I have written an indexed module called Text::Statistics::Latin that performs some works related to your needs. You can start by choosing a way to open your files, one each time, and writing output that will be opened again etc. Would be indispensable to use regex, in my opinion. So, start thinking in a procedure to open-write-read sequencies and handling data with regex. I also sugest to convert your Excel (c) files into CSV files. If so, my module could realy be usefull as inspiration.