Input of scraping is a curriculum, not a list of challenges

[benoit74]

As of 1.1, input to the scraper is a list of challenges. Indeed, FCC maintains a list of curriculum, with challenges in every of them.

The scraper must accept a curriculum as entrypoint instead of a list of challenges.

[benoit74]

This is needed to avoid maintaining manually the list of courses in one curriculum, since we mostly want to create one ZIM per curriculum.

But at the same time, it means that we cannot create our own curriculum easily, and that we will need a new parameter to exclude courses / challenges which are known not to work.

I don't think this is really very clear yet what is the best compromise between having something automatically adding courses to existing ZIMs as they are added to the curriculum (which is probably what we want) and avoiding breaking something "automatically".