---

FastAPI is a modern, fast (high-performance), web framework for building APIs with Python based on standard Python type hints.

The key features are:

* **Fast**: Very high performance, on par with **NodeJS** and **Go** (thanks to Starlette and Pydantic). [One of the fastest Python frameworks available](#performance).
* **Fast to code**: Increase the speed to develop features by about 200% to 300%. *

fast-deep-equal@^3.1.1:
  version "3.1.3"
  resolved "https://registry.yarnpkg.com/fast-deep-equal/-/fast-deep-equal-3.1.3.tgz#3a7d56b559d6cbc3eb512325244e619a65c6c525"
  integrity sha512-f3qQ9oQy9j2AhBe/H9VC91wLmKBCCU/gDOnKNAYG5hswO7BLKj09Hc5HYNz9cGI++xlpDCIgDaitVs03ATR84Q==

fast-json-stable-stringify@^2.0.0:
  version "2.1.0"

      logger.info("served $requestLine")
      return MockResponse(code = HttpURLConnection.HTTP_NOT_FOUND)
    }

    if (failFastResponse != null && responseQueue.peek() == null) {
      // Fail fast if there's no response queued up.
      return failFastResponse!!
    }

    val result = responseQueue.take()

    // If take() returned because we're shutting down, then enqueue another dead letter so that any

 * support even a "normal" `TypeParameter<T>` when `<T>` has a nullable bound. (See the discussion
 * on TypeToken.where.) So, in the interest of failing fast and encouraging the user to switch to a
 * non-null bound if possible, let's require a non-null bound here.
 *
 * TODO(cpovirk): Elaborate on "wouldn't behave as users might expect."
 */

run:
  # Timeout for analysis, e.g. 30s, 5m.
  # Default: 1m
  timeout: 20m
  build-tags:
  - integ
  - integfuzz
linters:
  disable-all: true
  enable:
  - goimports
  - gofumpt
  - gci
  fast: false
linters-settings:
  gci:
    sections:
      - standard # Captures all standard packages if they do not match another section.
      - default # Contains all imports that could not be matched to another section type.

      matrix:
        python-version:
          - "3.12"
          - "3.11"
          - "3.10"
          - "3.9"
          - "3.8"
        pydantic-version: ["pydantic-v1", "pydantic-v2"]
      fail-fast: false
    steps:
      - name: Dump GitHub context
        env:
          GITHUB_CONTEXT: ${{ toJson(github) }}
        run: echo "$GITHUB_CONTEXT"
      - uses: actions/checkout@v4
      - name: Set up Python

Nesse caso, se você tivesse **múltiplos contêineres**, por padrão, quando o Prometheus fosse **ler as métricas**, ele receberia as métricas de **um único contêiner cada vez** (para o contêiner que tratou essa solicitação específica), em vez de receber as **métricas acumuladas** de todos os contêineres replicados.

    strategy:
      matrix:
        os: [ubuntu-latest, windows-latest]
        java: [17, 21]
      fail-fast: false

    runs-on: ${{ matrix.os }}

    steps:
      - uses: actions/checkout@v4
        with:
          persist-credentials: false

      - uses: actions/setup-java@v4
        with:

permissions:
  contents: read

jobs:
  build:
    if: github.repository == 'tensorflow/tensorflow' # Don't do this in forks
    runs-on: [self-hosted, linux, ARM64]
    strategy:
      fail-fast: false
      matrix:
        pyver: ['3.9', '3.10', '3.11', '3.12']
    steps:
      - name: Stop old running containers (if any)
        shell: bash
        run: |
          running_containers=$(docker ps -q) && \

```

Ao declarar o parâmetro com o tipo sendo um `Request` em sua *função de operação de rota*, o **FastAPI** saberá como passar o `Request` neste parâmetro.

/// tip | "Dica"

Note que neste caso, nós estamos declarando o parâmetro da rota ao lado do parâmetro da requisição.

Assim, o parâmetro da rota será extraído, validado, convertido para o tipo especificado e anotado com OpenAPI.