Example

./extract.sh | wc -l

extract.sh

#!/bin/bash

cat <(./facebook_messages.rb) <(./facebook_messages.rb) |
  ./split.rb |
  ./lowercase.rb |
  egrep '.{2,}' |
  sort | uniq |
  ./mystem -w -n -e utf-8 |
  grep -v '??' |
  sed -n -e '$!N' -e 's/\n//g' -e 's/.*{\([^{][^{]*\)}.*/\1/gp' |
  awk -F\| '{ for (i = 0; ++i <= NF; )print $i }' |
  sort | uniq

facebook_messages.rb

#!/usr/bin/env ruby

require 'bundler/setup'
require 'nokogiri'

input = './facebook-amikhailov83/html/messages.htm'

doc = Nokogiri::HTML(open(input))

doc.css('.message').each do |header|
  author = header.css('.message_header span.user').first

  if author.content == 'Алексей Михайлов'
    message = header.next

    puts message.content
  end
end

facebook_wall.rb

#!/usr/bin/env ruby

require 'bundler/setup'
require 'nokogiri'

input = './facebook-amikhailov83/html/wall.htm'

doc = Nokogiri::HTML(open(input))

doc.css('.comment').each do |comment|
  header = comment.previous.content

  if header.include?('Алексей Михайлов') && !header.include?('posted something via')
    puts comment.content
  end
end

Gemfile

source 'https://rubygems.org'

gem 'nokogiri'
gem 'sanitize'
gem 'activesupport'

lowercase.rb

#!/usr/bin/env ruby
# coding: utf-8

require 'bundler/setup'
require 'active_support/core_ext/string/multibyte'

$stdin.each_line { |x| puts x.mb_chars.downcase }

split.rb

#!/usr/bin/env ruby

$stdin.each_line do |line|
  line.scan(/([а-я]+)(-[а-я]+)?/i).each do |x|
    puts x.join
  end
end