…xpression

As implied, this is basically a copy-paste of the functionality of expect. It's not complete, and it's not well-tested. But it gets the idea across.
Future work is cleaning this up, backfilling tests for features (such as the unwrap function), and adding support for polling requirements (i.e. toEventually).

Other than functionality, require also files the errorThrown issue type with XCTest, whereas expect files the assertionFailed issue type. This has minor differences that are mostly semantics.

Also, in addition to the require dsl, this also adds unwrap, which is a shorthand for `require(...).toNot(beNil())`.

Support async, but not yet polling.

Oh, gosh. This is terrible. I think a cleaner way to do this would be to change MatcherResult to also store the most recent value.
But that would be a breaking change, and might require updating existing matchers, so... no.

Rename requireAsync to requirea - matching the example set forth by expecta.
Adds requires, which is requirea, but for specifying SyncRequirement.

Update the table of contents with doctoc.

…corded as test assertion failures, which is what they are anyway.

Also actually mark the non-polling require dsl as discardableResult

Fixed test failure on linux