Turns out there’s no need to speculate. Someone pointed out on GH [0] that the AI was literally prompted to copy the existing code: > *Context:* The registry m

Turns out there’s no need to speculate. Someone pointed out on GH [0] that the AI was literally prompted to copy the existing code: > *Context:* The registry maps every supported encoding to its metadata. Era assignments MUST match chardet 6.0.0's `chardet/metadata/charsets.py` at https://raw.githubusercontent.com/chardet/chardet/f0676c0d6a4263827924b78a62957547fca40052/chardet/metadata/charsets.py https://raw.githubusercontent.com/chardet/chardet/f0676c0d6a... > Fetch that file and use it as the authoritative reference for which encodings belong to which era. Do not invent era assignments. [0] https://github.com/chardet/chardet/issues/327#issuecomment-4003870261 https://github.com/chardet/chardet/issues/327#issuecomment-4...