I’m currently scraping a few studios which have 2 different languages for each studio. They have several naming conventions for scene numbers. Some use numbers(21), some use roman numerals(XXI) some use written out numbers (Twenty- One).Sometimes they are misspelled written out numbers(TwentOne). And half of them are not in english.
There are 4 sources of titles. The websites release, the IMDB page, the video metadata and the title card.
I was told we should be naming files based on what the studio calls them, but when it is inconsistent, sorting becomes challenging. Normalizing around the idea of what the studio does is pretty brittle too.
Furthermore, Stash can’t sort based on number names, only on actual numbers. Naming something Twenty One means it cannot be sorted in order.