From: Daniel Firu <dannyfiru@...>
Date: 2009-01-17T08:28:53+09:00
Subject: Re: Convert String to Float if and only if the content of the string 	are digits

On Jan 16, 3:03 pm, Daniel Firu <dannyf...@gmail.com> wrote:
> Hi All,
>
> Long time reader -- first time poster.  Basically, I have a bunch of
> data that is sucked in from a file.  From reading the pick axe book,
> it is my understanding that this data by default will be a string even
> if the contents of that string are a bunch of digits.  Well, in the
> case that a string is solely composed of digits, I want to convert
> them to a float for the purpose of using the <=> operator.  I'll just
> go ahead and throw in my method and let you guys tear it a part.  The
> goal of the function is to soft my array of arrays based on a
> particular element defined by a 'key'.  Maybe there is an easier way
> to do this all together.
>
>   def sort(*keys) options = {}
>     regex = "^(\d)*$"
>     newData = Array.new
>     keys.reverse!
>     keys.each do |key|
>       newData = @data
>       keyIndex = @header.index(key)
>       newData.sort! do |entry1, entry2|
>         if (!(entry1[keyIndex].match(regex).nil? && entry2
> [keyIndex].match(regex).nil?))
>           print "I found only digits\n"
>           entry1[keyIndex].to_f <=> entry2[keyIndex].to_f
>         else
>           #print "I found something other than digits\n"
>           entry1[keyIndex] <=> entry2[keyIndex]
>         end
>       end
>     end
>     newTable = Table.new({
>                            :header => @header,
>                            :data => newData
>                          })
>     newTable
>   end
>
> Thanks in advance for you feedback.  Feel free to be brutal -- I have
> thick skin and am still learning.

Hi All,

I went ahead and changed my method to:

  def sort(*keys) options = {}
    regex = Regexp.new('^\d+$')
    newData = Array.new
    keys.reverse!
    keys.each do |key|
      newData = @data
      keyIndex = @header.index(key)
      newData.sort! do |entry1, entry2|
        if (entry1[keyIndex] =~ regex && entry2[keyIndex] =~ regex)
          #print "I found only digits\n"
          entry1[keyIndex].to_f <=> entry2[keyIndex].to_f
        else
          #print "I found something other than digits\n"
          entry1[keyIndex] <=> entry2[keyIndex]
        end
      end
    end
    newTable = Table.new({
                           :header => @header,
                           :data => newData
                         })
    newTable
  end

Now the question still remains if there is a better way to do this
sorting.