This is part 2 of the ‘TurboRedis By Example’ tutorial. I recommend reading part 1 if you haven’t done so yet.
We’re back. And now that we have the boilerplate stuff done, it’s time
for a more useful example.
This example is a little application that lists a GitHub user’s repositories
in an HTML list.
It only uses a few basic Redis commands: GET, TTL and SETEX.
The list of repositories is retrieved from the public GitHub API and the responses are cached in Redis for 10 seconds. It will also print bold text under the repositories list indicating wether it was a cache hit or a cache miss. If it was a cache-hit, the remaining time-to-live for the Redis key is also printed.
Things to take away from this example:
- Commands can be called with
corountine.yield()(which usesturbo.async.taskunder the hood) or with traditional callbacks. - TurboRedis auto-converts replies from many commands. In this example
the time-to-live value is returned as an integer.
This behaviour is somewhat inconsistent at the moment and can be
disabled completely by passing
purist=truetoturboredis.Connection:new(). - When using callbacks with TurboRedis you have to call
self:set_async(true)in the RequestHandler to prevent Turbo from finishing the request before you intend to. This also means that you have to:finish()when the request is done. - Host and/or port can be passed to
turboredis.Connection:new()in a table. - TurboRedis never touches arguments to commands, it only converts them to strings, so all command-functions uses the same argument order as described in the Redis Command Reference.
That’s it. Part 3 will be covering Publish/Subscribe.
If you’re bored to death and have nothing else to do, or just want more examples, you can check out the Git Repo or the TurboRedis docs.