Using indexOf() instead of Regexp()

[gorhill]

Performance improvement seen at (along with the intermediate experiments using binary search):
http://jsperf.com/uri-js-sld-regex-vs-binary-search/4

Memory footprint on my 32-bit OS:
Regexp: > 500K
indexOf: < 20K

A test has been added to ensure all SLD are correctly processed by the new code.

[gorhill]

Moving on.

[rodneyrehm]

Why did you close the PR? I've only not merged this PR because I wasn't sure if I'd want to go with 1.13 or 2.0 and fix some other things as well. Meanwhile it's clear that 1.13 will have to come first, as 2.0 is a longer way out. Please reopen the PR.

[gorhill]

I was wondering if you would be interested in the ability to support the full Public Suffix List? I think it's not too much work to fit in there publicsuffixlist.js.

User could optionally supply his own list, while the default (stock) list (if no external list supplied) would be a match of what SecondLevelDomain currently handles.

[rodneyrehm]

I would prefer using PublicSuffix as it used by browsers and maintained properly. I haven't used PS because it also contains entries like github.io, which are not SLDs, but used by the browser for cookie origin control.

If your list filters these out, we can talk about depending on or integrating your implementation (or a derivation thereof). If it doesn't, don't see how this could work.

btw. this should probably be a separate issue. care to open one?