Hello Dr. Ted please
1. text-nsp does not see the unicode or utf-8 file 2. I think to put the Arabic in windows-1256 file and then replace the while loop in count.pl # push @tokenRegex, "\\w+"; push @tokenRegex, "[\w\x80-\xff]+"; the problem is still exist C:\text_nsp>count output.txt ara_def.txt No input (SOURCE) file supplied! Type count.pl --help for help.