Subsequent-sibling combinator (~) support

[nfriedly]

Hi, thanks for this library!

I'd like to request support for the subsequent-sibling combinator (~).

It's similar to the next sibling combinator (+), except it selects all of the sibling elements that match the right side and come after the left side match, not just the first sibling after a left-side match.

For example, if you have:

<p>one</p>
<img src="foo">
<p>two</p>
<p>three</p>

The selector img + p would match the "two" paragraph (I believe CssXPath already handles this correctly).

Then the selector img ~ p would match the "two" and "three" paragraphs. (Currently CssXPath fails to match anything.)

See https://developer.mozilla.org/en-US/docs/Learn_web_development/Core/Styling_basics/Combinators#subsequent-sibling_combinator for more info.

I'd like to use this in FreshRSS to filter out a bunch of junk that gets shoved into the end of every article in one of my feeds.

(I'd really like support for the :contains("Your ad here!") pseudoclass, but I figured this one was an actual standard, so it'd be a better request to start with :)

[Alkarex]

@nfriedly Tests welcome #231