///

### Pydantic v1 e v2 na mesma aplicação { #pydantic-v1-and-v2-on-the-same-app }

Não é suportado pelo Pydantic ter um modelo do Pydantic v2 com campos próprios definidos como modelos do Pydantic v1, ou vice-versa.

```mermaid
graph TB
    subgraph "❌ Not Supported"
        direction TB
        subgraph V2["Pydantic v2 Model"]
            V1Field["Pydantic v1 Model"]
        end

   *       instances.
   *   <li>For the constructor or static factory method used to construct instances, it's checked
   *       that when equal parameters are passed, the result instance should also be equal; and vice
   *       versa.
   *   <li>If a non-private constructor or non-private static factory method exists:
   *       <ul>
   *         <li>Test will fail if default value for a parameter cannot be determined.

Como parte de eso, necesitaba investigar, probar y usar muchas alternativas.

La historia de **FastAPI** es en gran parte la historia de sus predecesores.

Como se dice en la sección [Alternativas](alternatives.md){.internal-link target=_blank}:

<blockquote markdown="1">

**FastAPI** no existiría si no fuera por el trabajo previo de otros.

Si no lo recibe, devuelve un error HTTP 401 "Unauthorized".

Y devuelve un header `WWW-Authenticate` con un valor de `Basic`, y un parámetro `realm` opcional.

Eso le dice al navegador que muestre el prompt integrado para un nombre de usuario y contraseña.

Luego, cuando escribes ese nombre de usuario y contraseña, el navegador los envía automáticamente en el header.

   * have the provided key, an empty collection is returned.
   *
   * <p>Changes to the returned collection will update the underlying multimap, and vice versa.
   *
   * <p>Because a {@code SortedSetMultimap} has unique sorted values for a given key, this method
   * returns a {@link SortedSet}, instead of the {@link java.util.Collection} specified in the

   * the table lacks a mapping for a given row and column, the corresponding array element is null.
   *
   * <p>Subsequent table changes will not modify the array, and vice versa.
   *
   * @param valueClass class of values stored in the returned array
   */
  @GwtIncompatible // reflection
  public @Nullable V[][] toArray(Class<V> valueClass) {

   */
  public abstract double transform(double x);

  /**
   * Returns the inverse linear transformation. The inverse of a horizontal transformation is a
   * vertical transformation, and vice versa. The inverse of the {@link #forNaN} transformation is
   * itself. In all other cases, the inverse is a transformation such that applying both the

     * clone. The only requirement is that invocations of {@link #addRepository(Repository)} on the clone do not affect
     * the state of the original resolver and vice versa.
     *
     * @return The cloned resolver, never {@code null}.
     */
    ModelResolver newCopy();

    name = "is_cuda_enabled",
    actual = if_oss(
        "@local_config_cuda//:is_cuda_enabled",
        "@local_config_cuda//cuda:using_config_cuda",
    ),
)

# Config setting that is satisfied when CUDA device code should be compiled
# with clang. It does not imply that CUDA support has been enabled.
alias(
    name = "is_cuda_compiler_clang",
    actual = if_oss(
        "@local_config_cuda//:is_cuda_compiler_clang",

///

### Pydantic v1 and v2 on the same app { #pydantic-v1-and-v2-on-the-same-app }

It's **not supported** by Pydantic to have a model of Pydantic v2 with its own fields defined as Pydantic v1 models or vice versa.

```mermaid
graph TB
    subgraph "❌ Not Supported"
        direction TB
        subgraph V2["Pydantic v2 Model"]
            V1Field["Pydantic v1 Model"]
        end