largest-series-product forces implementation

[exercism]

largest-series-product forces implementation

[kytrinyx]

I've started playing around with aruba for making acceptance tests that make no choices about the design/classes/methods. I don't really want to introduce a new dependency, but we could definitely write fewer tests and not specify the steps along the way.

[markijbema]

I think it's mostly the tests should require more functionality each time, but through the one tested final method. But by having a trivial first example, and keep on adding bordercases, this can be done most of the time. I don't think the testing framework itself matters too much.

[kotp]

Related information:

• largest series exercise test and soltuion
• Meta for exercise
• Readme data

Listed here for my convenience.

[kotp]

@markijbema and @kytrinyx hopefully this is what you had in mind.

[kotp]

BTW: in the spirit of "Look to exercism for awesome code" I took the following measurements:

Flogging your code!
    44.7: flog total
     5.6: flog/method average

    11.9: Series#separate                  example.rb:30
     7.1: Series#largest_product           example.rb:9
     6.7: Series#collection_of_digits      example.rb:36
     6.4: Series#analyzer                  example.rb:17
Your code reeks this much!
example.rb -- 2 warnings:
  [19]:Series#analyzer contains iterators nested 2 deep (NestedIterators)
  [26]:Series#reduce_to_product doesn't depend on instance state (UtilityFunction)

[kotp]

   42.6: flog total                                     
    4.3: flog/method average                            

    6.7: Series#collection_of_digits      example.rb:45 
    6.5: Series#analyzer                  example.rb:22 
    6.4: Series#validate_length           example.rb:17 
    6.1: Series#separate                  example.rb:39 

[kytrinyx]

That's a much nicer result.