BUG: get_contents does not return ContentStream (#1847)

Closes #1846

pypdf/_page.py

@@ -701,7 +701,11 @@ def get_contents(self) -> Optional[ContentStream]:
  ``/Contents`` is optional, as described in PDF Reference 7.7.3.3
  """
  if PG.CONTENTS in self:
- return self[PG.CONTENTS].get_object() # type: ignore
+ try:
+ pdf = cast(IndirectObject, self.indirect_reference).pdf
+ except AttributeError:
+ pdf = None
+ return ContentStream(self[PG.CONTENTS].get_object(), pdf)
  else:
  return None
 
@@ -819,7 +823,6 @@ def _merge_page(
 
  page2content = page2.get_contents()
  if page2content is not None:
- page2content = ContentStream(page2content, self.pdf)
  rect = getattr(page2, MERGE_CROP_BOX)
  page2content.operations.insert(
  0,
@@ -955,7 +958,6 @@ def _merge_page_writer(
 
  page2content = page2.get_contents()
  if page2content is not None:
- page2content = ContentStream(page2content, self.pdf)
  rect = getattr(page2, MERGE_CROP_BOX)
  page2content.operations.insert(
  0,
@@ -1491,12 +1493,7 @@ def compress_content_streams(self) -> None:
  """
  content = self.get_contents()
  if content is not None:
- content_obj: Any
- if not isinstance(content, ContentStream):
- content_obj = ContentStream(content, self.pdf)
- else:
- content_obj = content
- content_obj = content_obj.flate_encode()
+ content_obj = content.flate_encode()
  try:
  content.indirect_reference.pdf._objects[ # type: ignore
  content.indirect_reference.idnum - 1 # type: ignore

tests/test_page.py

@@ -264,7 +264,9 @@ def test_compress_content_streams(pdf_path, password):
  writer = PdfWriter()
  if password:
  reader.decrypt(password)
+ assert isinstance(reader.pages[0].get_contents(), ContentStream)
  writer.clone_document_from_reader(reader)
+ assert isinstance(writer.pages[0].get_contents(), ContentStream)
  for page in writer.pages:
  page.compress_content_streams()
 
@@ -330,7 +332,10 @@ def test_page_scale():
 
 def test_add_transformation_on_page_without_contents():
  page = PageObject()
+ assert page.get_contents() is None
  page.add_transformation(Transformation())
+ page[NameObject("/Contents")] = ContentStream(None, None)
+ assert isinstance(page.get_contents(), ContentStream)
 
 
 @pytest.mark.enable_socket()