natural-pdf 0.1.5__py3-none-any.whl → 0.1.7__py3-none-any.whl

tests/test_optional_deps.py

@@ -1,298 +0,0 @@
-import importlib  # Use importlib for checking
-import sys
-
-import pytest
-
-from natural_pdf import PDF, PDFCollection  # Import PDFCollection
-from natural_pdf.core.page import Page
-
-# --- Fixtures --- #
-
-# Define PDF paths relative to the project root (where pytest is usually run)
-TEST_PDF_URL = "https://github.com/jsoma/natural-pdf/raw/refs/heads/main/pdfs/01-practice.pdf"
-NEEDS_OCR_PDF_PATH = "pdfs/needs-ocr.pdf"
-STANDARD_PDF_PATH = "pdfs/01-practice.pdf"
-
-
-@pytest.fixture(scope="module")
-def standard_pdf_page():
-    """Fixture to load the first page of the standard test PDF."""
-    try:
-        # Use the local path if available, otherwise fallback to URL?
-        # For consistency in tests, let's stick to the local path for now.
-        # Assume the pdfs directory is in the root alongside tests/
-        pdf = PDF(STANDARD_PDF_PATH)
-        if not pdf.pages:
-            pytest.fail(f"Standard PDF has no pages: {STANDARD_PDF_PATH}")
-        return pdf.pages[0]
-    except Exception as e:
-        pytest.fail(f"Failed to load standard PDF ({STANDARD_PDF_PATH}) for module tests: {e}")
-
-
-@pytest.fixture(scope="module")
-def needs_ocr_pdf_page():
-    """Fixture to load the first page of the OCR test PDF."""
-    try:
-        pdf = PDF(NEEDS_OCR_PDF_PATH)
-        if not pdf.pages:
-            pytest.fail(f"OCR PDF has no pages: {NEEDS_OCR_PDF_PATH}")
-        return pdf.pages[0]
-    except Exception as e:
-        pytest.fail(f"Failed to load OCR PDF ({NEEDS_OCR_PDF_PATH}) for module tests: {e}")
-
-
-@pytest.fixture(scope="module")
-def standard_pdf_collection():
-    """Fixture to create a PDFCollection with the standard test PDF."""
-    try:
-        # Use a list containing the path
-        collection = PDFCollection([STANDARD_PDF_PATH])
-        assert len(collection.pdfs) == 1
-        return collection
-    except Exception as e:
-        pytest.fail(f"Failed to create PDFCollection ({STANDARD_PDF_PATH}) for module tests: {e}")
-
-
-# --- Helper --- #
-
-
-def is_extra_installed(extra_name):
-    """Checks if packages associated with an extra appear importable."""
-    extra_packages = {
-        "interactive": ["ipywidgets"],
-        "easyocr": ["easyocr"],
-        "paddle": ["paddleocr"],
-        "surya": ["surya"],
-        "layout_yolo": ["doclayout_yolo"],  # Correct package name for this extra
-        "haystack": ["haystack"],
-        "qa": ["transformers"],
-    }
-    if extra_name not in extra_packages:
-        return False
-
-    packages_to_check = extra_packages[extra_name]
-    try:
-        for pkg_name in packages_to_check:
-            importlib.import_module(pkg_name)
-        return True
-    except ImportError:
-        return False
-
-
-# --- Interactive Viewer (ipywidgets) Tests --- (Existing)
-
-
-def test_ipywidgets_availability_flag():
-    """Tests the internal _IPYWIDGETS_AVAILABLE flag based on environment."""
-    try:
-        from natural_pdf.widgets.viewer import _IPYWIDGETS_AVAILABLE
-
-        flag_value = _IPYWIDGETS_AVAILABLE
-    except ImportError:
-        pytest.fail("Could not import or find _IPYWIDGETS_AVAILABLE in natural_pdf.widgets.viewer")
-    should_be_installed = is_extra_installed("interactive")
-    assert (
-        flag_value == should_be_installed
-    ), f"_IPYWIDGETS_AVAILABLE flag mismatch. Expected {should_be_installed}, got {flag_value}."
-
-
-def test_page_viewer_widget_creation_when_installed(standard_pdf_page):
-    """Tests that Page.viewer() returns a widget when ipywidgets is installed."""
-    pytest.importorskip("ipywidgets")
-    from natural_pdf.widgets.viewer import SimpleInteractiveViewerWidget
-
-    viewer_instance = standard_pdf_page.viewer()
-    assert (
-        viewer_instance is not None
-    ), "Page.viewer() should return an object when ipywidgets is installed."
-    assert isinstance(
-        viewer_instance, SimpleInteractiveViewerWidget
-    ), f"Page.viewer() returned type {type(viewer_instance)}, expected SimpleInteractiveViewerWidget."
-
-
-def test_page_viewer_widget_creation_when_not_installed(standard_pdf_page):
-    """Tests that Page.viewer() returns None when ipywidgets is missing."""
-    if is_extra_installed("interactive"):
-        pytest.skip("Skipping test: ipywidgets IS installed in this environment.")
-    viewer_instance = standard_pdf_page.viewer()
-    assert (
-        viewer_instance is None
-    ), "Page.viewer() should return None when ipywidgets is not installed."
-
-
-# --- EasyOCR Tests --- #
-
-
-def test_ocr_easyocr_works_when_installed(needs_ocr_pdf_page):
-    """Test running EasyOCR when installed."""
-    pytest.importorskip("easyocr")
-    try:
-        # Use extract_ocr_elements which doesn't modify the page state
-        ocr_elements = needs_ocr_pdf_page.extract_ocr_elements(engine="easyocr")
-        assert isinstance(ocr_elements, list)
-        assert len(ocr_elements) > 0, "EasyOCR should find text elements on the OCR PDF."
-        # Check if the first element looks like a TextElement (basic check)
-        assert hasattr(ocr_elements[0], "text"), "OCR result should have text attribute."
-        assert hasattr(ocr_elements[0], "bbox"), "OCR result should have bbox attribute."
-    except Exception as e:
-        pytest.fail(f"EasyOCR extraction failed when installed: {e}")
-
-
-def test_ocr_easyocr_fails_gracefully_when_not_installed(needs_ocr_pdf_page):
-    """Test calling EasyOCR when not installed."""
-    if is_extra_installed("easyocr"):
-        pytest.skip("Skipping test: EasyOCR IS installed.")
-    # Check how OCRManager handles unavailable engines - assuming it returns empty list
-    ocr_elements = needs_ocr_pdf_page.extract_ocr_elements(engine="easyocr")
-    assert (
-        ocr_elements == []
-    ), "extract_ocr_elements should return empty list for unavailable engine."
-
-
-# --- PaddleOCR Tests --- #
-
-
-def test_ocr_paddle_works_when_installed(needs_ocr_pdf_page):
-    """Test running PaddleOCR when installed."""
-    pytest.importorskip("paddleocr")
-    if sys.platform == "darwin":
-        pytest.skip("PaddleOCR tests skipped on macOS")
-    try:
-        ocr_elements = needs_ocr_pdf_page.extract_ocr_elements(engine="paddle")
-        assert isinstance(ocr_elements, list)
-        assert len(ocr_elements) > 0, "PaddleOCR should find text elements on the OCR PDF."
-        assert hasattr(ocr_elements[0], "text")
-        assert hasattr(ocr_elements[0], "bbox")
-    except Exception as e:
-        pytest.fail(f"PaddleOCR extraction failed when installed: {e}")
-
-
-def test_ocr_paddle_fails_gracefully_when_not_installed(needs_ocr_pdf_page):
-    """Test calling PaddleOCR when not installed."""
-    if is_extra_installed("paddle"):
-        pytest.skip("Skipping test: PaddleOCR IS installed.")
-    if sys.platform == "darwin":  # Also skip if check fails but platform is darwin
-        pytest.skip("PaddleOCR tests skipped on macOS")
-    # Check how OCRManager handles unavailable engines - assume it returns empty list (KEEPING THIS)
-    # It might be reasonable for OCR manager to return [] if engine isn't there,
-    # vs layout which is explicitly requested.
-    # Alternatively, OCRManager could also be changed to raise errors.
-    ocr_elements = needs_ocr_pdf_page.extract_ocr_elements(engine="paddle")
-    assert (
-        ocr_elements == []
-    ), "extract_ocr_elements should return empty list for unavailable engine."
-
-
-# --- Surya Tests --- #
-
-
-def test_layout_surya_works_when_installed(standard_pdf_page):  # Use standard PDF for layout
-    """Test running Surya layout analysis when installed."""
-    pytest.importorskip("surya")
-    if sys.version_info < (3, 10):
-        pytest.skip("Surya tests skipped on Python < 3.10")
-    try:
-        layout_regions = standard_pdf_page.analyze_layout(engine="surya")
-        from natural_pdf.elements.collections import ElementCollection  # Import needed for check
-
-        assert isinstance(
-            layout_regions, ElementCollection
-        ), "analyze_layout should return an ElementCollection"
-        # Layout might return empty list, check type
-        # assert len(layout_regions) > 0, "Surya should find layout regions." # Keep commented unless specific PDF guarantees regions
-    except Exception as e:
-        pytest.fail(f"Surya layout analysis failed when installed: {e}")
-
-
-def test_layout_surya_fails_gracefully_when_not_installed(standard_pdf_page):
-    """Test calling Surya layout analysis when not installed raises error."""
-    if is_extra_installed("surya"):
-        pytest.skip("Skipping test: Surya IS installed.")
-    if sys.version_info < (3, 10):  # Also skip if check fails but Python is < 3.10
-        pytest.skip("Surya tests skipped on Python < 3.10")
-    # Expect RuntimeError because engine is known but unavailable
-    with pytest.raises(RuntimeError, match="not available"):
-        _ = standard_pdf_page.analyze_layout(engine="surya")
-
-
-# --- Layout YOLO Tests --- #
-
-
-def test_layout_yolo_works_when_installed(standard_pdf_page):
-    """Test running YOLO layout analysis when installed."""
-    # Check for the *actual* package associated with the extra
-    pytest.importorskip("doclayout_yolo")
-    try:
-        layout_regions = standard_pdf_page.analyze_layout(engine="yolo")
-        from natural_pdf.elements.collections import ElementCollection
-
-        assert isinstance(layout_regions, ElementCollection)
-    except Exception as e:
-        pytest.fail(f"YOLO layout analysis failed when installed: {e}")
-
-
-def test_layout_yolo_fails_gracefully_when_not_installed(standard_pdf_page):
-    """Test calling YOLO layout analysis when not installed raises error."""
-    if is_extra_installed("layout_yolo"):
-        pytest.skip("Skipping test: Layout YOLO IS installed.")
-    # Expect RuntimeError because engine is known but unavailable
-    with pytest.raises(RuntimeError, match="not available"):
-        _ = standard_pdf_page.analyze_layout(engine="yolo")
-
-
-# --- Haystack Tests --- #
-
-
-def test_search_haystack_works_when_installed(standard_pdf_collection):
-    """Test basic Haystack initialization via PDFCollection when installed."""
-    pytest.importorskip("haystack")
-    pytest.importorskip("chromadb")
-    try:
-        # Initialize search on the collection
-        # This should setup the SearchService internally
-        standard_pdf_collection.init_search(
-            index=False
-        )  # index=False to avoid actual indexing time
-        # Basic check: did init_search run without error?
-        # Optionally check if _search_service is now initialized on the collection
-        assert hasattr(standard_pdf_collection, "_search_service")
-        assert standard_pdf_collection._search_service is not None
-        # TODO: Add actual indexing and searching tests later
-    except ImportError as ie:
-        pytest.fail(f"Import failed even though haystack/chromadb seem installed: {ie}")
-    except Exception as e:
-        pytest.fail(f"Haystack integration failed when installed: {e}")
-
-
-def test_search_haystack_fails_gracefully_when_not_installed(standard_pdf_collection):
-    """Test initializing Haystack features via PDFCollection when not installed."""
-    if is_extra_installed("haystack"):
-        pytest.skip("Skipping test: Haystack IS installed.")
-
-    # Expect an ImportError when trying to initialize search
-    with pytest.raises(
-        ImportError,
-        match="Search Service could not be created. Ensure Haystack extras are installed",
-    ):
-        standard_pdf_collection.init_search(index=False)
-
-
-# --- QA Tests --- #
-
-
-def test_qa_works_when_installed(standard_pdf_page):
-    """Test basic QA functionality (requires transformers core dep)."""
-    # No importorskip needed as transformers is core
-    try:
-        # Simple question
-        result = standard_pdf_page.ask("What is this document about?")
-        assert isinstance(result, dict)
-        assert "answer" in result
-        assert "confidence" in result
-        # We don't know the answer, but it should run
-    except Exception as e:
-        pytest.fail(f"QA execution failed: {e}")
-
-
-# No 'fails gracefully' test needed for QA as its core dep (transformers) is always installed.
-# We might need tests for *specific models* if they require separate downloads/setup.