fixed bugs in promise rejection tests

[dvisztempacct]

when you use Js.String.make() as your error transformer, i'm guessing that maps to javascripts' String() function, which means you'll get a string based on the specific JS representation that bucklescript has chosen for your error value.

these changes should work no matter how bsb decides to represent those values

[dvisztempacct]

To see the old tests break, just add another exception type to your project. It's probably pretty likely to break the tests.

[coveralls]

Pull Request Test Coverage Report for Build 22

• 0 of 0 changed or added relevant lines in 0 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage remained the same at 97.938%

---

Totals
Change from base Build 18:	0.0%
Covered Lines:	82
Relevant Lines:	83

---

💛 - Coveralls

[gilbert]

Thanks for the PR. On further thought, shouldn't we be testing JavaScript exceptions? I believe TestError is an ocaml exception, which has a different runtime representation.

[dvisztempacct]

@gilbert yes I think that would be a good idea. Once this is merged I'm going to write some

[gilbert]

Sounds good. Thank you for you contribution!