I’m working on an Express.js that uses Sequelize as the ORM, my Character model is declared like this on models/character_model.js:

class Character extends Sequelize.Model{
  //fields and stuff
}

And I have model creation code that I put directly in the controller:

async create (req, res) {
  await models.Character.findOrCreate({
    where: { name: req.body.name },
    defaults: {
      //some fields
    }
  })
  //other code
  return res.status(200).json(data)
}

This seems ok to me at least, seems like the kind of code that should be in the controller, right?

But then I do this

const { Op } = require('sequelize')

async search (req, res) {
  let character = await models.Character.findOne({
    where: {
      name: { [Op.iLike]: `%${req.params.name}%` }
    }
  })
  //other code
  return res.status(200).json(data)
}

And that to me seems like code that should be somewhere else, it’s not separating the concerns very well in my opinion, I don’t think I should have that much database logic in the same place where I put endpoint responses. Also, I’m importing stuff from sequelize here and in models, instead of having all of it contained in one file.

What do you think? Should I put that findOne call somewhere else? Maybe a separate file (a utils file or something, or maybe some character_manager.js file in the models folder), or make a method inside the Character class, or maybe in the same file, but as a different export.

And if I’m putting that code somewhere else, maybe I should do the same for the findOrCreate call as well, right? That way the controller doesn’t do any sequelize logic.

I really need to follow the best practices for this project.