BUG: Handle IndirectObject as image filter (#2355)

stefan6419846 · web-flow · commit 133ccb1c9154 · 2023-12-23T11:48:33.000+01:00
Previously, we might pass "4bits" as image mode to Pillow, leading to "ValueError: unrecognized image mode".

Example: `lfilters = IndirectObject(26, 0, 139771595681120)`, whose `get_object()` would yield `['/FlateDecode']` (going into the else branch of the filter handling until now).
diff --git a/pypdf/filters.py b/pypdf/filters.py
@@ -779,7 +779,7 @@ def _xobj_to_image(x_object_obj: Dict[str, Any]) -> Tuple[Optional[str], bytes,
         )
     extension = None
     alpha = None
-    filters = x_object_obj.get(SA.FILTER, [None])
+    filters = x_object_obj.get(SA.FILTER, NullObject()).get_object()
     lfilters = filters[-1] if isinstance(filters, list) else filters
     if lfilters in (FT.FLATE_DECODE, FT.RUN_LENGTH_DECODE):
         img, image_format, extension, _ = _handle_flate(

Original file line number	Diff line number	Diff line change
`@@ -779,7 +779,7 @@ def _xobj_to_image(x_object_obj: Dict[str, Any]) -> Tuple[Optional[str], bytes,`
`779`	`779`	`)`
`780`	`780`	`extension = None`
`781`	`781`	`alpha = None`
`782`		`- filters = x_object_obj.get(SA.FILTER, [None])`
	`782`	`+ filters = x_object_obj.get(SA.FILTER, NullObject()).get_object()`
`783`	`783`	`lfilters = filters[-1] if isinstance(filters, list) else filters`
`784`	`784`	`if lfilters in (FT.FLATE_DECODE, FT.RUN_LENGTH_DECODE):`
`785`	`785`	`img, image_format, extension, _ = _handle_flate(`