Hamming: Add a test case to avoid wrong recursion solution

[tqa236]

This solution will pass all the current test cases but fail on the new one:

module Hamming (distance) where


distance :: String -> String -> Maybe Int
distance [] []         = Just 0
distance (x:xs) (y:ys)
  | length(xs) /= length(ys) = Nothing
  | x /= y = fmap (1 + ) (distance xs ys)
  | x == y = distance xs ys

[sshine]

Thanks a lot for your pull request!

And you are quite right, this test case is missing.

If you don't mind, I would like to propagate this test case to the central Exercism repository for test cases shared across language tracks, since I imagine this case may trigger at least other functional languages.

Before this can happen, I just noticed that the Hamming exercise is one version bump behind on the Haskell track. I've bumped the Hamming test suite version to 2.2.0 in #797, so that you don't also have to fix this.

1. Fork the repo exercism/problem-specifications
2. In that repo, make an update to the file exercises/hamming/canonical-data.json: Add this test case and increase "version": "2.2.0" to "version": "2.3.0". Send a PR there and reference this PR by writing exercism/haskell#796 in the message.
3. In this repo, rebase your current feature branch on upstream now that I've merged Hamming: Bump test suite to 2.2.0 #797. (Let me know if you want hints on how to do that.)
4. In this repo, bump the version number in the file exercises/hamming/package.yaml from version: 2.2.0.9 to version: 2.3.0.10. The first three numbers are derived from the version in canonical-data, and the fourth is our track's incremental pointer.

I'll merge this PR once the canonical data is merged and the version in package.yaml reflects 2.3.0.10. :-)

[tqa236]

Hello,

I did all 4 points you said and request a new pull request. Please tell me if you notice anything wrong.

[tqa236]

@sshine Is this PR ok to merge now or do we still need some changes?

[sshine]

@tqa236: We're good. Thanks a lot, and sorry for the late response.