From: James Edward Gray II <james@...>
Date: 2007-06-20T02:19:42+09:00
Subject: Re: [QUIZ] Verbal Arithmetic (#128)

On Jun 19, 2007, at 9:30 AM, Eugene Kalenkovich wrote:

> On Jun 18, 4:59 pm, James Edward Gray II <j...@grayproductions.net>
> wrote:
>>> It's of the dumb-brute-force-slow-as-hell variety:
>> Mine was too:
>> #!/usr/bin/env ruby -wKU
>>
>> EQUATION = ARGV.shift.to_s.downcase.sub("=", "==")
>> LETTERS  = EQUATION.scan(/[a-z]/).uniq
>> CHOICES  = LETTERS.inject(Hash.new) do |all, letter|
>>    all.merge(letter => EQUATION =~ /\b#{letter}/ ? 1..9 : 0..9)
>> end
>>
>> def search(choices, mapping = Hash.new)
>>    if choices.empty?
>>      letters, digits = mapping.to_a.flatten.partition { |e| e.is_a?
>> String }
>>      return mapping if eval(EQUATION.tr(letters.join, digits.join))
>>    else
>>      new_choices = choices.dup
>>      letter      = new_choices.keys.first
>>      digits      = new_choices.delete(letter).to_a - mapping.values
>>
>>      digits.each do |choice|
>>        if result = search(new_choices, mapping.merge(letter =>  
>> choice))
>>          return result
>>        end
>>      end
>>
>>      return nil
>>    end
>> end
>>
>> if solution = search(CHOICES)
>>    LETTERS.each { |letter| puts "#{letter}: #{solution[letter]}" }
>> else
>>    puts "No solution found."
>> end
>>
>> __END__
>>
>
> I really like it - your solution covers any type of expression :).
> Perhaps some preprocessing of CHOICES can save the performance....

Thanks.  There have been far more clever submissions though.

> The only thing, assuming that all examples are always integers, I'd  
> add
>
> .sub(/([a-z])([^a-z])/,'\1.0\2')
>
> to EQUATION to cover examples with division (to drop the  
> assumption, before
> adding .0 one can regexp it for '.'s)
>
> for  'boaogr/foo=bar' (arbitrary example) your code gives
>
> b: 1
> o: 0
> a: 2
> g: 3
> r: 7
> f: 8
>
> and modified one corrects it to
>
> b: 1
> o: 0
> a: 2
> g: 3
> r: 7
> f: 8
>
> I can expect that this modification may break on rounding errors,  
> but in
> this task domain I do not think it will

Good points.

James Edward Gray II