BUG: Avoid isolating the graphics state multiple times (fixes py-pdf#…

…2219)

pypdf/generic/_data_structures.py

@@ -1023,6 +1023,8 @@ def __init__(
  super().set_data(b_(stream_data))
  self.forced_encoding = forced_encoding
 
+ self._has_isolated_graphics_state: Optional[bool] = None
+
  def clone(
  self,
  pdf_dest: Any,
@@ -1229,12 +1231,31 @@ def operations(self, operations: List[Tuple[Any, Any]]) -> None:
  self._operations = operations
  self._data = b""
 
+ @property
+ def has_isolated_graphics_state(self) -> bool:
+ if self._has_isolated_graphics_state is None:
+ if self._operations:
+ self._has_isolated_graphics_state = self._operations[0] == "q" and self._operations[-1] == "Q"
+ elif self._data:
+ # Check for the character with the linebreak as inserted by `isolate_graphics_state`.
+ self._has_isolated_graphics_state = self._data[:2] == b"q\n" and self._data[-2:] == b"Q\n"
+ else:
+ # Empty stream.
+ self._has_isolated_graphics_state = True
+
+ return self._has_isolated_graphics_state
+
  def isolate_graphics_state(self) -> None:
+ if self.has_isolated_graphics_state:
+ # No need to isolate again.
+ return
+
  if self._operations:
  self._operations.insert(0, ([], "q"))
  self._operations.append(([], "Q"))
  elif self._data:
  self._data = b"q\n" + b_(self._data) + b"Q\n"
+ self._has_isolated_graphics_state = True
 
  # This overrides the parent method:
  def write_to_stream(

tests/test_page.py

@@ -1288,3 +1288,54 @@ def test_get_contents_from_nullobject():
  p = writer.add_blank_page(100, 100)
  p[NameObject("/Contents")] = writer._add_object(NullObject())
  p.get_contents()
+
+
+@pytest.mark.enable_socket()
+def test_has_isolated_graphics_state():
+ # Real example.
+ url = "https://github.com/py-pdf/pypdf/files/12428859/out1.pdf"
+ name = "isolate-graphics-state.pdf"
+ page = PdfReader(BytesIO(get_data_from_url(url, name=name))).pages[0]
+ content_stream = page.get_contents()
+ assert content_stream is not None
+
+ assert content_stream.has_isolated_graphics_state is False
+ content_stream.isolate_graphics_state()
+ assert content_stream.has_isolated_graphics_state is True
+
+ # Empty stream handling.
+ content_stream = ContentStream(stream=None, pdf="dummy.pdf")
+ assert content_stream.has_isolated_graphics_state is True
+
+ # Handling of string-based checks.
+ content_stream = ContentStream(stream=None, pdf="dummy.pdf")
+ content_stream.set_data(b"q\n 841.680 0 0 595.200 0.000 0.000 cm\n/Im0 Do\nQ\n\n \n")
+ assert content_stream.has_isolated_graphics_state is False
+
+ content_stream = ContentStream(stream=None, pdf="dummy.pdf")
+ content_stream.set_data(b"q\n 841.680 0 0 595.200 0.000 0.000 cm\n/Im0 Do\nQ\n")
+ assert content_stream.has_isolated_graphics_state is True
+
+ # Dummy example to test caching.
+ content_stream = ContentStream(stream=None, pdf="dummy.pdf")
+ assert content_stream._has_isolated_graphics_state is None
+ content_stream._has_isolated_graphics_state = True
+ assert content_stream.has_isolated_graphics_state is True
+ content_stream._has_isolated_graphics_state = False
+ assert content_stream.has_isolated_graphics_state is False
+
+
+@pytest.mark.enable_socket()
+def test_isolate_graphics_state():
+ url = "https://github.com/py-pdf/pypdf/files/12428859/out1.pdf"
+ name = "isolate-graphics-state.pdf"
+ page = PdfReader(BytesIO(get_data_from_url(url, name=name))).pages[0]
+ content_stream = page.get_contents()
+ assert content_stream is not None
+
+ # This page is not considered isolated at the beginning due to the final characters.
+ assert content_stream._data == b"q\n 841.680 0 0 595.200 0.000 0.000 cm\n/Im0 Do\nQ\n\n \n"
+ content_stream.isolate_graphics_state()
+ assert content_stream._data == b"q\nq\n 841.680 0 0 595.200 0.000 0.000 cm\n/Im0 Do\nQ\n\n \nQ\n"
+ content_stream.isolate_graphics_state()
+ assert content_stream._data == b"q\nq\n 841.680 0 0 595.200 0.000 0.000 cm\n/Im0 Do\nQ\n\n \nQ\n"