Memoisasikan Fungsi yang Mahal

Sometimes only one slow helper is the problem, not the whole page. Memoization caches a function result keyed by its arguments.

You wrap a function with @cache.memoize(). Same arguments return the stored result instead of recomputing it. 💡

Memoize builds the key from the arguments. heavy(2) and heavy(5) are cached separately, each remembering its own answer.

Use cached for views keyed by URL, and memoize for plain functions keyed by their inputs. They solve different layers.

Just like views, you give memoize a timeout so results expire. After it passes, the next call recomputes and re-stores.

Memoize shines on pure functions: same input always gives the same output, with no side effects to lose.

Do not memoize a function that sends email or writes a row. On a cache hit the body is skipped, so the effect never runs.

For instance methods, the key includes self. Two objects cache separately, which is usually exactly what you want.

You can evict a single entry with delete_memoized, passing the function and the exact arguments to clear.

Pass arguments to delete_memoized to drop just that variant, leaving every other cached result untouched.

Only memoize what is actually slow. Profile first, because caching cheap work adds key overhead with little real gain.

You memoize a function that also logs each call to your database.

You can memoize slow pure functions, key results by arguments, set timeouts, and evict with delete_memoized. 🎯