Character escaping inside selectors

[SuggonM]

Using the online demo, I'm getting this error in browser console when trying to parse a rule with special characters escaped:

.wendy\'s {}

parse.js:62 Uncaught 
Error: Unmatched ' (line 1, char 8)
    at parseError (parse.js:62:15)
    at Object.parse (parse.js:110:6)
    at outputUpdated (parse.js:151:19)
    at HTMLTextAreaElement.changed (parse.js:185:2)

which is equivalent to this block:

CSSOM/lib/parse.js

Lines 106 to 113 in 64dfbde

	case "'":
	index = i + 1;
	do {
	index = token.indexOf("'", index) + 1;
	if (!index) {
	parseError("Unmatched '");
	}
	} while (token[index - 2] === '\\');

Another minor issue is that, round braces and many other invalid-but-escapable selectors remain unchanged when using toString(). Perhaps we could somehow ensure that the returned selector of toString() method is always valid?

.wendy(s) {}

// current: .wendy(s) {}
// suggestion: .wendy\(s\) {}
CSSOM.parse('.wendy(s) {}').toString();

---

I'm trying to parse a fairly large CSS spritesheet that has a lot of these characters, so fixing this would be of huge help!

Another suggestion that may or may not be relevant to this repository is, using a CSS escaping module. Mathias has made one in cssesc and it addresses both cases mentioned above, for example here and here.

[xemlock]

Same problem is present in the original repo, see NV#118.