Cache API Reference

Cache

from lightapi import Cache

Cache(
    ttl: int,                        # seconds — required
    vary_on: list[str] | None = None,
)

Enables response caching for GET list and detail endpoints. Configured via Meta.cache:

from lightapi import RestEndpoint, Cache

class ProductEndpoint(RestEndpoint):
    name: str
    price: float

    class Meta:
        cache = Cache(ttl=300)

Parameter	Type	Description
ttl	int	Cache lifetime in seconds. Must be ≥ 1.
vary_on	list[str] \| None	Query parameter names included in the cache key.

Raises ConfigurationError if ttl < 1.

RedisCache

from lightapi import RedisCache

RedisCache(
    host: str = "localhost",
    port: int = 6379,
    db: int = 0,
)

The built-in Redis cache backend. Serializes cached values as JSON.

Parameter	Default	Description
host	"localhost"	Redis server hostname
port	6379	Redis server port
db	0	Redis database index

BaseCache

Base class for custom cache backends:

from lightapi.cache import BaseCache
from typing import Any, Dict, Optional

class MyCache(BaseCache):
    def get(self, key: str) -> Optional[Dict[str, Any]]:
        ...

    def set(self, key: str, value: Dict[str, Any], timeout: int = 300) -> bool:
        ...

Cache key

The default cache key includes the endpoint class name and the full request URL (path + query string). When vary_on is set, only the listed query parameters are included in the key.

Notes

• Caching applies to GET requests only.
• Write operations (POST, PUT, PATCH, DELETE) do not automatically invalidate cached entries.
• Use a short ttl or implement manual Redis key invalidation for consistency.