how count words in each line of a spark data frame?

Question

I have a huge data set, in each row there are some titles that have been separated by , . I want to do two things:

1- remove , if they are follo