Par exemple, vous ne pouvez pas mettre un modèle Pydantic dans une `JSONResponse` sans d'abord le convertir en un `dict` avec tous les types de données (comme `datetime`, `UUID`, etc.) convertis en types compatibles avec JSON.

Pour ces cas, vous pouvez spécifier un appel à `jsonable_encoder` pour convertir vos données avant de les passer à une réponse :

```Python hl_lines="6-7  21-22"

  public static short fromBytes(byte b1, byte b2) {
    return (short) ((b1 << 8) | (b2 & 0xFF));
  }

  private static final class ShortConverter extends Converter<String, Short>
      implements Serializable {
    static final Converter<String, Short> INSTANCE = new ShortConverter();

    @Override
    protected Short doForward(String value) {
      return Short.decode(value);
    }

    @Override

  public static short fromBytes(byte b1, byte b2) {
    return (short) ((b1 << 8) | (b2 & 0xFF));
  }

  private static final class ShortConverter extends Converter<String, Short>
      implements Serializable {
    static final Converter<String, Short> INSTANCE = new ShortConverter();

    @Override
    protected Short doForward(String value) {
      return Short.decode(value);
    }

    @Override

/// tip | Consejo

`JSONResponse` en sí misma es una subclase de `Response`.

///

Y cuando devuelves una `Response`, **FastAPI** la pasará directamente.

No hará ninguna conversión de datos con modelos Pydantic, no convertirá el contenido a ningún tipo, etc.

Esto te da mucha flexibilidad. Puedes devolver cualquier tipo de dato, sobrescribir cualquier declaración de datos o validación, etc.

## Usando el `jsonable_encoder` en una `Response`

    return b1 << 24 | (b2 & 0xFF) << 16 | (b3 & 0xFF) << 8 | (b4 & 0xFF);
  }

  private static final class IntConverter extends Converter<String, Integer>
      implements Serializable {
    static final Converter<String, Integer> INSTANCE = new IntConverter();

    @Override
    protected Integer doForward(String value) {
      return Integer.decode(value);
    }

        "the total number of elements (%s) in the arrays must fit in an int",
        result);
    return (int) result;
  }

  private static final class FloatConverter extends Converter<String, Float>
      implements Serializable {
    static final Converter<String, Float> INSTANCE = new FloatConverter();

    @Override
    protected Float doForward(String value) {
      return Float.valueOf(value);
    }

    @Override

        "the total number of elements (%s) in the arrays must fit in an int",
        result);
    return (int) result;
  }

  private static final class DoubleConverter extends Converter<String, Double>
      implements Serializable {
    static final Converter<String, Double> INSTANCE = new DoubleConverter();

    @Override
    protected Double doForward(String value) {
      return Double.valueOf(value);
    }

        "the total number of elements (%s) in the arrays must fit in an int",
        result);
    return (int) result;
  }

  private static final class FloatConverter extends Converter<String, Float>
      implements Serializable {
    static final Converter<String, Float> INSTANCE = new FloatConverter();

    @Override
    protected Float doForward(String value) {
      return Float.valueOf(value);
    }

    @Override

        "the total number of elements (%s) in the arrays must fit in an int",
        result);
    return (int) result;
  }

  private static final class DoubleConverter extends Converter<String, Double>
      implements Serializable {
    static final Converter<String, Double> INSTANCE = new DoubleConverter();

    @Override
    protected Double doForward(String value) {
      return Double.valueOf(value);
    }

    * Query parameters.
    * Cookies.
    * Headers.
    * Form.
    * File.
* <abbr title="detta anche: serialization, parsing, marshalling">Conversione</abbr> dei dati di output: converte dati e tipi di Python a dati per la rete (come JSON):
    * Converte i tipi di Python (`str`, `int`, `float`, `bool`, `list`, ecc).
    * Oggetti `datetime`.
    * Oggetti `UUID`.
    * Modelli del database.
    * ...e molto di più.