This week, I've learned the names of numerous comic characters I'd never heard before ... Beside that, my record count is not exactly identical to the solution, difference seems to be caused by handling of issues like "12.1" or "0.5".
Okay, I must admit, I had a small challenge in getting the result to match the provided output, but in the end, nothing a little data cleansing couldn't handle, so my workflow ended looking like this: My final workflow