Potential optimization for charsetFromStream (yes, I've read the wiki)

[Rudxain]

I know I should do benchmarks, but my suggestion really seems to be fast.

Original code:
https://github.com/gorhill/uBlock/blob/8629f07138749e7c6088fbfda84a381f2cd3bc66/src/js/traffic.js#L748-L760

My suggestion (warning: diff link will break):

// would this be faster?:
// https://en.wikipedia.org/wiki/Boyer%E2%80%93Moore_string-search_algorithm
let i = 0;
while ( i < (65536 - 39) ) {
    // every pattern-char is unique,
    // so we can skip multiple offsets
    /* c */ if ( bytes[i+0] !== 0x63 ) { i += 1; continue; }
    /* h */ if ( bytes[i+1] !== 0x68 ) { i += 1; continue; }
    /* a */ if ( bytes[i+2] !== 0x61 ) { i += 2; continue; }
    /* r */ if ( bytes[i+3] !== 0x72 ) { i += 3; continue; }
    /* s */ if ( bytes[i+4] !== 0x73 ) { i += 4; continue; }
    /* e */ if ( bytes[i+5] !== 0x65 ) { i += 5; continue; }
    /* t */ if ( bytes[i+6] !== 0x74 ) { i += 6; continue; }
    break;
}
if ( i >= (65536 - 40) ) { return; }

I haven't tested if it has bugs. If anyone from the community wants to test and benchmark it, please do, the maintainers have limited time