redis-node-client

A Redis client implementation for Node.js which runs atop Google V8.

This project lets you access a Redis instance using server-side JavaScript.

Asynchronicity

Node.js does not block on I/O operations.

This means that while a typical Redis client might have code that accesses a
Redis server in a blocking call, Node.js-based code cannot.

Typical Redis client (e.g. Python):

foo = client.get('counter')

This Node.js-based Redis client:

var sys = require("sys");
var redis = require("./redis");

var client = new redis.Client();
client.connect(learn_to_count);

function learn_to_count () {
    client.incr('counter').addCallback(function (value) {
        sys.puts("counter is now " + value);
        client.close();
    });
}

Running this example, we’d see:

$ node counter-example.js
counter is now 1
$ node counter-example.js
counter is now 2
$ node counter-example.js
counter is now 3

That is, you must supply a callback function that is called when Redis returns,
even if Redis queries are extremely fast.

A potential upside to this slightly awkward requirement is that you can enjoy
the benefits of pipelining many Redis queries in a non-blocking way. Redis
returns replies for requests in the order received.

See the test.js
file as a good example of this.

Status

• The full Redis 1.1 command specification is supported.
• All tests pass using Redis HEAD at commit (09f6f7020952cd93e178da11e66e36f8a98398d1; Dec 1 2009)
• All tests pass using Node.js v0.1.20-3-g5b1a535 (Dec 1 2009)
• See the TODO file for known issues.

Testing

To test:

1. fire up redis-server on 127.0.0.1:6379 (the default)
2. install node.js
3. run node test.js

Author

Brian Hammond, Fictorial (brian at fictorial dot com)

Copyright

Copyright © 2009 Fictorial LLC

License

See LICENSE (it’s MIT; go nuts).