[WAIT] Issue 14: Add Request to ResponsePredicate

haskell-servant / servant-quickcheck

40 stars 21 forks source link

[WAIT] Issue 14: Add Request to ResponsePredicate #36

Closed erewok closed 6 years ago

erewok commented 6 years ago

Description

This PR does the following:

Adds the Request to ResponsePredicate so that PredicateFailure can pretty print the request as well as the failing response.
Removes the Maybe from PredicateFailure (Maybe Request) now that both RequestFailure and ResponseFailure require the request.
Adds tests for the following predicates: not500 and unauthorizedContainsWWWAuthenticate

Fixes #14

coveralls commented 6 years ago

Coverage increased (+0.3%) to 43.284% when pulling 064b6601a7867c524ff28a6fd2d168ac67eeae35 on issue-14/req_in_resp_preds into 54a05a53a9f227bd147298e53eac50371098f220 on master.

coveralls commented 6 years ago

Coverage decreased (-0.1%) to 43.284% when pulling d4f6aa527252e74f20063efa9745409547399dc1 on issue-14/req_in_resp_preds into d65abc856f6708558d96fa8aa1973d578bc85dbe on master.

erewok commented 6 years ago

@jkarni I think this one's ready, I just need a bit of help thinking through any broader implications (CHANGELOG, etc.).

erewok commented 6 years ago

I have been thinking about this PR and I think I'd like to withdraw it. I have some thoughts that I'd like to explore a bit before coming up with a suggestion for representing predicates based on Request + Response versus predicates based only on a Response.

I have been thinking about this in light of this comment from the Predicates module:

-- The idea with all this footwork is to not waste any requests. Rather than
-- generating new requests and only applying one predicate to the response, we
-- apply as many predicates as possible.

While I have some ideas on representing these things, this part is kind of stumping me.