PI: Check duplicate objects in writer._sweep_indirect_references

PyPDF2/_writer.py

@@ -92,6 +92,7 @@ class (typically :class:`PdfReader<PyPDF2.PdfReader>`).
     def __init__(self) -> None:
         self._header = b"%PDF-1.3"
         self._objects: List[Optional[PdfObject]] = []  # array of indirect objects
+        self._idnum_hash: Dict[bytes, int] = {}
 
         # The root of our page tree node.
         pages = DictionaryObject()
@@ -868,8 +869,16 @@ def _sweep_indirect_references(
                 if newobj is None:
                     try:
                         newobj = data.pdf.get_object(data)
+                        hash_value = None
+                        if newobj is not None:
+                            hash_value = newobj.hash_value()
+                        # Check if object is already added to pdf.
+                        if hash_value in self._idnum_hash:
+                            return IndirectObject(self._idnum_hash[hash_value], 0, self)
                         self._objects.append(None)  # placeholder
                         idnum = len(self._objects)
+                        if hash_value is not None:
+                            self._idnum_hash[hash_value] = idnum
                         newobj_ido = IndirectObject(idnum, 0, self)
                         if data.pdf not in extern_map:
                             extern_map[data.pdf] = {}

PyPDF2/generic.py

@@ -34,11 +34,12 @@
 
 import codecs
 import decimal
+import hashlib
 import logging
 import re
 import warnings
 from io import BytesIO
-from typing import Any, Dict, List, Optional, Tuple, Union, cast
+from typing import Any, Callable, Dict, List, Optional, Tuple, Union, cast
 
 from ._codecs import (  # noqa: rev_encoding
     _pdfdoc_encoding,
@@ -77,6 +78,21 @@
 
 
 class PdfObject:
+    # function for calculating a hash value
+    hash_func: Callable[..., "hashlib._Hash"] = hashlib.sha1
+
+    def hash_value_data(self) -> bytes:
+        return ("%s" % self).encode()
+
+    def hash_value(self) -> bytes:
+        return (
+            "%s:%s"
+            % (
+                self.__class__.__name__,
+                self.hash_func(self.hash_value_data()).hexdigest(),
+            )
+        ).encode()
+
     def get_object(self) -> Optional["PdfObject"]:
         """Resolve indirect references."""
         return self
@@ -1001,6 +1017,9 @@ def __init__(self) -> None:
         self.__data: Optional[str] = None
         self.decoded_self: Optional[DecodedStreamObject] = None
 
+    def hash_value_data(self) -> bytes:
+        return b_(self._data)
+
     @property
     def decodedSelf(self) -> Optional["DecodedStreamObject"]:  # pragma: no cover
         deprecate_with_replacement("decodedSelf", "decoded_self")