Grouping text entries by associated tags into hierarchical structure by theirs occurence across entries

tag_hierarchy.rb

#!/usr/bin/env ruby

# Following is a code which groups some entries hierarchically by it's tags composition. The hierarchy is determined by tag occurence in other entries.

require 'pry'

# entry = [id, some text description, array of tags]
entries = [
  [1, 'lorem ipsum', ['#customer_x', '#administration', '#paperwork']],
  [2, 'lorem ipsum', ['#customer_x', '#development']],
  [3, 'lorem ipsum', ['#customer_x', '#administration']],
  [4, 'lorem ipsum', ['#customer_y', '#administration']],
  [5, 'lorem ipsum', ['#customer_y', '#administration', '#invoice']],
  [6, 'lorem ipsum', ['#bike']],
  [7, 'lorem ipsum', ['#project_foo', '#mvp']],
  [8, 'lorem ipsum', ['#project_foo', '#docs']],
  [9, 'lorem ipsum', ['#project_bar', '#docs']],
  [10, 'lorem ipsum', ['#blog', '#article']],
  [11, 'lorem ipsum', ['#customer_x', '#maintenance', '#infrastructure']],
  [12, 'lorem ipsum', ['#customer_x', '#maintenance', '#infrastructure']],
  [13, 'lorem ipsum', []]
]

expected_result = [
  {"#customer_x"=>[{"#administration"=>[{"#paperwork"=>[{}, [[1, "lorem ipsum", []]]]}, [[3, "lorem ipsum", []]]], "#development"=>[{}, [[2, "lorem ipsum", []]]], ["#infrastructure", "#maintenance"]=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []],
   "#administration"=>[{"#customer_x"=>[{"#paperwork"=>[{}, [[1, "lorem ipsum", []]]]}, [[3, "lorem ipsum", []]]], "#customer_y"=>[{"#invoice"=>[{}, [[5, "lorem ipsum", []]]]}, [[4, "lorem ipsum", []]]]}, []],
   "#customer_y"=>[{"#administration"=>[{"#invoice"=>[{}, [[5, "lorem ipsum", []]]]}, [[4, "lorem ipsum", []]]]}, []],
   "#bike"=>[{}, [[6, "lorem ipsum", []]]],
   "#project_foo"=>[{"#mvp"=>[{}, [[7, "lorem ipsum", []]]], "#docs"=>[{}, [[8, "lorem ipsum", []]]]}, []],
   "#docs"=>[{"#project_foo"=>[{}, [[8, "lorem ipsum", []]]], "#project_bar"=>[{}, [[9, "lorem ipsum", []]]]}, []],
   ["#article", "#blog"]=>[{}, [[10, "lorem ipsum", []]]]},
  [[13, "lorem ipsum", []]]
]

def tag_hierarchy(entries)
  without_tag = []

  grouped = entries.inject({}) do |memo, entry|
    id = entry[0]
    desc = entry[1]
    tags = entry[2]

    without_tag << entry if tags.empty?

    tags.each do |tag|
      without = tags - [tag]
      current = memo[tag] || []
      memo[tag] = current + [[id, desc, without]]
    end

    memo
  end

  merged = {}
  grouped.each do |key, value|
    other_tag_combinations = value.map { _1[2] }.uniq
    other_tags = other_tag_combinations.flatten.uniq

    # Entries don't have other tags
    if other_tags.empty?
      next
    end

    # Entries which are details of only one other tag
    if other_tag_combinations.size == 1
      grouped.delete key
    end

    entry_ids = value.map &:first

    # Entries which all have only same combination of tags
    if other_tags.all? { |tag| grouped.fetch(tag, []).map(&:first) == entry_ids }
      new_key = other_tags << key
      new_key.each { grouped.delete _1 }
      merged[new_key] = value.map { [_1[0], _1[1], []]}
    end
  end

  merged = grouped.merge(merged)
  [merged.transform_values { |vs| tag_hierarchy vs }, without_tag]
end

result = tag_hierarchy entries

if result == expected_result
  puts "\e[32mCool, result matches expected value\e[0m"
else
  puts "\e[31mResult doesn't match the expected value\e[0m"
end

tag_pyramid.rb

#!/usr/bin/env ruby

# Following is a code which groups entries by theirs's tags and then recursively does the some with subgroups

require 'pry'

# entry = [id, some text description, array of tags]
entries = [
  [1, 'lorem ipsum', ['#customer_x', '#administration', '#paperwork']],
  [2, 'lorem ipsum', ['#customer_x', '#development']],
  [3, 'lorem ipsum', ['#customer_x', '#administration']],
  [4, 'lorem ipsum', ['#customer_y', '#administration']],
  [5, 'lorem ipsum', ['#customer_y', '#administration', '#invoice']],
  [6, 'lorem ipsum', ['#bike']],
  [7, 'lorem ipsum', ['#project_foo', '#mvp']],
  [8, 'lorem ipsum', ['#project_foo', '#docs']],
  [9, 'lorem ipsum', ['#project_bar', '#docs']],
  [10, 'lorem ipsum', ['#blog', '#article']],
  [11, 'lorem ipsum', ['#customer_x', '#maintenance', '#infrastructure']],
  [12, 'lorem ipsum', ['#customer_x', '#maintenance', '#infrastructure']],
  [13, 'lorem ipsum', []]
]

expected_result = [
  {"#customer_x"=>
   [{"#administration"=>[{"#paperwork"=>[{}, [[1, "lorem ipsum", []]]]}, [[3, "lorem ipsum", []]]],
     "#paperwork"=>[{"#administration"=>[{}, [[1, "lorem ipsum", []]]]}, []],
     "#development"=>[{}, [[2, "lorem ipsum", []]]],
     "#maintenance"=>[{"#infrastructure"=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []],
     "#infrastructure"=>[{"#maintenance"=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []]},
   []],
   "#administration"=>
   [{"#customer_x"=>[{"#paperwork"=>[{}, [[1, "lorem ipsum", []]]]}, [[3, "lorem ipsum", []]]],
     "#paperwork"=>[{"#customer_x"=>[{}, [[1, "lorem ipsum", []]]]}, []],
     "#customer_y"=>[{"#invoice"=>[{}, [[5, "lorem ipsum", []]]]}, [[4, "lorem ipsum", []]]],
     "#invoice"=>[{"#customer_y"=>[{}, [[5, "lorem ipsum", []]]]}, []]},
   []],
   "#paperwork"=>[{"#customer_x"=>[{"#administration"=>[{}, [[1, "lorem ipsum", []]]]}, []], "#administration"=>[{"#customer_x"=>[{}, [[1, "lorem ipsum", []]]]}, []]}, []],
   "#development"=>[{"#customer_x"=>[{}, [[2, "lorem ipsum", []]]]}, []],
   "#customer_y"=>[{"#administration"=>[{"#invoice"=>[{}, [[5, "lorem ipsum", []]]]}, [[4, "lorem ipsum", []]]], "#invoice"=>[{"#administration"=>[{}, [[5, "lorem ipsum", []]]]}, []]}, []],
   "#invoice"=>[{"#customer_y"=>[{"#administration"=>[{}, [[5, "lorem ipsum", []]]]}, []], "#administration"=>[{"#customer_y"=>[{}, [[5, "lorem ipsum", []]]]}, []]}, []],
   "#bike"=>[{}, [[6, "lorem ipsum", []]]],
   "#project_foo"=>[{"#mvp"=>[{}, [[7, "lorem ipsum", []]]], "#docs"=>[{}, [[8, "lorem ipsum", []]]]}, []],
   "#mvp"=>[{"#project_foo"=>[{}, [[7, "lorem ipsum", []]]]}, []],
   "#docs"=>[{"#project_foo"=>[{}, [[8, "lorem ipsum", []]]], "#project_bar"=>[{}, [[9, "lorem ipsum", []]]]}, []],
   "#project_bar"=>[{"#docs"=>[{}, [[9, "lorem ipsum", []]]]}, []],
   "#blog"=>[{"#article"=>[{}, [[10, "lorem ipsum", []]]]}, []],
   "#article"=>[{"#blog"=>[{}, [[10, "lorem ipsum", []]]]}, []],
   "#maintenance"=>[{"#customer_x"=>[{"#infrastructure"=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []], "#infrastructure"=>[{"#customer_x"=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []]}, []],
   "#infrastructure"=>[{"#customer_x"=>[{"#maintenance"=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []], "#maintenance"=>[{"#customer_x"=>[{}, [[11, "lorem ipsum", []], [12, "lorem ipsum", []]]]}, []]}, []]},
[[13, "lorem ipsum", []]]
]

def tag_pyramid(entries)
  without_tag = []

  grouped = entries.reduce({}) do |memo, entry|
    id = entry[0]
    desc = entry[1]
    tags = entry[2]

    without_tag << entry if tags.empty?

    tags.each do |tag|
      current = memo[tag] || []
      memo[tag] = current + [[id, desc, tags - [tag]]]
    end

    memo
  end

  grouped.transform_values! { |subentries| tag_pyramid subentries }

  [grouped, without_tag]
end

result = tag_pyramid entries

if result == expected_result
  puts "\e[32mCool, result matches expected value\e[0m"
else
  puts "\e[31mResult doesn't match the expected value\e[0m"
end