Vendored version of chardet is buggy, can't detect utf-8 with bom #2296
Description
Here's the commit that fixed it in upstream.
Because of the bug, if I fetch a UTF-8 file that starts with a BOM, I get (e.g.):
<?xml version="1.0" encoding="utf-8"?>
...
and response.encoding is ISO-8859-1 or -2.
I think you should be able to just pull the newest version, and it will be fixed.