Fix Buffer.toString encoding to match Node.js

[prydt]

Current implementation of Buffer.toString does not match Node.js behavior when using undefined as encoding.

Folks have run into this here: mariadb-corporation/mariadb-connector-nodejs#322 (comment)

[github-actions]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[prydt]

I have read the CLA Document and I hereby sign the CLA

[anonrig]

Can you add a test?

[prydt]

Can you add a test?

Added a test. My initial thought was to reuse normalizeEncoding but found that Buffer.toString should fail on null while streams which were using the same encoding function accept null as well (readable_setencoding_null), so I went back to using getEncodingOps but adding a special case for undefined.