[huuan26]

Accept buffer inputs in \Codec.decode\ across \stdlib/encodings/cp1250.pyi\ through \cp1258.pyi, matching their runtime CPython implementation as well as \cp1252.pyi\ (updated in #16665).

In Python, standard library codec \decode\ methods accept any object implementing the buffer protocol (\ReadableBuffer\ like \�ytes, \�ytearray, \memoryview).

[github-actions]

According to mypy_primer, this change has no effect on the checked open source code. 🤖🎉

[srittau]

Do you have a link to the implementation in CPython to verify this?

[huuan26]

Here are the links to the CPython implementation:

1. Lib/encodings/cp1250.py – cp1258.py (generated via Tools/unicode/gencodec.py):
   Codec.decode delegates directly to codecs.charmap_decode(input, errors, decoding_table) (just like IncrementalDecoder.decode, which is already typed with ReadableBuffer in these stubs):

   • https://github.com/python/cpython/blob/main/Lib/encodings/cp1250.py#L14-L15
   • https://github.com/python/cpython/blob/main/Lib/encodings/cp1252.py#L14-L15 (note: cp1252.pyi in typeshed already uses ReadableBuffer for Codec.decode)

2. Modules/_codecsmodule.c (_codecs.charmap_decode):
   In Argument Clinic, charmap_decode takes data: Py_buffer, accepting any ReadableBuffer (�ytes, �ytearray, memoryview, etc.), which matches _codecs.charmap_decode(data: ReadableBuffer, ...) in stdlib/_codecs.pyi:

   • https://github.com/python/cpython/blob/main/Modules/_codecsmodule.c#L351-L368