Implement Resistor Colors exercise

[tejasbubane]

Problem Specification: exercism/problem-specifications#1466

[tejasbubane]

I keep getting this segmentation fault while running the ensure-readmes-updated script:

➜ bin/ensure-readmes-are-updated.sh
Cloning into 'problem-specifications'...
remote: Enumerating objects: 6, done.
remote: Counting objects: 100% (6/6), done.
remote: Compressing objects: 100% (6/6), done.
remote: Total 8462 (delta 0), reused 6 (delta 0), pack-reused 8456
Receiving objects: 100% (8462/8462), 1.70 MiB | 493.00 KiB/s, done.
Resolving deltas: 100% (4986/4986), done.
Checking readme for resistor-colors
bin/ensure-readmes-are-updated.sh: line 20:  1144 Segmentation fault: 11  $configlet generate . --only "$exercise" --spec-path "problem-specifications"

PS: It runs fine on master, but fails on this branch.

[petertseng]

Since I couldn't find output containing segmentation fault in the Travis CI output, I am taking your word for it.
Since the output you have displayed says that configlet is segfaulting, it sounds like the thing to do is to find a minimal set of conditions to reproduce this and then report it to https://github.com/exercism/configlet/issues

Since I don't use configlet I regret being unable to assist in the specifics of that

[sshine]

I think we need some kind of error handling built in.

I like this exercise over allergies for the reason that you don't need exponentiation, so it's less mathy.

If convert Black == 0, then what is value []? (This is perhaps a question better asked on problem-specifications.)

[sshine]

Since we're converting ["black", "brown"], I think it would make sense to

1. Have error handling.
2. Return an Integer.

Since it makes sense to have error handling, it also makes sense to ask if we should go with Maybe or Either.

Then readMaybe or readEither from Text.Read can come into use.

[tejasbubane]

From the tests in specification, I assumed the input will always be valid. Let me know if I should make this change or wait for more discussion (in the problem spec as well).

[sshine]

bin/ensure-readmes-are-updated.sh: line 20:  1144 Segmentation fault: 11  $configlet generate . --only "$exercise" --spec-path "problem-specifications"

That's strange, line 20 only contains git, grep, cut and sort.

configlet isn't run until line 28.

What if you echo the line that begins with $configlet and run that manually?

It sounds like configlet fails because of the way it's compiled...

[petertseng]

I notice that it would be possible to use http://hackage.haskell.org/package/base-4.12.0.0/docs/Prelude.html#t:Enum here.

I have not yet determined whether it would be a good idea to do so.

[sshine]

I had the same thought.

What you don't get with deriving Enum is an explicit annotation that Black is 0, etc.

Making an explicit instance Enum where ... has the drawback that you need to specify both toEnum and fromEnum, even though the exercise only needs the one. There are ways to avoid that, but the first way (using a dynamic lookup table) should be discouraged, and the second way (using Data and Typeable) is probably a little too complicated to serve as a good example.

So I'm content with having a simple convert function.

[emcoding]

I noticed that the Readme here is an older version of the Readme we actually implemented.

[emcoding]

@tejasbubane Can you regenerate the readme with
configlet regenerate -o resistor-colors? what Erik posted below?

(thanks/cc @ErikSchierboom )

[ErikSchierboom]

Oh, my mistake. That command should actually be: ./bin/configlet generate . -o resistor-colors
And that assumes that there is an up-to-date problem-specifications repository located at ../problem-specifications

[tejasbubane]

Regarding the configlet segfault issue - Sorry that was my bad. I had an old version (which was segfaulting) lying in my $PATH and that was picked up instead of bin/configlet:

haskell/bin/ensure-readmes-are-updated.sh

Lines 2 to 5 in 189f262

	configlet=$(which configlet)
	if [ ! -x $configlet ]; then
	configlet=bin/configlet
	fi

I have updated the README and the ensure script works fine now.

[sshine]

Travis CI reports a type error in the stub:

/home/travis/build/exercism/haskell/build/resistor-colors/stub/test/Tests.hs:20:29: error:
    • Couldn't match type ‘Color’ with ‘[Char]’
      Expected type: [String]
        Actual type: [Color]
    • In the first argument of ‘value’, namely ‘input’
      In the first argument of ‘shouldBe’, namely ‘value input’
      In the expression: value input `shouldBe` expected
   |
20 |         assertion   = value input `shouldBe` expected
   |                             ^^^^^

Other than that, I have no further comments before hitting approve.