""" Rate limiter for FastAPI — Redis-backed in production, in-memory fallback. Redis mode: works correctly across multiple uvicorn workers. Memory mode: single-process only (development / single-worker deploys). Env vars: REDIS_URL — e.g. "redis://redis:6379/0" (enables Redis mode) If unset, falls back to in-memory sliding window. """ from __future__ import annotations import os import time import logging import threading from collections import defaultdict, deque from fastapi import Request, HTTPException log = logging.getLogger("tebyan.ratelimit") _REDIS_URL = os.getenv("REDIS_URL", "") _IS_PROD = os.getenv("ENVIRONMENT", "development") == "production" _TRUSTED_PROXY = os.getenv("TRUSTED_PROXY_IPS", "").split(",") def _extract_ip(request: Request) -> str: """Return real client IP. Only trust X-Forwarded-For in production behind nginx.""" if _IS_PROD: forwarded = request.headers.get("X-Forwarded-For", "") if forwarded: return forwarded.split(",")[0].strip() return request.client.host if request.client else "unknown" # ── Redis-backed limiter ──────────────────────────────────────────────────── class _RedisLimiter: """ Sliding-window rate limiter backed by Redis sorted sets. Works correctly with multiple uvicorn workers. """ def __init__(self, redis_url: str) -> None: import redis as _redis self._r = _redis.from_url(redis_url, decode_responses=True, socket_connect_timeout=2) self._r.ping() log.info("[ratelimit] Redis mode | url=%s", redis_url.split("@")[-1]) def _key(self, request: Request, endpoint: str) -> str: ip = _extract_ip(request) return f"rl:{endpoint}:{ip}" async def check(self, request: Request, limit: int, window: int = 60, endpoint: str = "") -> None: key = self._key(request, endpoint or request.url.path) now = time.time() cutoff = now - window pipe = self._r.pipeline() pipe.zremrangebyscore(key, "-inf", cutoff) pipe.zcard(key) pipe.zadd(key, {str(now): now}) pipe.expire(key, window + 1) _, count, *_ = pipe.execute() if count >= limit: raise HTTPException( status_code=429, detail={ "error": "rate_limit_exceeded", "message": "لقد تجاوزت الحد المسموح به من الطلبات. يرجى الانتظار قليلاً.", "retry_after": window, }, headers={"Retry-After": str(window)}, ) # ── In-memory limiter (fallback) ──────────────────────────────────────────── class _MemoryLimiter: """Sliding-window per (IP, endpoint). Thread-safe. Single-process only.""" def __init__(self) -> None: self._windows: dict[str, deque[float]] = defaultdict(deque) self._lock = threading.Lock() log.info("[ratelimit] memory mode (single-process)") def _key(self, request: Request, endpoint: str) -> str: forwarded = request.headers.get("X-Forwarded-For") ip = forwarded.split(",")[0].strip() if forwarded else ( request.client.host if request.client else "unknown" ) return f"{ip}:{endpoint}" async def check(self, request: Request, limit: int, window: int = 60, endpoint: str = "") -> None: key = self._key(request, endpoint or request.url.path) now = time.monotonic() cutoff = now - window with self._lock: dq = self._windows[key] while dq and dq[0] < cutoff: dq.popleft() if len(dq) >= limit: retry_after = int(window - (now - dq[0])) + 1 raise HTTPException( status_code=429, detail={ "error": "rate_limit_exceeded", "message": "لقد تجاوزت الحد المسموح به من الطلبات. يرجى الانتظار قليلاً.", "retry_after": retry_after, }, headers={"Retry-After": str(retry_after)}, ) dq.append(now) # ── Singleton: Redis → fakeredis → memory ────────────────────────────────── def _build_limiter(): if _REDIS_URL: try: return _RedisLimiter(_REDIS_URL) except Exception as e: log.warning("[ratelimit] Redis unavailable (%s) — trying fakeredis", e) try: import fakeredis server = fakeredis.FakeServer() fake_r = fakeredis.FakeRedis(server=server, decode_responses=True) fake_r.ping() limiter = _RedisLimiter.__new__(_RedisLimiter) limiter._r = fake_r log.info("[ratelimit] fakeredis mode (dev/single-process)") return limiter except Exception as fe: log.warning("[ratelimit] fakeredis unavailable (%s) — memory fallback", fe) return _MemoryLimiter() _limiter = _build_limiter() async def limit_analyze(request: Request) -> None: """5 analysis requests per minute per IP.""" await _limiter.check(request, limit=5, window=60, endpoint="analyze") async def limit_chat(request: Request) -> None: """30 chat requests per minute per IP.""" await _limiter.check(request, limit=30, window=60, endpoint="chat") async def limit_search(request: Request) -> None: """60 search requests per minute per IP.""" await _limiter.check(request, limit=60, window=60, endpoint="search")