Hello,

Thanks a lot for this article 👏

Some though: IMHO, it's better if cache is agnostic. I mean if developers doesn't have to worry about how and what is cached while developing api.

Another thing: I don't understand why you put the cache retrieve code in a middleware. I might be wrong, but right now, if I query GET /lmao/:username, and if the cache exist, looks like it will return me the same response as GET /repos/:username.

So why not putting all the cache logic inside the cached route. I.E

app.get('/repos/:username', cache, async (req, res) => {
  const { username } = req.params

  client.get(username, (err, data) => {
    if (err) throw err
    if (data !== null) {
      res.send(data)
    } else {
      try {
        const response = await fetch(`api.github.com/users${username}/repos`)
        const data = await response.json()
        let repos = data.map(item => item.name)

        // save the repos to redis using the username as key and with an expiration
        client.setex(username, 60, repos)
        res.send(repos)
      } catch (err) {
        res.status(500)
      }
    }
  })
});