Merge branch 'main' into simplify-create-string

PyPDF2/_reader.py

@@ -691,6 +691,9 @@ def _get_outlines(
                     # so continue to load the file without the Bookmarks
                     return outlines
 
+                if isinstance(lines, NullObject):
+                    return outlines
+
                 # TABLE 8.3 Entries in the outline dictionary
                 if "/First" in lines:
                     node = cast(DictionaryObject, lines["/First"])
@@ -1052,6 +1055,10 @@ def get_object(self, indirect_reference: IndirectObject) -> Optional[PdfObject]:
             indirect_reference.generation in self.xref
             and indirect_reference.idnum in self.xref[indirect_reference.generation]
         ):
+            if self.xref_free_entry.get(indirect_reference.generation, {}).get(
+                indirect_reference.idnum, False
+            ):
+                return NullObject()
             start = self.xref[indirect_reference.generation][indirect_reference.idnum]
             self.stream.seek(start, 0)
             idnum, generation = self.read_object_header(self.stream)
@@ -1225,6 +1232,7 @@ def read(self, stream: StreamType) -> None:
 
         # read all cross reference tables and their trailers
         self.xref: Dict[int, Dict[Any, Any]] = {}
+        self.xref_free_entry: Dict[int, Dict[Any, Any]] = {}
         self.xref_objStm: Dict[int, Tuple[Any, Any]] = {}
         self.trailer = DictionaryObject()
         while True:
@@ -1380,9 +1388,12 @@ def _read_standard_xref_table(self, stream: StreamType) -> None:
                     stream.seek(-1, 1)
 
                 offset_b, generation_b = line[:16].split(b" ")
+                entry_type_b = line[17:18]
+
                 offset, generation = int(offset_b), int(generation_b)
                 if generation not in self.xref:
                     self.xref[generation] = {}
+                    self.xref_free_entry[generation] = {}
                 if num in self.xref[generation]:
                     # It really seems like we should allow the last
                     # xref table in the file to override previous
@@ -1391,6 +1402,7 @@ def _read_standard_xref_table(self, stream: StreamType) -> None:
                     pass
                 else:
                     self.xref[generation][num] = offset
+                    self.xref_free_entry[generation][num] = entry_type_b == b"f"
                 cnt += 1
                 num += 1
             read_non_whitespace(stream)

tests/test_reader.py

@@ -691,7 +691,7 @@ def test_extract_text_hello_world():
 
 
 def test_read_path():
-    path = Path(os.path.join(RESOURCE_ROOT, "crazyones.pdf"))
+    path = Path(RESOURCE_ROOT, "crazyones.pdf")
     reader = PdfReader(path)
     assert len(reader.pages) == 1
 

tests/test_workflows.py

@@ -325,3 +325,13 @@ def test_get_fields():
 
     # Cleanup
     os.remove("tmp.txt")
+
+
+def test_scale_rectangle_indirect_object():
+    url = "https://corpora.tika.apache.org/base/docs/govdocs1/999/999944.pdf"
+    name = "tika-999944.pdf"
+    data = BytesIO(get_pdf_from_url(url, name=name))
+    reader = PdfReader(data)
+
+    for page in reader.pages:
+        page.scale(sx=2, sy=3)

tox.ini

@@ -8,4 +8,4 @@ deps =
 	pytest
 	pytest-cov
 	pycryptodome
-commands = pytest tests --cov --cov-report term-missing -vv
+commands = pytest tests --cov --cov-report term-missing -vv {posargs}