RecID3 int64 `gorm:"column:recid3;type:numeric(9,2);not null" json:"recid3"`
		}
		expectedSql := []string{
			`ALTER TABLE "migrate_decimal_columns" ALTER COLUMN "recid1" TYPE numeric(8) USING "recid1"::numeric(8)`,
			`ALTER TABLE "migrate_decimal_columns" ALTER COLUMN "recid2" TYPE numeric(9,1) USING "recid2"::numeric(9,1)`,

        // Test registration with null/empty values
        String regId1 = mockService.registerWitness(null, "192.168.1.100", 1);
        assertNotNull(regId1);

        String regId2 = mockService.registerWitness("", "192.168.1.100", 1);
        assertNotNull(regId2);

        String regId3 = mockService.registerWitness("\\\\server\\share", null, 1);
        assertNotNull(regId3);

        assertEquals(3, mockService.getRegistrationCount());

///

## Crear un middleware { #create-a-middleware }

Para crear un middleware usas el decorador `@app.middleware("http")` encima de una función.

La función middleware recibe:

* La `request`.
* Una función `call_next` que recibirá la `request` como parámetro.
    * Esta función pasará la `request` a la correspondiente *path operation*.

Para eso, **FastAPI** proporciona una función `jsonable_encoder()`.

## Usando el `jsonable_encoder` { #using-the-jsonable-encoder }

Imaginemos que tienes una base de datos `fake_db` que solo recibe datos compatibles con JSON.

Por ejemplo, no recibe objetos `datetime`, ya que no son compatibles con JSON.

Entonces, un objeto `datetime` tendría que ser convertido a un `str` que contenga los datos en [formato ISO](https://en.wikipedia.org/wiki/ISO_8601).

          "format": "time_series",
          "instant": false,
          "interval": "",
          "intervalFactor": 1,
          "legendFormat": "{{bucket,range}}",
          "refId": "A",
          "step": 300
        }
      ],
      "title": "Object Size Distribution",
      "type": "bargauge"
    },
    {
      "datasource": {
        "type": "prometheus",

          "instant": true,
          "interval": "",
          "intervalFactor": 1,
          "legendFormat": "{{instance}}",
          "metric": "process_start_time_seconds",
          "refId": "A",
          "step": 60
        }
      ],
      "title": "Uptime",
      "type": "stat"
    },
    {
      "datasource": {
        "type": "prometheus",
        "uid": "${DS_PROMETHEUS}"

          "hide": false,
          "instant": true,
          "interval": "",
          "intervalFactor": 1,
          "legendFormat": "",
          "metric": "process_start_time_seconds",
          "refId": "A",
          "step": 60
        }
      ],
      "title": "Total Drives",
      "type": "stat"
    },
    {
      "datasource": {
        "type": "prometheus",

import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;

/**
 * Use this to meta-annotate {@code XxxFeature.Require} annotations, so that those annotations can
 * be used to decide whether to apply a test to a given class-under-test.
 *
 * <p>This is needed because annotations can't implement interfaces, which is also why reflection is
 * used to extract values from the properties of the various annotations.

import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;

/**
 * Use this to meta-annotate {@code XxxFeature.Require} annotations, so that those annotations can
 * be used to decide whether to apply a test to a given class-under-test.
 *
 * <p>This is needed because annotations can't implement interfaces, which is also why reflection is
 * used to extract values from the properties of the various annotations.

      render: shell
    validations:
      required: true
  - type: textarea
    id: what-did-you-do
    attributes:
      label: "What did you do?"
      description: "If possible, provide a recipe for reproducing the error. A complete runnable program is good. A link on [go.dev/play](https://go.dev/play) is better. A failing unit test is the best."
    validations:
      required: true
  - type: textarea