Using a Django Request Cache in place of Memoized Cache.

[robrap]

Hello.

I was implementing my own version of a TieredCache in Django when a colleague pointed me to this caching solution. I like it, but the local cache I want is a Request Cache, rather than the Timed Memoized Cache. I also heard you may already have a Request Cache version available?

Thank you.

Robert

[dannyroberts]

There are currently two ways that you achieve the effect of request-scoping all in-memory caching:

• Option 1: Pass in a session_function which takes no arguments and returns a short string that varies on request or threading.local() or whatever else you use to define a "session" or single conceptual execution period. For an example of how we've done that in another project, you can refer to: https://github.com/dimagi/commcare-hq/blob/08dde6f16afaf3aed1d1ee500986e3b06aab6bc3/corehq/util/quickcache.py#L19-L47
• Option 2: Write a Django cache backend that stores the cache on the request or threading.local(). If you name this cache 'locmem' in your settings.py, quickcache will use it.

It occurs to me that I have not documented that get_django_quickcache uses your Django caches named 'default' (for shared cache) and 'locmem' (for in-memory cache), and that these should be something you can specify when calling get_django_quickcache. I've opened an issue about this https://github.com/dimagi/quickcache/issues/13

[robrap]

Thanks for your response Danny. I'll let you know if I have any questions, but hopefully this should cover it.