Merge consecutive whitespace

[LeopoldArkham]

Currently the parser ends whitespace whenever a newline is reached. When appending a WS item, check that it does not follow another WS item,and merge them together if it does.

Look for the todo in parse_item(), currently on line 135.

[jeremydavis519]

I'd like to take this one if you don't mind. I should have a working version soon. But just to be clear, when you refer to a "WS item," that means a line that contains nothing but whitespace, right? Nothing else seems to make sense, given the source code that I've read since discovering this project a few hours ago. (In particular, I thought at first that any whitespace surrounding a newline would count, but that's clearly not right.)

[LeopoldArkham]

Hey, great to have you!

Yes. I do specifically mean Item::WS elements should be concatenated into a single one if multiple of them occur consecutively.

Couple things I didn't mention in the OP:

• Ideally this would stay allocation-free. Meaning you can keep track of the start index of the last WS element and create a new slice if another WS element follow it. You could also go unsafe and use `slice::from_raw_parts if you feel like it.
• Ideally this would happen as the items are parsed, not as a second pass
• There should be a test for it, but it doesn't need to be in the first draft of the PR.

Let me know if you need anything!