So, this is one of those non-magic places in Ruby

app/models/spam_wave.rb

@@ -65,11 +65,14 @@ def post_matches?(post, site_ids = nil)
       # We only care about everything matching.
       # Returning here saves testing the longer strings if a shorter one doesn't match.
       post_text = post.send(f.to_sym)
-      post_text = '' if post_text.nil?
-      # UTF-8 -> UTF-16 -> UTF-8 idea and code from [answer to: "ruby 1.9: invalid byte sequence in UTF-8"](https://stackoverflow.com/a/8873922)
-      # by [RubenLaguna](https://stackoverflow.com/users/90580/rubenlaguna), which is under a CC BY-SA 3.0 license.
-      post_text = post_text.encode!('UTF-16', 'UTF-8', invalid: :replace, replace: '')
-      post_text = post_text.encode!('UTF-8', 'UTF-16')
+      if post_text.nil?
+        post_text = ''
+      else
+        # UTF-8 -> UTF-16 -> UTF-8 idea and code from [answer to: "ruby 1.9: invalid byte sequence in UTF-8"](https://stackoverflow.com/a/8873922)
+        # by [RubenLaguna](https://stackoverflow.com/users/90580/rubenlaguna), which is under a CC BY-SA 3.0 license.
+        post_text.encode!('UTF-16', 'UTF-8', invalid: :replace, replace: '')
+        post_text.encode!('UTF-8', 'UTF-16')
+      end
       Rails.logger.debug "[spam-wave] id: #{id}: #{name}:: post #{f}: encoding: #{post_text.encoding}"
       return false unless regex.match?(post_text)
       Rails.logger.debug "[spam-wave] id: #{id}: #{name}:: post_matches?: #{f}: MATCHES: post id: #{post.id}: #{post.title}"