JSON Field Names Using Snake Case

[nabati]

Is it possible to define a model which uses camelCase for the attributes, but underscore/snake case when converted to json, or maybe even a field name that will be used for JSON only?

Currently;

var User = seq.define('user', {
    id: {
        type: Sequelize.INTEGER,
        primaryKey: true
    },
    shortLivedToken: {
        type: Sequelize.STRING,
        field: 'short_lived_token'
    },
    longLivedToken: {
        type: Sequelize.STRING,
        field: 'long_lived_token'
    }
}, {
    underscored: true,
    underscoredAll: true,
});

Turns into when calling response.json(user)

{
  "id": null,
  "shortLivedToken": "abcde",
  "longLivedToken": "fghji",
  "updated_at": "2015-11-21T18:32:20.181Z",
  "created_at": "2015-11-21T18:32:20.181Z"
}

And I want it to be

{
  "id": null,
  "short_lived_token": "abcde",
  "long_lived_token": "fghij",
  "updated_at": "2015-11-21T18:32:20.181Z",
  "created_at": "2015-11-21T18:32:20.181Z"
}

Any ideas?

[mickhansen]

Not possible. The attribute names are used for getting values, toJSON uses methods for getting values.
Overwrite toJSON or don't use camel cased attribute names.