MoinMoin: MelitaMihaljevic/RefactoringWikidicts/WikidictsTrace

This is some wikidicts tracing and questions in order to understand how wikidicts and caching work together.

User requests a page, for example: http://127.0.0.1:8080/MyPage Than it calls wikidicts.py:

• init GroupDict object

• load_dicts:

  • instance cacheEntry object with defined key and arena

  • get the currently disk_cache_id-> cache.uid()

    • uid is a tuple (write here what that represents- it is mtime, string,...)

  • try to:

    •       self.__dict__.update(self.cfg.cache.DICTS_DATA)

    • update dict ( DICTS_DATA ) - loading data from cache

  • if there is no cache or is no uptodate - reset and force to load from disk (check the new uid and compare - current_disk_cache_id != self.disk_cache_id )
  • except:
    • got the cache content - this is something that's a bit confusing (see questions)
    • update dictionary with currently read data from cache
    • update cache id

    • save data and cache id to -> data and update DICTS_DATA

  • later if someone requests for a group page again it only reads DICTS_DATA if a page content wasn't changed

Questions

Here are some questions I'm not sure that I understand the process and I tried to give an answer/s to it, or some options,

1.What is the difference between getting the cache.DICTS_DATA and cache.content()

• cache.content() - it reads data from a disk (it's cached but it's on a disk and has to access disk) Does is need always to access the disk?

  • No, if it's pickled it does: data = pickle.loads(data)

• cache.DICTS_DATA - it's not only data from disk - it has cache_id , data, ... - but I'm not sure where is DICTS_DATA saved - it cant be read from disk but not sure

  • see into request.init. request parameters needs to be initialized