TST: Allow loading truncated images if required (#2586)

* TST: Allow loading truncated images if required

tests/__init__.py

@@ -143,3 +143,25 @@ def test_csv_consistency():
 
     # Ensure the urls are unique
     assert len(pdfs) == len({pdf["url"] for pdf in pdfs})
+
+
+class PILContext:
+    """Allow changing the PIL/Pillow configuration for some limited scope."""
+
+    def __init__(self):
+        self._saved_load_truncated_images = False
+
+    def __enter__(self):
+        # Allow loading incomplete images.
+        from PIL import ImageFile
+        self._saved_load_truncated_images = ImageFile.LOAD_TRUNCATED_IMAGES
+        ImageFile.LOAD_TRUNCATED_IMAGES = True
+        return self
+
+    def __exit__(self, type_, value, traceback):
+        from PIL import ImageFile
+        ImageFile.LOAD_TRUNCATED_IMAGES = self._saved_load_truncated_images
+        if type_:
+            # Error.
+            return
+        return True

tests/test_filters.py

@@ -21,7 +21,7 @@
 )
 from pypdf.generic import ArrayObject, DictionaryObject, NameObject, NumberObject
 
-from . import get_data_from_url
+from . import PILContext, get_data_from_url
 from .test_encryption import HAS_AES
 from .test_images import image_similarity
 
@@ -371,13 +371,14 @@ def test_tiff_predictor():
 @pytest.mark.enable_socket()
 def test_rgba():
     """Decode rgb with transparency"""
-    reader = PdfReader(BytesIO(get_data_from_url(name="tika-972174.pdf")))
-    data = reader.pages[0].images[0]
-    assert ".jp2" in data.name
-    similarity = image_similarity(
-        data.image, BytesIO(get_data_from_url(name="tika-972174_p0-im0.png"))
-    )
-    assert similarity > 0.99
+    with PILContext():
+        reader = PdfReader(BytesIO(get_data_from_url(name="tika-972174.pdf")))
+        data = reader.pages[0].images[0]
+        assert ".jp2" in data.name
+        similarity = image_similarity(
+            data.image, BytesIO(get_data_from_url(name="tika-972174_p0-im0.png"))
+        )
+        assert similarity > 0.99
 
 
 @pytest.mark.enable_socket()

tests/test_workflows.py

@@ -26,7 +26,7 @@
     read_object,
 )
 
-from . import get_data_from_url, normalize_warnings
+from . import PILContext, get_data_from_url, normalize_warnings
 
 TESTS_ROOT = Path(__file__).parent.resolve()
 PROJECT_ROOT = TESTS_ROOT.parent
@@ -672,12 +672,13 @@ def test_image_extraction(url, name):
     if not root.exists():
         root.mkdir()
 
-    for page in reader.pages:
-        for image in page.images:
-            filename = root / image.name
-            with open(filename, "wb") as img:
-                img.write(image.data)
-            images_extracted.append(filename)
+    with PILContext():
+        for page in reader.pages:
+            for image in page.images:
+                filename = root / image.name
+                with open(filename, "wb") as img:
+                    img.write(image.data)
+                images_extracted.append(filename)
 
     # Cleanup
     do_cleanup = True  # set this to False for manual inspection