Hug a été l'un des premiers frameworks à implémenter la déclaration des types de paramètres d'API en utilisant les type hints Python. C'était une excellente idée qui a inspiré d'autres outils à faire de même.

Il utilisait des types personnalisés dans ses déclarations au lieu des types Python standard, mais c'était tout de même un énorme pas en avant.

 */
package okhttp3.survey

import java.security.Security
import okhttp3.Cache
import okhttp3.OkHttpClient
import okhttp3.survey.ssllabs.SslLabsClient
import okhttp3.survey.types.Client
import okhttp3.survey.types.SuiteId
import okio.FileSystem
import okio.Path.Companion.toPath
import org.conscrypt.Conscrypt

@Suppress("ktlint:standard:property-naming")
suspend fun main() {
  val includeConscrypt = false

 * Maven provides implementations of those types and may inject them in plugins.
 * <p>
 * A type can be marked {@link Consumer} or {@link Provider} but not both. A type is assumed to be
 * {@link Consumer} if it is not marked either {@link Consumer} or {@link Provider}.
 * <p>
 * A package can be marked {@link Provider}. In this case, all types in the package are considered

     * <p>
     * Important: this "static" list of types should be in-sync with core provided types.
     */
    protected ArtifactTypeRegistry getArtifactTypeRegistry() {
        DefaultArtifactTypeRegistry stereotypes = new DefaultArtifactTypeRegistry();
        new DefaultTypeProvider().types().forEach(stereotypes::add);
        return stereotypes;
    }

#include "tensorflow/c/tf_datatype.h"
#include "tensorflow/c/tf_status.h"
#include "tensorflow/core/framework/tensor_shape.h"
#include "tensorflow/core/platform/casts.h"
#include "tensorflow/core/platform/types.h"

namespace tensorflow {

// Represents the results of the execution of an operation.
struct OutputList {
  std::vector<AbstractTensorHandle*> outputs;
  int expected_num_outputs = -1;
};

     * ({@value #JAR} or {@value #CLASSPATH_JAR} types), then the test artifact will also be placed on the class-path.
     * Otherwise, if the main artifact is placed on the module-path ({@value #JAR} or {@value #MODULAR_JAR} types),
     * then the test artifact will be added using {@code --patch-module} option.
     */
    String TEST_JAR = "test-jar";

{!> ../../docs_src/body/tutorial001.py!}
```

////

## Erstellen Sie Ihr Datenmodell

Dann deklarieren Sie Ihr Datenmodell als eine Klasse, die von `BaseModel` erbt.

Verwenden Sie Standard-Python-Typen für die Klassenattribute:

//// tab | Python 3.10+

```Python hl_lines="5-9"
{!> ../../docs_src/body/tutorial001_py310.py!}
```

////

//// tab | Python 3.8+

```Python hl_lines="7-11"

And these models are all sharing a lot of the data and duplicating attribute names and types.

We could do better.

We can declare a `UserBase` model that serves as a base for our other models. And then we can make subclasses of that model that inherit its attributes (type declarations, validation, etc).

  pymdownx.tilde:

  # pymdownx blocks
  pymdownx.blocks.admonition:
    types:
    - note
    - attention
    - caution
    - danger
    - error
    - tip
    - hint
    - warning
    # Custom types
    - info
    - check
  pymdownx.blocks.details:
  pymdownx.blocks.tab:
    alternate_style: True

  # Other extensions

    when (type) {
      // Special case types that have 0 or 1 flag.
      TYPE_SETTINGS, TYPE_PING -> return if (flags == FLAG_ACK) "ACK" else BINARY[flags]
      TYPE_PRIORITY, TYPE_RST_STREAM, TYPE_GOAWAY, TYPE_WINDOW_UPDATE -> return BINARY[flags]
    }
    val result = if (flags < FLAGS.size) FLAGS[flags]!! else BINARY[flags]
    // Special case types that have overlap flag values.
    return when {