Incorrect encoding for cyrillic symbols

[fourslashw]

When I trying to debug socket.io events addon shows event data in incorrect encoding

42["TURN_SWITCHED",{"turn":{"id":0,"number":0,"attackerSide":"NORTH","state":"ENDED","defenderId":null,"attackedField":[1,1],"targetOwnership":"NORTH","currentAttack":{"question":{"text":"�акое ме��о зан�ла па��и� ��езиден�а на в�бо�а� в �о��да���венн�� ��м� в 1995 год�?","answers":["I","II","III*","IV"]}}}}]

[eliihen]

Hi,

Thanks for reporting. As you may have guessed, we don't use cyrillic much, so we're always happy to have people who do test the extension and report bugs :)

I'll have a look at this when I have time, which probably is either today or sometime next week. Stay tuned!

[eliihen]

I can see that where we receive frames the received frame needs to be utf8 decoded twice in order to get a meaningful string.

@janodvarko is this a bug in the API or do we need to decode the frame twice for some reason?

STR:

1. Place this line at the beginning of frameReceived in view.js console.log(decodeURIComponent(escape(decodeURIComponent(escape(frame)))));
2. Build and run with jpm
3. Open http://janodvarko.cz/test/websockets/ and the browser console
4. Connect and send это or anything cyrillic
5. In the browser console you see "это" printed because of the double decode console log, but the single decode we do in view.js only results in some binary stuff, as seen in the extension output