Spaces:
Running
Running
| import fitz | |
| from backend.services.pdf_processor import PDFProcessor | |
| def test_text_cleaning_and_hyphenation(): | |
| assert PDFProcessor()._clean_text("impor-\ntant words") == "important words" | |
| def test_content_doc_id_is_deterministic_and_content_based(tmp_path): | |
| processor = PDFProcessor() | |
| a = tmp_path / "a.pdf"; b = tmp_path / "b.pdf" | |
| a.write_bytes(b"same bytes"); b.write_bytes(b"same bytes") | |
| assert processor._generate_doc_id(a) == processor._generate_doc_id(b) | |
| b.write_bytes(b"different bytes") | |
| assert processor._generate_doc_id(a) != processor._generate_doc_id(b) | |
| def test_real_pdf_page_metadata_and_short_chunk_skipping(tmp_path): | |
| path = tmp_path / "sample.pdf" | |
| pdf = fitz.open(); page = pdf.new_page() | |
| page.insert_text((72, 72), "This is enough meaningful text to create a chunk with page metadata.") | |
| pdf.save(path); pdf.close() | |
| result = PDFProcessor().process(str(path)) | |
| assert result.total_pages == 1 | |
| assert result.chunks and all(chunk.page_number == 1 for chunk in result.chunks) | |
| assert all(len(chunk.text) >= 20 for chunk in result.chunks) | |
| def test_scanned_or_empty_pdf_raises_value_error(tmp_path): | |
| path = tmp_path / "blank.pdf" | |
| pdf = fitz.open(); pdf.new_page(); pdf.save(path); pdf.close() | |
| try: | |
| PDFProcessor().process(str(path)) | |
| except ValueError as exc: | |
| assert "No extractable text" in str(exc) | |
| else: | |
| raise AssertionError("Textless PDF must be rejected") | |