Normalizing this data into a proper SQL schema seems like a lot of work
Which is why Koha is not doing it