FastAPI utiliza um padrão para construir frameworks e servidores web em Python chamado <abbr title="Asynchronous Server Gateway Interface">ASGI</abbr>. FastAPI é um framework web ASGI.

A principal coisa que você precisa para executar uma aplicação **FastAPI** (ou qualquer outra aplicação ASGI) em uma máquina de servidor remoto é um programa de servidor ASGI como o **Uvicorn**, que é o que vem por padrão no comando `fastapi`.

    }
    sink1.writeUtf8("def")
    sink1.close()
    assertFailsWith<ConnectionShutdownException> {
      connection.newStream(headerEntries("c", "cola"), true)
    }
    assertThat(stream1.isOpen).isTrue()
    assertThat(stream2.isOpen).isFalse()
    assertThat(connection.openStreamCount()).isEqualTo(1)

    // Verify the peer received what was expected.

*   Include unit tests when you contribute new features, as they help to a)
    prove that your code works correctly, and b) guard against future breaking
    changes to lower the maintenance cost.
*   Bug fixes also generally require unit tests, because the presence of bugs
    usually indicates insufficient test coverage.
*   Keep API compatibility in mind when you change code in core TensorFlow,
    e.g., code in

        as described in Section 3.1, and You must inform recipients of the
        Executable Form how they can obtain a copy of such Source Code Form by
        reasonable means in a timely manner, at a charge no more than the cost
        of distribution to the recipient; and

     b. You may distribute such Executable Form under the terms of this License,
        or sublicense it under different terms, provided that the license for

        as described in Section 3.1, and You must inform recipients of the
        Executable Form how they can obtain a copy of such Source Code Form by
        reasonable means in a timely manner, at a charge no more than the cost
        of distribution to the recipient; and

     b. You may distribute such Executable Form under the terms of this
        License, or sublicense it under different terms, provided that the

  DEVICE_PLACEMENT_EXPLICIT = 0,
  // Copy the tensor to the right device but log a warning.
  DEVICE_PLACEMENT_WARN = 1,
  // Silently copy the tensor, which has a performance cost since the operation
  // will be blocked till the copy completes. This is the default policy.
  DEVICE_PLACEMENT_SILENT = 2,
  // Placement policy which silently copies int32 tensors but not other dtypes.

Ao validar os dados com o Pydantic nós podemos garantir que temos, por exemplo, exatamente uma `list` de `str` com os escopos e uma `str` com o `username`.

No lugar de, por exemplo, um `dict`, ou alguma outra coisa, que poderia quebrar a aplicação em algum lugar mais tarde, tornando isso um risco de segurança.

    for (int i = 0; i < length; i++) {
      checkElementNotNull(array[i], i);
    }
    return array;
  }

  // We do this instead of Preconditions.checkNotNull to save boxing and array
  // creation cost.
  @CanIgnoreReturnValue
  static Object checkElementNotNull(@CheckForNull Object element, int index) {
    if (element == null) {
      throw new NullPointerException("at index " + index);
    }
    return element;

import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.concurrent.atomic.AtomicBoolean;

import org.apache.pdfbox.cos.COSInputStream;
import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDDocumentInformation;
import org.apache.pdfbox.pdmodel.PDDocumentNameDictionary;

    for (int i = 0; i < length; i++) {
      checkElementNotNull(array[i], i);
    }
    return array;
  }

  // We do this instead of Preconditions.checkNotNull to save boxing and array
  // creation cost.
  @CanIgnoreReturnValue
  static Object checkElementNotNull(Object element, int index) {
    if (element == null) {
      throw new NullPointerException("at index " + index);
    }
    return element;
  }