Useful Unicode Regex (Ruby)

StringSanitizer.rb

module StringSanitizer
  extend self

  ##
  # This is used for different kinds of cleanup of a string
  # Use cases:
  #   1. When creating reports in different formats with varying support for unicode.
  #   2. For Data inputs to remove whitespace and strange characters which are not "language"
  #   3. External services which may not be able to handle different character sets
  #   4. Comparisons - when comapring two strings which have whitespace and weird unicode
  #
  # Examples:
  #
  # str = "bob, g, Germany 🇩🇪"
  # StringSanitizer.call(str, strategy: :letters_accents_and_numbers_only)
  # => "bob g Germany"
  #
  # str = "Добро пожаловать в джунгли.             "
  # StringSanitizer.call(str, strategy: :letters_accents_and_numbers_only)
  # => "Добро пожаловать в джунгли"
  #
  # str = "ABCDEпожаловатьF"
  # StringSanitizer.call(str, strategy: :ascii_only)
  # => "ABCDEF"

  # List of useful REGEX
  # https://www.regular-expressions.info/refunicode.html
  # https://stackoverflow.com/questions/24672834/how-do-i-remove-emoji-from-string
  PICTURE_REGEX = /[\u{1f300}-\u{1f5ff}]/
  ENCLOSED_CHAR_REGEX = /[\u{2500}-\u{2BEF}]/
  EMOTICONS_REGEX = /[\u{1f600}-\u{1f64f}]/
  DINGBATS_REGEX = /[\u{2702}-\u{27b0}]/
  TRANSPORT_AND_MAP_REGEX = /[\u{1f680}-\u{1f6ff}]/
  REGIONAL_INDICATOR_SYMBOL_REGEX = /[\u{1f1e6}-\u{1f1ff}]/ # flags
  SPACE_BEFORE_COMMA_REGEX = / \,/

  # Inverted Regex
  ASCII_ONLY_REGEX = /[^\u{0000}-\u{007f}]/
  LETTERS_ACCENTS_AND_DIGITS_ONLY_REGEX = /[^\p{L}\p{M}\d+\s]/
  ASCII_AND_LANGUAGE_CHARACTERS_ONLY_REGEX = /[^\u{0000}-\u{007f}\p{L}\p{M}\d+]/

  def call_multiple(str, strategies: [:strip_extra_whitespace_only])
    strategies.each do |strategy|
      str = self.call(str, strategy: strategy)
    end

    str
  end

  def call(str, strategy: :strip_extra_whitespace_only)
    return str unless str.is_a?(String)
    str = str.force_encoding('utf-8').encode

    case(strategy)
    when :ascii_and_language_symbols_only
      str = ascii_and_language_symbols_only(str)
    when :ascii_only
      str = ascii_only(str)
    when :letters_accents_and_numbers_only
      str = letters_accents_and_numbers_only(str)
    when :strip_special_unicode_classes_only
      str = strip_special_unicode_classes_only(str)
    when :strip_html_tags
      str = strip_html_tags(str)
    else
      str = strip_space_before_comma(str)
    end

    str.squish
  end

  private

  # Dangerous to just strip all characters which are not-ascii because then we cannot
  # support other locales like ru, de, etc ...
  def ascii_and_language_symbols_only(str)
    str = strip_space_before_comma(str)
    str.gsub(ASCII_AND_LANGUAGE_CHARACTERS_ONLY_REGEX, '')
  end

  def ascii_only(str)
    str = strip_space_before_comma(str)
    str.gsub(ASCII_ONLY_REGEX, '')
  end

  def letters_accents_and_numbers_only(str)
    str = strip_space_before_comma(str)
    str.gsub(LETTERS_ACCENTS_AND_DIGITS_ONLY_REGEX, '')
  end

  def strip_special_unicode_classes_only(str)
    str = strip_space_before_comma(str)
    str = str.gsub(PICTURE_REGEX, '')
    str = str.gsub(ENCLOSED_CHAR_REGEX, '')
    str = str.gsub(EMOTICONS_REGEX, '')
    str = str.gsub(DINGBATS_REGEX, '')
    str = str.gsub(TRANSPORT_AND_MAP_REGEX, '')
    str.gsub(REGIONAL_INDICATOR_SYMBOL_REGEX, '')
  end

  def strip_space_before_comma(str)
    str.gsub(SPACE_BEFORE_COMMA_REGEX, '')
  end

  def strip_html_tags(str)
    ActionController::Base.helpers.strip_tags(str)
  end
end