[TASK] Add tests for the selector specificity

[oliverklee]

Mostly move tests from ParserTest that belong into
the unit tests.

Part of #757
Part of #758

[coveralls]

coverage: 55.753% (+0.5%) from 55.282%
when pulling efe6034 on task/tests/selector-1
into 19ffd07 on main.

[JakeQZ]

I think there's a naming issue. MDN does indeed say:

In CSS, selectors are patterns used to match, or select, the elements you want to style...

"pattern" is a description of how they work. "selector" is the (official) name for what they are. I think the test names should reflect what the 'object' is called (or known as), rather than what it does or how it works.

Wittgenstein might disagree (if I sit on a table, it's now a chair), but he wasn't a computer scientist.

Also, I'd like this PR to be split for more manageable review - the addition of the deprecated test class can be done separately I think.

[oliverklee]

I'd like to find a way around that a selector can have a selector (which would mean that we need to deal with a selector's selector) … any ideas?

[JakeQZ]

I'd like to find a way around that a selector can have a selector (which would mean that we need to deal with a selector's selector) … any ideas?

The Selector class represents and/or encapsulates a selector, which in the rendered or to-be-parsed CSS is just a string like a:hover. I guess the string property of the class could be named something like $rawContent or $stringRepresentation, or something better that I can't come up with right now. Note we alse have Value.value, and others (e.g. Rule.rule).

[oliverklee]

Okay, then I'll keep using selector for the tests for now, and we do the renaming in #1016. I'll also split this up into smaller chunks.

[oliverklee]

The other two PRs are merged so that now this PR is a lot smaller. Rebased, reworded and repushed.

[JakeQZ]

There's a typo in a test method name.

Specificites cannot be represented by a single number, so we should address that in future...

[JakeQZ]

Specificites cannot be represented by a single number, so we should address that in future...

Added #1042 to capture this.