Added 7 new CSS4Pseudo tests and 3 references

[TalbotG]

cascade-highlight-001.html
cascade-highlight-001-ref.html
cascade-highlight-002.html
cascade-highlight-004.html
cascade-highlight-004-ref.html
grammar-error-002-manual.html
grammar-error-003-manual.html
highlight-z-index-001.html
highlight-z-index-001-ref.html
highlight-z-index-002.html

[fantasai]

The rest of the tests look correct, however I think the z-index tests need some changes:

• The descendant comes before the selected text (makes the test stronger, since usually text after paints over anyway).
• The descendant's text is not highlighted. (Currently you are highlighting the entire contents of #test, including the positioned descendant.)

[TalbotG]

the z-index tests need some changes:

http://www.gtalbot.org/BrowserBugsSection/CSS4Pseudo/highlight-z-index-001-new.html

http://www.gtalbot.org/BrowserBugsSection/CSS4Pseudo/highlight-z-index-001-ref-new.html

http://www.gtalbot.org/BrowserBugsSection/CSS4Pseudo/highlight-z-index-002-new.html

http://www.gtalbot.org/BrowserBugsSection/CSS4Pseudo/highlight-z-index-002-ref-new.html

[fantasai]

Well, now you say you're testing a positioned descendant, but you don't have a descendant. :) I'm not sure which is better, to have a descendant or to have a previous sibling, but either way the comment and the code should match.

I might also suggest combining both tests into one file. They're very closely related, almost all the code is duplicated, so I think it will be easier to understand if they are in one file.

[TalbotG]

Well, now you say you're testing a positioned descendant, but you don't have a descendant. :) I'm not sure which is better, to have a descendant or to have a previous sibling, but either way the comment and the code should match.

The statement that I was trying to test is:

Each highlight pseudo-element draws its background over the corresponding portion of the highlight overlay, painting it immediately below any positioned descendants

coming from § 3.5 Painting the Highlight

At first, such statement was not (to me) perfectly explicit or clear since a descendant could refer to the body element (as the referring ancestor) or to the highlighted element itself (as the referring ancestor). Anyway... So, I created 2 tests (z-index-001 involving 'position: relative' and z-index-002 involving 'position: absolute'). And now, I have tuned and tweaked accordingly the css identifiers and the meta assert text of those 2 tests at my website:

http://www.gtalbot.org/BrowserBugsSection/CSS4Pseudo/highlight-z-index-001-new.html

http://www.gtalbot.org/BrowserBugsSection/CSS4Pseudo/highlight-z-index-002-new.html

I might also suggest combining both tests into one file. They're very closely related, almost all the code is duplicated, so I think it will be easier to understand if they are in one file.

I think it would be difficult and not best to combine both tests (abs pos and rel pos) into 1 ... but it's doable. I prefer not to combine both tests into 1.

In a few min., I will try to make the changes into
highlight-z-index-001.html
highlight-z-index-002.html
in this PR. If it fails because of Community-TC integration check and the need to rebase, then I will just close this PR and create a new PR. So, stay tuned ...

[TalbotG]

@fantasai , the modifications I did to the PR went as hoped. So, please review the z-index tests one last time. Thank you!