Depends on #77. This PR only contains 2674de3980cb6bd4beb9825f5d926e94ceac2572.
Currently, in non ascii mode, we implement case insensitive matching by case folding. The spec makes a difference between unicode mode and non unicode mode. In non unicode mode the toUppercase algorithm is used instead of case folding (see https://tc39.es/ecma262/#sec-runtime-semantics-canonicalize-ch).
This PR adds the relevant unicode table for toUppercase and adjusts the folding and unfolding functions. I also added a test from 262 that checks this behaviour.
Depends on #77. This PR only contains 2674de3980cb6bd4beb9825f5d926e94ceac2572.
Currently, in non ascii mode, we implement case insensitive matching by case folding. The spec makes a difference between unicode mode and non unicode mode. In non unicode mode the toUppercase algorithm is used instead of case folding (see https://tc39.es/ecma262/#sec-runtime-semantics-canonicalize-ch).
This PR adds the relevant unicode table for toUppercase and adjusts the folding and unfolding functions. I also added a test from 262 that checks this behaviour.