MAINT: Add _get_page_number_from_indirect in writer (#2493)

Create similar function to have same API as in reader. Used in future development.

Author: pubpub-zz

pypdf/_writer.py

@@ -424,6 +424,28 @@ def _get_num_pages(self) -> int:
         pages = cast(Dict[str, Any], self.get_object(self._pages))
         return int(pages[NameObject("/Count")])
 
+    def _get_page_number_by_indirect(
+        self, indirect_reference: Union[None, int, NullObject, IndirectObject]
+    ) -> Optional[int]:
+        """
+        Generate _page_id2num.
+
+        Args:
+            indirect_reference:
+
+        Returns:
+            The page number or None
+        """
+        # to provide same function as in PdfReader
+        if indirect_reference is None or isinstance(indirect_reference, NullObject):
+            return None
+        if isinstance(indirect_reference, int):
+            indirect_reference = IndirectObject(indirect_reference, 0, self)
+        obj = indirect_reference.get_object()
+        if isinstance(obj, PageObject):
+            return obj.page_number
+        return None
+
     @property
     def pages(self) -> List[PageObject]:
         """

tests/test_writer.py

@@ -2011,3 +2011,16 @@ def test_reattach_fields():
     assert isinstance(writer.pages[0]["/Annots"][0], IndirectObject)
     del writer.pages[1]["/Annots"]
     assert len(writer.reattach_fields(writer.pages[1])) == 0
+
+
+def test_get_pagenumber_from_indirectobject():
+    """Test test_get_pagenumber_from_indirectobject"""
+    pdf_path = RESOURCE_ROOT / "crazyones.pdf"
+    writer = PdfWriter(clone_from=pdf_path)
+    assert writer._get_page_number_by_indirect(None) is None
+    assert writer._get_page_number_by_indirect(NullObject()) is None
+
+    ind = writer.pages[0].indirect_reference
+    assert writer._get_page_number_by_indirect(ind) == 0
+    assert writer._get_page_number_by_indirect(ind.idnum) == 0
+    assert writer._get_page_number_by_indirect(ind.idnum + 1) is None