strict name object

pypdf/generic/_base.py

@@ -648,6 +648,8 @@ def read_from_stream(stream: StreamType, pdf: Any) -> "NameObject":  # PdfReader
         if name != NameObject.surfix:
             raise PdfReadError("name read error")
         name += read_until_regex(stream, NameObject.delimiter_pattern)
+        if pdf and pdf.strict and not name.isascii():
+            raise PdfReadError(f"Illegal character in NameObject ({name})")
         name = NameObject.unnumber(name)
         try:
             name = NameObject(name.decode(NameObject.encoding))

tests/test_generic.py

@@ -1076,20 +1076,15 @@ def test_checkboxradiobuttonattributes_opt():
 
 
 def test_name_object_invalid_decode():
-    charsets = deepcopy(NameObject.CHARSETS)
-    try:
-        NameObject.CHARSETS = ("utf-8",)
-        stream = BytesIO(b"/\x80\x02\x03")
-        # strict:
-        with pytest.raises(PdfReadError) as exc:
-            NameObject.read_from_stream(stream, ReaderDummy(strict=True))
-        assert "Illegal character in NameObject " in exc.value.args[0]
-
-        # non-strict:
-        stream.seek(0)
-        NameObject.read_from_stream(stream, ReaderDummy(strict=False))
-    finally:
-        NameObject.CHARSETS = charsets
+    stream = BytesIO(b"/\x80\x02\x03")
+    # strict:
+    with pytest.raises(PdfReadError) as exc:
+        NameObject.read_from_stream(stream, ReaderDummy(strict=True))
+    assert "Illegal character in NameObject " in exc.value.args[0]
+
+    # non-strict:
+    stream.seek(0)
+    NameObject.read_from_stream(stream, ReaderDummy(strict=False))
 
 
 def test_indirect_object_invalid_read():