diff --git a/lib/wayback_machine_downloader/to_regex.rb b/lib/wayback_machine_downloader/to_regex.rb
new file mode 100644
index 0000000..a059ccf
--- /dev/null
+++ b/lib/wayback_machine_downloader/to_regex.rb
@@ -0,0 +1,81 @@
+module ToRegex
+  module StringMixin
+    class << self
+      def literal?(str)
+        REGEXP_DELIMITERS.none? { |s, e| str.start_with?(s) and str =~ /#{e}#{INLINE_OPTIONS}\z/ }
+      end
+    end
+
+    INLINE_OPTIONS = /[imxnesu]*/
+    REGEXP_DELIMITERS = {
+      '%r{' => '}',
+      '/' => '/',
+    }
+
+    # Get a regex back
+    #
+    # Without :literal or :detect, `"foo".to_regex` will return nil.
+    #
+    # @param [optional, Hash] options
+    # @option options [true,false] :literal Treat meta characters and other regexp codes as just text; always return a regexp
+    # @option options [true,false] :detect If string starts and ends with valid regexp delimiters, treat it as a regexp; otherwise, interpret it literally
+    # @option options [true,false] :ignore_case /foo/i
+    # @option options [true,false] :multiline /foo/m
+    # @option options [true,false] :extended /foo/x
+    # @option options [true,false] :lang /foo/[nesu]
+    def to_regex(options = {})
+      if args = as_regexp(options)
+        ::Regexp.new *args
+      end
+    end
+
+    # Return arguments that can be passed to `Regexp.new`
+    # @see to_regexp
+    def as_regexp(options = {})
+      unless options.is_a?(::Hash)
+        raise ::ArgumentError, "[to_regexp] Options must be a Hash"
+      end
+      str = self
+
+      return if options[:detect] and str == ''
+
+      if options[:literal] or (options[:detect] and ToRegexp::String.literal?(str))
+        content = ::Regexp.escape str
+      elsif delim_set = REGEXP_DELIMITERS.detect { |k, _| str.start_with?(k) }
+        delim_start, delim_end = delim_set
+        /\A#{delim_start}(.*)#{delim_end}(#{INLINE_OPTIONS})\z/u =~ str
+        content = $1
+        inline_options = $2
+        return unless content.is_a?(::String)
+        content.gsub! '\\/', '/'
+        if inline_options
+          options[:ignore_case] = true if inline_options.include?('i')
+          options[:multiline] = true if inline_options.include?('m')
+          options[:extended] = true if inline_options.include?('x')
+          # 'n', 'N' = none, 'e', 'E' = EUC, 's', 'S' = SJIS, 'u', 'U' = UTF-8
+          options[:lang] = inline_options.scan(/[nesu]/i).join.downcase
+        end
+      else
+        return
+      end
+
+      ignore_case = options[:ignore_case] ? ::Regexp::IGNORECASE : 0
+      multiline = options[:multiline] ? ::Regexp::MULTILINE : 0
+      extended = options[:extended] ? ::Regexp::EXTENDED : 0
+      lang = options[:lang] || ''
+      if ::RUBY_VERSION > '1.9' and lang.include?('u')
+        lang = lang.delete 'u'
+      end
+
+      if lang.empty?
+        [ content, (ignore_case|multiline|extended) ]
+      else
+        [ content, (ignore_case|multiline|extended), lang ]
+      end
+    end
+  end
+end
+
+class String
+  include ToRegex::StringMixin
+end