烧瓶缓存:基于模式列出密钥? [英] flask cache: list keys based on a pattern?

问题描述

我将Flask Cache插件与Redis用作后端来缓存我的API响应.假设我有API可以吸引用户并像这样创建用户:

I'm using the Flask Cache plugin with Redis as backend to cache my API response. Say I have APIs to get users and create user like this:

/api/users?page=1  GET
/api/users         POST

将以完整URL作为键来缓存GET结果.创建新用户后，我想删除所有以/api/users 开头的键-当前，我正在执行 cache.clear()，但几乎没有似乎是必要的.

The GET result will be cached with full URL as key. When a new user is created, I'd like to delete all keys that start with /api/users - currently I'm doing cache.clear() but it hardly seems necessary.

但是，我似乎找不到用于获取键列表的API.使用 redis-py 时，有一个 keys(* pattern) API用于此目的.Flask Cache是​​否有类似的API?

However, I can't seem to find an API to get a list of keys. With redis-py, there is a keys(*pattern) API for that purpose. Is there a similar API for Flask Cache?

推荐答案

Flask-Cache看起来已经过时了.您可以切换到烧瓶缓存，它是Flask-Cache的分支.

Flask-Cache looks outdated. You can switch to Flask-Caching which is maintained fork of Flask-Cache.

按模式删除Redis键

您有两个选择:

1. 使用 redis 软件包分别连接到Redis.使用程序包提供的核心命令根据模式搜索键，然后将其删除.
2. 使用Flask-Cache的受保护方法/烧瓶缓存可以访问底层Redis对象，并使用该对象上的核心命令来基于模式搜索键并将其删除.

1. Connect to Redis separately using the redis package. Use core commands provided by the package to search for keys based on a pattern, and delete them.
2. Use protected methods of the Flask-Cache / Flask-Caching to get access to the underlying Redis object, and use core commands on that object to search for keys based on a pattern, and delete them.

选项1:与Redis的单独连接

# python utility for redis
import redis

r = redis.Redis(host=REDIS_HOST, port=REDIS_PORT, db=REDIS_DB, password=REDIS_PASSWORD)

def delete_pattern(pattern: str) -> int:
    """Remove all keys matching pattern.
    """
    count = 0
    for key in r.scan_iter(pattern):
        r.delete(key)
        count += 1
    return count

# pass pattern to delete
CACHE_URL_PAT = "flask_cache*auth*"
delete_pattern(CACHE_URL_PAT)

选项2:使用受保护的方法来访问底层的Redis连接对象

⚠️即使此方法工作正常，但这些都是未记录的方法.我在GitHub上浏览了源代码来创建此功能.仅针对CACHE_TYPE ='redis'

⚠️ Even though this works fine, but these are undocumented methods. I went through the source code on GitHub to create this function. Tested only for CACHE_TYPE = 'redis'

# initialised cache object stored in extensions file
from app.extensions import cache

def delete_pattern(pattern):
    status = False
    binary_keys = cache.cache._read_clients.keys(pattern)
    keys = [k.decode("utf-8", errors="ignore") for k in binary_keys if k]
    if keys:
        status = cache.cache._write_client.delete(*keys)
    return status

# pass pattern to delete
CACHE_URL_PAT = "flask_cache*auth*"
delete_pattern(CACHE_URL_PAT)

---

注意:在Flask-Caching中， flask_cache _ 是默认的 CACHE_KEY_PREFIX .如果您对 CACHE_KEY_PREFIX 使用了其他值，请使用该值(而不是 flask_cache _ )作为搜索模式的前缀.

---

NOTE : flask_cache_ is default CACHE_KEY_PREFIX in Flask-Caching. If you have used some other value for CACHE_KEY_PREFIX, please use that (instead of flask_cache_) as the prefix for your search pattern.

这篇关于烧瓶缓存:基于模式列出密钥?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！