Test vectors refactor

[amika-sq]

Previously, test vectors lived alongside normal unit tests. This caused a few pain points:

1. Test classes did not follow the naming conventions defined in the sdk-development test vector spec.
   • e.g. Test classes must be prefixed with Web5TestVectors.
2. It was difficult to understand which tests represented a test vector vs which tests represented a unit test.
   • e.g. Does test_resolve run the resolve test vector, or does it run a unit test on the resolve function? It was impossible to tell without intimate knowledge of the test itself, or the use of rigorous naming conventions of the tests themselves (which isn't possible, because the test names have their own naming convention according to the spec).
3. The test result output didn't break down each test vector. When one test vector failed, it gave no indication as to which one of the test vectors in the file failed.
   • e.g.: did_jwk/resolve has three separate vectors defined within the resource JSON file. If one vector failed, but the other two succeeded, the test would fail and the output gave no indication as to which of the vectors failed. The developer would need to manually inspect the test with breakpoints in order to determine which one it was. This was time consuming, and just overall a bad experience.

To address these pain points, I have refactored how test vectors work in this package.

Changes made:

1. All test vectors are now defined in a separate testing target called Web5TestVectors.
2. All test vectors are now named with the appropriate naming conventions.
3. All common utility code shared between unit tests & test vectors are defined in a new TestUtilities target.
4. When running a test vector, each vector is run within its own named XCTActivity, allowing us to know which of the vectors actually failed.