This is still kind of a half-baked feature, which is why I
currently don't really want to document it. If you're curious,
you can have a look at the
getFromCache method in the
The purpose of this feature: when you're working on a crawler, you will probably run it again and again. Performing actual HTTP requests during all the test runs makes the runs unnecessary slow and produces a lot of network traffic and traffic on the crawled servers. Using the response cache during development allows for very fast test runs, once the needed pages have been loaded for the first time.