Documentation questions about panko_serializer HOT 3 OPEN

Hey!

"new(...).to_json" / "new.serialize_to_json(...)" - will return a JSON serialized String, so if you do render json: { .. } you will get an output of object with value as string, this is something I don't think you expect. The way to solve this String issue, is with Panko::Response (option 5).

"new.serialize(...)" / "new.serialize(...).to_json" - serialize creates an array of hashes, together with to_json will return a JSON string on it. For performance reasons it's not recommended to use it, since:

• Panko iterates the users and convert the AR objects to Array of Hashes
• And another iteration is required to convert Array of Hashes to JSON
  while with serialize_to_json / to_json - You do it on one pass, See https://panko.dev/docs/design-choices#ojstringwriter for more explanation.

Let me know if you have more questions.

from panko_serializer.

Thanks! very clear.

from panko_serializer.

Hey @yosiat , thanks for this answer - I think it deserves a place in the documentation on its own :) What's the equivalent for single object serialization?

ObjectSerializer.new.serialize_to_json(object) is preferable over ObjectSerializer.new.serialize(object).to_json, right? In the "getting started" documentation I don't think it's clear which one is recommended.

I think existing documentation could greatly benefit from a page with common use cases, like:

• serialize a single object in a Rails controller;
• serialize an ActiveRecord collection in a Rails controller;
• serialize nested relations (e.g. User with many Comments) and eager loading;
• serialize an array of objects;
• serialize an array of objects using ActiveSupport::Cache as a cache backend;
• and so on.

I know most of these things are outside the scope of a serialization library, but they could give hints about possible ways of integrating Panko with the existing rails ecosystem for starters.

Thanks for the great work on this gem!

from panko_serializer.