Serlo scraper: needs an items filter

[AOHPI]

The current results of Serlo metadata are showing too much of organizational stuff, which is not relevant for any lessons. (Although the information might be interesting itself.)

Recommended fix: implement a filter based on the keywords, because thats the place where you can tell.

[torsten-simon]

You might simply handle this yourself in the spider by overriding the shouldImport(self, response) method and return false or true depending whether you want to import the specific object. Or did I miss something?

[MRuecklCC]

Can we close this as won't fix / outdated?