non-breaking space causes parse-domain to fail

[AurielleP]

ParseDomain('\xa0')
TypeError: Cannot read property '3' of null
at parseDomain (/node_modules/parse-domain/lib/parseDomain.js:58:22)

related to #32

[AurielleP]

new lines also throw this error

ParseDomain('\n\nhttp://somewebsite.com/privacy-statement/')
TypeError: Cannot read property '3' of null
at parseDomain (/node_modules/parse-domain/lib/parseDomain.js:48:22)

[kulinsj]

In the case of line breaks in the url, the normalizedUrl is not falsy, so this referenced PR is not sufficient to prevent the error. It would probably be more prudent to sanitize the string in the normalize.js file. I'm now pre-sanitizing the urls i want to parse before handing them to the parse-domain function
cleanedUrl = url.replace(/(\r\n|\n|\r)/gm, '');
With all that said, it's still good practice to do a null check on the urlSplit in case there are other unforseen cases where the regex won't match

[jhnns]

We delegate all the URL parsing logic to the WHATWG URL constructor in a major version update. URL parsing is too complicated. See #14

[jhnns]

In the case of line breaks in the url, the normalizedUrl is not falsy, so this referenced PR is not sufficient to prevent the error

normalize.js uses .trim() which removes line breaks. A falsy check should be sufficient after that. So I'm going to merge and publish this PR as a fix.

[jhnns]

Fixed with v2.1.8