Wrong charset detect!

[NB-Dragon]

Reproduce

>>> chardet.detect_all("中".encode("gb2312"))
[{'encoding': 'ISO-8859-5', 'confidence': 0.99, 'language': 'Russian'}, {'encoding': 'ISO-8859-1', 'confidence': 0.365, 'language': ''}]

Environment

Python 3.7.2
chardet 4.0.0

[NB-Dragon]

谷歌浏览器可以正常识别gb2312的编码，不知道为何。

$filename = mb_convert_encoding("中", "gb2312");
header("content-disposition:attachment; filename=$filename");

[NB-Dragon]

And now, I have fix it in AdvancedDownloader.
Welcome to learn and turn on the star, thanks.

[samamorgan]

I'm seeing a similar, but opposite issue. Chardet 3.0.4 correctly detected a file as "Windows-1252" (a superset of ISO 8859-1), but 4.0.0 detects gb2312, which is clearly incorrect given the file contents I'm reading. Apologies, I can't share here, this is confidential data.