Port test suite so it supports Data.Text

[sshine]

In #780, exercises that might be applicable for teaching the Data.Text module were found. In some of those cases, the test suite was extended with OverloadedStrings so that both String and Text solutions work, and an example solution was provided. For a notable example of this, one of the early exercises, Bob, had an example added, its test suite updated, and had the following hint added. (See 28cc7d5.)

This issue contains the remaining exercises for which Data.Text may be applicable. Contributors are free to pick an exercise they're comfortable with, (1) make an example solution, (2) make the test suite compatible with multiple string types, (3) add a hint, and optionally (4) make Data.Text the default in the stub file.

List of exercises that don't yet support Data.Text but could:

• rotational-cipher: Easy.
• raindrops: Easy.
• diamond: Easy.
• twelve-days: Easy.
• atbash-cipher: Easy.
• run-length-encoding: Doable.
• ocr-numbers: Doable.
• rail-fence-cipher: Doable.
• luhn: Easy, but hardly relevant as the exercise mainly consists of working on integers.
• series: Easy. Test suite already contains OverloadedStrings and already has ByteString -> Seq (Seq a). So a Data.Text example is probably not necessary unless the stub should encourage something other than String by default.

I've looked at the example solution(s) for each of these and made a quick estimate where "Easy" means that it would be quick to translate this to Data.Text by finding equivalent combinators, and "Doable" means that the contributor probably needs to learn (part of) the algorithm in the exercise to port it. The estimation was made very quickly and subjectively, so your mileage may vary.

[chiroptical]

I can add Data.Text support to diamond. I am working my way through the exercises and I don't want to contibute to ones I haven't completed. That being said, the next closest exercises are raindrops and twelve-days.

[sshine]

@barrymoo: That sounds good! Since you're adding property-based tests for Diamond in #843 right now, I would recommend that you either (1) wait with adding Data.Text support until after, and submit a separate PR, or (2) add support for Data.Text in a feature branch forked from master, submit a separate PR, and once this is merged, rebase #843 onto it.

I would recommend (1) because it involves fewer merge conflicts.

[chiroptical]

I will work on twelve-days next.