ROB: Merge documents with named destinations with invalid page (#2857)

Closes #2842.

pypdf/_doc_common.py

@@ -855,7 +855,7 @@ def _get_outline(
                 # §12.3.3 Document outline, entries in the outline dictionary
                 if not is_null_or_none(lines) and "/First" in lines:
                     node = cast(DictionaryObject, lines["/First"])
-            self._namedDests = self._get_named_destinations()
+            self._named_destinations = self._get_named_destinations()
 
         if node is None:
             return outline
@@ -996,7 +996,7 @@ def _build_outline_item(self, node: DictionaryObject) -> Optional[Destination]:
             # TODO : keep named destination instead of replacing it ?
             try:
                 outline_item = self._build_destination(
-                    title, self._namedDests[dest].dest_array
+                    title, self._named_destinations[dest].dest_array
                 )
             except KeyError:
                 # named destination not found in Name Dict

pypdf/_writer.py

@@ -2699,10 +2699,13 @@ def merge(
                 position += 1
             srcpages[pg.indirect_reference.idnum].original_page = pg
 
-        reader._namedDests = (
+        reader._named_destinations = (
             reader.named_destinations
         )  # need for the outline processing below
-        for dest in reader._namedDests.values():
+
+        arr: Any
+
+        def _process_named_dests(dest: Any) -> None:
             arr = dest.dest_array
             if "/Names" in self._root_object and dest["/Title"] in cast(
                 List[Any],
@@ -2718,7 +2721,10 @@ def merge(
             elif isinstance(dest["/Page"], int):
                 # the page reference is a page number normally not a PDF Reference
                 # page numbers as int are normally accepted only in external goto
-                p = reader.pages[dest["/Page"]]
+                try:
+                    p = reader.pages[dest["/Page"]]
+                except IndexError:
+                    return
                 assert p.indirect_reference is not None
                 try:
                     arr[NumberObject(0)] = NumberObject(
@@ -2733,6 +2739,9 @@ def merge(
                 ].indirect_reference
                 self.add_named_destination_array(dest["/Title"], arr)
 
+        for dest in reader._named_destinations.values():
+            _process_named_dests(dest)
+
         outline_item_typ: TreeObject
         if outline_item is not None:
             outline_item_typ = cast(

tests/test_writer.py

@@ -2468,3 +2468,15 @@ def test_increment_writer(caplog):
     assert writer.metadata is None
     b = BytesIO()
     writer.write(b)
+
+
+@pytest.mark.enable_socket()
+def test_append_pdf_with_dest_without_page(caplog):
+    """Tests for #2842"""
+    url = "https://github.com/user-attachments/files/16990834/test.pdf"
+    name = "iss2842.pdf"
+    reader = PdfReader(BytesIO(get_data_from_url(url, name=name)))
+    writer = PdfWriter()
+    writer.append(reader)
+    assert "/__WKANCHOR_8" not in writer.named_destinations
+    assert len(writer.named_destinations) == 3