files(self, subdirectory: str) -> list[Path]:
+        """Zwraca listę plików próbek z danego podkatalogu."""
+        sample_dir = SAMPLES_DIR / subdirectory
+        if not sample_dir.exists():
+            return []
+        
+        files = []
+        for ext in ['*.pdf', '*.jpg', '*.jpeg', '*.png', '*.xml', '*.html', '*.htm', '*.txt']:
+            files.extend(sample_dir.glob(ext))
+        return sorted(files)
+
+    def test_invoice_samples_generate_id(self, pipeline):
+        """Test generowania ID dla wszystkich próbek faktur."""
+        invoice_files = self.get_sample_files("invoices")
+        assert len(invoice_files) > 0, "Brak plików faktur w samples/invoices"
+        
+        results = []
+        for file_path in invoice_files:
+            try:
+                result = pipeline.process(file_path)
+                results.append({
+                    'file': file_path.name,
+                    'id': result.document_id,
+                    'type': result.document_type.value,
+                    'canonical': result.canonical_string,
+                    'confidence': result.ocr_confidence,
+                })
+                # Sprawdź czy ID zostało wygenerowane
+                assert result.document_id, f"Brak ID dla {file_path}"
+                assert result.document_id.startswith("EXEF-FV"), f"Nieprawidłowy prefix ID dla faktury: {result.document_id}"
+            except Exception as e:
+                pytest.fail(f"Błąd przetwarzania {file_path}: {e}")
+        
+        # Wypisz podsumowanie
+        print(f"\n=== Faktury ({len(results)} plików) ===")
+        for r in results:
+            print(f"  {r['file']}: {r['id']} (confidence: {r['confidence']:.2f})")
+
+    def test_receipt_samples_generate_id(self, pipeline):
+        """Test generowania ID dla wszystkich próbek paragonów."""
+        receipt_files = self.get_sample_files("receipts")
+        assert len(receipt_files) > 0, "Brak plików paragonów w samples/receipts"
+        
+        results = []
+        for file_path in receipt_files:
+            try:
+                result = pipeline.process(file_path)
+                results.append({
+                    'file': file_path.name,
+                    'id': result.document_id,
+                    'type': result.document_type.value,
+                    'confidence': result.ocr_confidence,
+                })
+                assert result.document_id, f"Brak ID dla {file_path}"
+            except Exception as e:
+                pytest.fail(f"Błąd przetwarzania {file_path}: {e}")
+        
+        print(f"\n=== Paragony ({len(results)} plików) ===")
+        for r in results:
+            print(f"  {r['file']}: {r['id']} (confidence: {r['confidence']:.2f})")
+
+    def test_contract_samples_generate_id(self, pipeline):
+        """Test generowania ID dla wszystkich próbek umów."""
+        contract_files = self.get_sample_files("contracts")
+        assert len(contract_files) > 0, "Brak plików umów w samples/contracts"
+        
+        results = []
+        for file_path in contract_files:
+            try:
+                result = pipeline.process(file_path)
+                results.append({
+                    'file': file_path.name,
+                    'id': result.document_id,
+                    'type': result.document_type.value,
+                    'confidence': result.ocr_confidence,
+                })
+                assert result.document_id, f"Brak ID dla {file_path}"
+            except Exception as e:
+                pytest.fail(f"Błąd przetwarzania {file_path}: {e}")
+        
+        print(f"\n=== Umowy ({len(results)} plików) ===")
+        for r in results:
+            print(f"  {r['file']}: {r['id']} (confidence: {r['confidence']:.2f})")
+
+    def test_invoice_cross_format_consistency(self, pipeline):
+        """Test czy różne formaty tej samej faktury mają ten sam ID."""
+        invoice_dir = SAMPLES_DIR / "invoices"
+        
+        # Znajdź wszystkie formaty faktury (faktura_full.*)
+        base_name = "faktura_full"
+        formats = []
+        
+        for ext in ['.jpg', '.jpeg', '.png', '.pdf', '.xml', '.html', '.txt']:
+            file_path = invoice_dir / f"{base_name}{ext}"
+            if file_path.exists():
+                formats.append(file_path)
+        
+        if len(formats) < 2:
+            pytest.skip(f"Za mało formatów faktury do testu (znaleziono: {len(formats)})")
+        
+        # Przetwórz wszystkie formaty
+        ids = {}
+        for file_path in formats:
+            try:
+                result = pipeline.process(file_path)
+                ids[file_path.suffix] = {
+                    'id': result.document_id,
+                    'canonical': result.canonical_string,
+                    'confidence': result.ocr_confidence,
+                }
+            except Exception as e:
+                pytest.fail(f"Błąd przetwarzania {file_path}: {e}")
+        
+        # Sprawdź czy wszystkie formaty mają ten sam ID
+        unique_ids = set(r['id'] for r in ids.values())
+        
+        print(f"\n=== Cross-format consistency dla {base_name} ===")
+        for ext, data in sorted(ids.items()):
+            print(f"  {ext}: {data['id']} (confidence: {data['confidence']:.2f})")
+        
+        assert len(unique_ids) == 1, (
+            f"Różne formaty {base_name} mają różne ID: {unique_ids}\n"
+            f"Szczegóły: {ids}"
+        )
+
+    def test_receipt_cross_format_consistency(self, pipeline):
+        """Test czy różne formaty tego samego paragonu mają ten sam ID."""
+        receipt_dir = SAMPLES_DIR / "receipts"
+        
+        base_name = "paragon_full"
+        formats = []
+        
+        for ext in ['.jpg', '.jpeg', '.png', '.pdf', '.xml', '.html', '.txt']:
+            file_path = receipt_dir / f"{base_name}{ext}"
+            if file_path.exists():
+                formats.append(file_path)
+        
+        if len(formats) < 2:
+            pytest.skip(f"Za mało formatów paragonu do testu (znaleziono: {len(formats)})")
+        
+        ids = {}
+        for file_path in formats:
+            try:
+                result = pipeline.process(file_path)
+                ids[file_path.suffix] = {
+                    'id': result.document_id,
+                    'canonical': result.canonical_string,
+                    'confidence': result.ocr_confidence,
+                }
+            except Exception as e:
+                pytest.fail(f"Błąd przetwarzania {file_path}: {e}")
+        
+        unique_ids = set(r['id'] for r in ids.values())
+        
+        print(f"\n=== Cross-format consistency dla {base_name} ===")
+        for ext, data in sorted(ids.items()):
+            print(f"  {ext}: {data['id']} (confidence: {data['confidence']:.2f})")
+        
+        assert len(unique_ids) == 1, (
+            f"Różne formaty {base_name} mają różne ID: {unique_ids}\n"
+            f"Szczegóły: {ids}"
+        )
+
+    def test_all_samples_summary(self, pipeline):
+        """Podsumowanie wszystkich próbek."""
+        all_results = []
+        
+        for subdir in ['invoices', 'receipts', 'contracts', 'universal']:
+            files = self.get_sample_files(subdir)
+            for file_path in files:
+                try:
+                    result = pipeline.process(file_path)
+                    all_results.append({
+                        'subdir': subdir,
+                        'file': file_path.name,
+                        'id': result.document_id,
+                        'type': result.document_type.value,
+                        'confidence': result.ocr_confidence,
+                    })
+                except Exception as e:
+                    all_results.append({
+                        'subdir': subdir,
+                        'file': file_path.name,
+                        'error': str(e),
+                    })
+        
+        print(f"\n{'='*60}")
+        print(f"PODSUMOWANIE WSZYSTKICH PRÓBEK ({len(all_results)} plików)")
+        print(f"{'='*60}")
+        
+        success_count = sum(1 for r in all_results if 'error' not in r)
+        error_count = sum(1 for r in all_results if 'error' in r)
+        
+        for r in sorted(all_results, key=lambda x: (x['subdir'], x['file'])):
+            if 'error' in r:
+                print(f"  [ERR] {r['subdir']}/{r['file']}: {r['error']}")
+            else:
+                print(f"  [OK]  {r['subdir']}/{r['file']}: {r['id']}")
+        
+        print(f"{'='*60}")
+        print(f"Sukcesy: {success_count}/{len(all_results)}, Błędy: {error_count}/{len(all_results)}")
+        print(f"{'='*60}")
+        
+        # Nie failujemy testu przy błędach - to tylko podsumowanie
+        assert success_count > 0, "Żaden plik nie został przetworzony pomyślnie"
+
+
+class TestSampleIDDeterminism:
+    """Testy determinizmu ID - te same dane = ten sam ID."""
+
+    def test_invoice_deterministic_id(self):
+        """Test czy faktura generuje ten sam ID przy każdym uruchomieniu."""
+        invoice_dir = SAMPLES_DIR / "invoices"
+        
+        # Znajdź pierwszy dostępny plik faktury
+        invoice_file = None
+        for ext in ['.txt', '.xml', '.html']:
+            candidate = invoice_dir / f"faktura_full{ext}"
+            if candidate.exists():
+                invoice_file = candidate
+                break
+        
+        if not invoice_file:
+            pytest.skip("Brak pliku faktury do testu determinizmu")
+        
+        # Generuj ID dwa razy
+        id1 = process_document(invoice_file).document_id
+        id2 = process_document(invoice_file).document_id
+        
+        assert id1 == id2, f"ID nie jest deterministyczne: {id1} != {id2}"
+
+    def test_receipt_deterministic_id(self):
+        """Test czy paragon generuje ten sam ID przy każdym uruchomieniu."""
+        receipt_dir = SAMPLES_DIR / "receipts"
+        
+        receipt_file = None
+        for ext in ['.txt', '.xml', '.html']:
+            candidate = receipt_dir / f"paragon_full{ext}"
+            if candidate.exists():
+                receipt_file = candidate
+                break
+        
+        if not receipt_file:
+            pytest.skip("Brak pliku paragonu do testu determinizmu")
+        
+        id1 = process_document(receipt_file).document_id
+        id2 = process_document(receipt_file).document_id
+        
+        assert id1 == id2, f"ID nie jest deterministyczne: {id1} != {id2}"