AshCache.Key (AshCache v0.1.0)

Builds deterministic cache keys for cached read and generic actions.

A key is a plain Erlang term (no hashing, so no collision risk) built from:

the resource module
the resource's current cache version (see AshCache.CacheVersion)
the action name
the whole query shape — arguments, filter, sort, limit, offset, page, distinct, select, load, calculations and aggregates (assembled in AshCache.Preparations.ReadThrough). filter is the load-bearing one: the policy authorizer has already folded its filter checks into it by the time the key is derived, so it is what stops an unauthorized read from being answered by an authorized one — not merely what tells two queries apart
for a generic action, only its arguments take the place of the query shape (AshCache.Actions.RunThrough)
for a calculation, {:calculation, name} takes the place of the action name, and the shape is one record's attributes, the fields the calculation loads and its arguments (AshCache.Calculations.CalculateThrough); there is no version
the actor and tenant, so cached results never leak across actors or tenants

Summary

Types

t()

Functions

build(resource, action_name, query_shape, version, opts)

Types

t()

@type t() ::
  {:ash_cache, resource :: module(), version :: term(), action :: atom(),
   query_shape :: term(), actor_ref :: term(), tenant_ref :: term()}

Functions

build(resource, action_name, query_shape, version, opts)

@spec build(module(), atom() | {:calculation, atom()}, map(), term(), keyword()) ::
  t()