ENH : automatic access to pointed object

alternative solution to py-pdf#2460
fixes py-pdf#2287

pypdf/_utils.py

@@ -362,21 +362,11 @@ def b_(s: Union[str, bytes]) -> bytes:
         return r
 
 
-@overload
-def str_(b: str) -> str:
-    ...
-
-
-@overload
-def str_(b: bytes) -> str:
-    ...
-
-
-def str_(b: Union[str, bytes]) -> str:
+def str_(b: Any) -> str:
     if isinstance(b, bytes):
         return b.decode("latin-1")
     else:
-        return b
+        return b.__str__()
 
 
 @overload

pypdf/generic/_base.py

@@ -286,6 +286,28 @@ def get_object(self) -> Optional["PdfObject"]:
             return None
         return obj.get_object()
 
+    def __deepcopy__(self, memo: Any) -> "IndirectObject":
+        return IndirectObject(self.idnum, self.generation, self.pdf)
+
+    def __getattr__(self, name: str) -> Any:
+        # Attribute not found in object: look in pointed object
+        if name == "pdf" or name not in dir(self.get_object()):
+            raise AttributeError  # xreturn lambda:None
+        try:
+            return getattr(self.get_object(), name)
+        except AttributeError:
+            raise AttributeError(
+                f"No attribute {name} found in IndirectObject or pointed object"
+            )
+
+    def __getitem__(self, key: Any) -> Any:
+        # items should be extracted from pointed Object
+        return self.get_object()[key]  # type: ignore
+
+    def __str__(self) -> str:
+        # in this case we are looking for the pointed data
+        return self.get_object().__str__()
+
     def __repr__(self) -> str:
         return f"IndirectObject({self.idnum!r}, {self.generation!r}, {id(self.pdf)})"
 

tests/test_generic.py

@@ -1235,3 +1235,25 @@ def test_encodedstream_set_data():
     assert cc["/Filter"] == ["/FlateDecode", "/FlateDecode", "/FlateDecode"]
     assert str(cc["/DecodeParms"]) == "[NullObject, NullObject, NullObject]"
     assert cc[NameObject("/Test")] == "/MyTest"
+
+
+def test_calling_indirect_objects():
+    """Cope with cases where attributes/items are called from indirectObject"""
+    url = (
+        "https://raw.githubusercontent.com/xrkk/tmpppppp/main/"
+        "2021%20----%20book%20-%20Security%20of%20biquitous%20Computing%20Systems.pdf"
+    )
+    name = "2021_book_security.pdf"
+    reader = PdfReader(BytesIO(get_data_from_url(url, name=name)))
+    reader.trailer.get("/Info")["/Creator"]
+    reader.pages[0]["/Contents"][0].get_data()
+
+
+@pytest.mark.enable_socket()
+def test_indirect_object_page_dimensions():
+    url = "https://github.com/py-pdf/pypdf/files/13302338/Zymeworks_Corporate.Presentation_FINAL1101.pdf.pdf"
+    name = "issue2287.pdf"
+    data = BytesIO(get_data_from_url(url, name=name))
+    reader = PdfReader(data, strict=False)
+    mediabox = reader.pages[0].mediabox
+    assert mediabox == RectangleObject((0, 0, 792, 612))

tests/test_writer.py

@@ -914,8 +914,9 @@ def test_write_dict_stream_object(pdf_file_path):
 
     for k, v in page_object.items():
         if k == "/Test":
-            assert str(v) != str(stream_object)
+            assert repr(v) != repr(stream_object)
             assert isinstance(v, IndirectObject)
+            assert str(v) == str(stream_object)  # expansion of IndirectObjects
             assert str(v.get_object()) == str(stream_object)
             break
     else: