Doubt developing extractors

[AllanDaemon]

I have a doubt about developing extractors.

When creating or developing an extractor, can I save data (state) in a instance of the extractor (like class MyExtractorIE(InfoExtractor)? In all the code I saw until now, it seems to avoid saving state in the IE object, working with a more functional approach and passing data just as functions arguments. Many helper methods in the class InfoExtractor (in extractor/common.py) kinda follow this behaving like static methods.

So, in an extractor, can I save data like the video id and others parameters in the object (eg. self.webpage = fetched_webpage), or assuming that each download item creates a new instance of an InfoExtractor is wrong, or something else is wrong.

[dstftw]

Extractors' instances are reused so you can't keep any data that affects extraction between downloads. Needless to say it does not even make any sense to store it.