Two nice ressources with plenty of examples based on real life needs, including some other urls to look at when you need to clean data before processing them. Maybe having similar examples in DCG would help making it more often used mainstream ?
At least flexible “bashing solutions” exist before processing with standard Prolog libraires:
Have fun !
EDIT - “real life needs” = the examples in those sites are based on the experience of a data auditor and retired zoologist.