Issues with encodings used in the Baltics

[adbar]

Describe the bug

• ISO-8859-15 detected correctly by chardet and cchardet, charset_normalizer says cp850,cp857,cp858
• WINDOWS-1252 detected correctly by chardet and cchardet, charset_normalizer says cp775,cp850,cp857,cp858

The language detection works in both cases (Estonian).

To Reproduce

• https://gitlab.freedesktop.org/uchardet/uchardet/-/blob/master/test/et/iso-8859-15.txt
• https://gitlab.freedesktop.org/uchardet/uchardet/-/blob/master/test/et/windows-1252.txt

Desktop (please complete the following information):

https://charsetnormalizerweb-ousret.vercel.app/

[Ousret]

I have taken a quick look over these.
Still borderline cases, looking at explain=True, it's a close call. The engine is already in a sweet spot, improving it further is difficult and challenging.

It's possible to fix those, but I lack the free time to make the fine-tuning for now.

[adbar]

I understand, but it's still something that other packages manage to handle, so if you find some time in the future...

[Ousret]

I have pushed/merged a solid patch for one of the two cases you presented me.

For the remaining file, I could not find a viable solution for now. It still produces some very readable Estonian.
https://gitlab.freedesktop.org/uchardet/uchardet/-/blob/master/test/et/iso-8859-15.txt

Feel free to check against dev-master@2.0.8.dev1

[adbar]

Nice, thanks!