lib/ebooks/generator.rb in ebooks-0.2.0 vs lib/ebooks/generator.rb in ebooks-0.2.1

@@ -17,11 +17,14 @@
       csv_text = CSV.read(@tweets_csv_path)
 
       # Create a new clean file of text that acts as the seed for your Markov chains
       File.open(@corpus_path, 'w') do |file|
         csv_text.reverse_each do |row|
-          # Strip links and new lines
-          tweet_text = row[5].gsub(/(?:f|ht)tps?:\/[^\s]+/, '').gsub(/\n/,' ')
+          tweet_text = row[5]
+                        .gsub(/(?:f|ht)tps?:\/[^\s]+/, '') # Strip links
+                        .gsub(/\n/,' ') # Strip new lines
+                        .gsub(/@[a-z0-9_]+/i, '') # Strip usernames
+                        .gsub(/[R|M]T/, '') # Strip RTs
           # Save the text
           file.write("#{tweet_text}\n")
         end
       end
     end