Why does chapter 2 section 7 solution use iteration instead of matching? about spacy-course HOT 2 CLOSED

Hi! This is a good question and definitely a valid point. There's no particular reason and the Matcher would definitely be the more elegant and scalable solution in this case. There are other cases where you might still want to iterate over the tokens – for instance, if you're working with the dependency tree or if you're extracting additional information after matching (e.g. check the previous sentence for something).

This particular exercise is based on a real example I once saw on Stack Overflow and I wanted to get the point of using the Doc as the "single source of truth" across. And since rewrite exercises are often a bit more challenging than "fill in the gaps", I didn't want to introduce any new concepts or new token attributes here, and keep the result a bit closer to the original code.

tl;dr: No particular reason, mostly to make the rewrite exercise more straightforward and not ask for too much at once. Matching is perfectly reasonable, too 🙂

from spacy-course.

Hi @ines, thank you for the reply (and the exercises too), that clears things up and adds more context to the tradeoffs between the two methods.

from spacy-course.