Question

I have an XML-File with <pb n="4-DIGIT-NUMBER" ... />. The number being in some cases identical, so I’d like to disambiguate, coming so far, but now problems with counting (do I have the right approach? => 3))

1) Reading the all numbers into an Array, yielding a very long list with:

Dir.chdir("./Tustep/luxneu")
sammel = []
open("lp42tags.txt").each do |x|
  if x =~ /<pb n="(\d\d\d\d)/
  sammel << $1
  end
end

2) Finding the numbers repeating and put them into subarrays

dupl_groups = sammel.select{|i| sammel.grep(i).size > 1}.group_by{|x| x}.values
p dupl_groups
# (much shorter example)=> [["0119", "0119"], ["0147", "0147"], ["0156", "0156", "0156"]]

3) Now I thought I could somehow count the elements of each subarray and put them back into (or into a copy..). I want e.g. [[“0119:1”, “0119:2”], […], [“0156:1”, “0156:2”, “0156:3”], maybe like this (but only got hilarious loops with almost endless number computations… :/)

dupl_counted = []
dupl_groups.each do |outer|
  count = 1
  dupl_groups do |inner|
    #puts inner.inspect
    inner_new = inner.to_s.sub(/(.+)/, "\\1:#{count}")
    dupl_counted << inner_new
    count += count
  end
end

Seriously flawed..? Maybe something instead using “each_with_index”? Also I need the groups for counting in meaningful chunks (slice 3 or so is unacceptable, because there are number-repetitions ranging from 2-6). If I could split the array in its subarrays yielding them all as normal arrays, would that be good?

Thanks in advance!
René T.

Answer 1

This should be just a nested application of map – once to the outer group, and then to each element within:

dupl_groups.map do |gp| 
  gp.map.with_index {|el, ix| el + ":#{ix+1}"}
end
# => [["0119:1", "0119:2"], ["0147:1", "0147:2"], ["0156:1", "0156:2", "0156:3"]]