Persisting lambdas for object-specific logic in Ruby and Rails

Stop me if you’ve heard this one before. You’re building your app, everything is going swimmingly, then BAM. New requirement – each instance of a particular object needs its own implementation of something. In my case, each instance of some object may want to structure an API request that it makes a little bit differently (including some optional fields and not others depending on state, building different values for some fields the API expects, etc.).

How do you do this? Well, I’m sure there are several ways. This is the one I’ve found works the best for my uses. As always, YMMV.

Essentially, we’re storing lambdas as strings of code. Decide on a signature for the blocks you’ll be using (which params they expect). Here’s mine:

The comment1 and comment2 fields are the ones that the PayPal API expects – the blocks specify how to build these for all products that the seller is selling. Since they’re just strings, we can serialize the whole hash out to DB and load it without issues.

When it comes time to process a payment for one of the seller’s products, you can see how simple it is:

This is obviously an extremely simple example, but it generalises to any situation that involves custom logic for particular object instances. What I haven't mentioned here is the security of `eval`ing whatever lives in your database as Ruby code. Don't do this in production code, obviously, but a cool example of the flexibility of my beloved Ruby.

Happy coding!