APICache

A lightweight proxy-cache that sits between your app and a public API.

It helps avoid rate limits, enables reproducible experiments, and incrementally builds up datasets by caching structured responses — not raw HTTP.

Solving the frustration of API rate limits, inconsistent responses, and unreproducible workflows by caching structured API responses during development for reuse, graphing, and analysis.

Installation

pip install apicache

Behind the scenes:

• All API responses are cached in a local SQLite store
• Cached responses are versioned with timestamps and TTL-aware
• The system gracefully falls back to live data when needed
• Cached entries can be purged, inspected, or used to reconstruct state

Basic Usage

"""proxied-cache to UK govs Companies House API"""
from cache import APICache
import requests
import os

API_KEY = os.environ["COMPANIES_HOUSE_API_KEY"]
BASE_URL = "https://api.company-information.service.gov.uk"

def fetch_from_ch_api(url, params):
    """function to call on cache miss"""
    full_url = url.format(**{**params, "BASE_URL": BASE_URL})
    response = requests.get(full_url, auth=(API_KEY, ""))
    if response.status_code != 200:
        raise ValueError(f"Failed [{response.status_code}]")
    return response.json()

# define the cache object
cache = APICache(request_fn=fetch_from_ch_api, ttl=3600)

# make a request to the backend API
# cache miss will call the `fetch_from_ch_api` function
cache_hit, data = cache.request("{BASE_URL}/company/{company_number}", {"company_number": "12345678"})
print("Cache hit?", cache_hit)
print("Company:", data["company_name"])
print("Stats:", cache.stats())

Use Cases

• Data scientists experimenting with public APIs
• Reproducible research pipelines without rate-limit worries
• Mockable APIs for CI/offline testing
• API usage minimization or budgeting
• Building incremental graphs or datasets over time

Examples

Google News RSS

import feedparser
from cache import APICache

def fetch_google_news(url, params):
    query = params["q"]
    full_url = f"https://news.google.com/rss/search?q={query}"
    return feedparser.parse(full_url)

cache = APICache(request_fn=fetch_google_news, ttl=86400)

hit, parsed = cache.request("https://news.google.com/rss/search?q={q}", {"q": "next high tide"})
for entry in parsed["entries"]:
    print(entry["title"])

GitHub API (with headers)

import requests
from cache import APICache

def fetch_github(url, params):
    full_url = url.format(**params)
    resp = requests.get(full_url, headers={"Accept": "application/vnd.github+json"})
    return resp.json()

cache = APICache(request_fn=fetch_github)

hit, user = cache.request("https://api.github.com/users/{username}", {"username": "torvalds"})
print("Name:", user["name"])

Extended usage

The cache indexs over url: str and params: dict, so any data may be cached via this method. Cache misses are resolved with the fetch_fn callback, which can custom process any actual interaction with the backend API.

Slow Fibonacci

Of course, you should explore other caching options like functools.lru_cache but it is sometimes useful to capture the cached data.

from cache import APICache
import time

def slow_fib(_, params):
    """Pretend this is a slow remote call."""
    n = params["n"]
    time.sleep(0.2)  # simulate latency
    if n <= 1:
        return n
    return fib_cache.request("fib", {"n": n - 1})[1] + fib_cache.request("fib", {"n": n - 2})[1]

fib_cache = APICache(request_fn=slow_fib)

for i in range(20):
    hit, result = fib_cache.request("fib", {"n": i})
    print(f"fib({i}) = {result} {'(cache hit)' if hit else '(computed)'}")

print("\nCache stats:", fib_cache.stats())

OpenAI library calls

In this example, we wrap the APICache in a simple object to proxy the OpenAI backend without losing the rich object responses defined in the OpenAI library.

NB: not fully tested

import openai
from openai.util import convert_to_openai_object
from cache import APICache
import hashlib
import json


class OpenAICache:
    def __init__(self, ttl: int = 86400, location: str = None):
        self.cache = APICache(request_fn=self._fetch_from_openai, ttl=ttl, location=location)

    def _fetch_from_openai(self, key: str, params: dict):
        if key == "openai:chat":
            response = openai.ChatCompletion.create(**params)
        elif key == "openai:image":
            response = openai.Image.create(**params)
        else:
            raise ValueError(f"Unsupported OpenAI API type in key: {key}")

        return response.to_dict()

    def chat(self, **params):
        """Cached chat completion"""
        hit, raw = self.cache.request("openai:chat", params)
        return convert_to_openai_object(raw), hit

    def image(self, **params):
        """Cached image generation"""
        hit, raw = self.cache.request("openai:image", params)
        return convert_to_openai_object(raw), hit

API Reference

cache = APICache(request_fn, ttl=3600, location="api-cache.sqlite")

cache.request(url, params) → (hit: bool, response: dict)
cache.read(url, params)
cache.write(url, params, response)
cache.has(url, params)
cache.clear_for_url("/some/api/prefix")
cache.prune_old_versions()
cache.stats() → dict