((fileFsFullSizeInfo.getCapacity() - fileFsFullSizeInfo.getFree()) * 100.0) / fileFsFullSizeInfo.getCapacity();

            // Then
            assertEquals(25.0, percentFree, 0.001);
            assertEquals(75.0, percentUsed, 0.001);
        }

        @Test
        @DisplayName("Should differentiate between caller available and actual free units")

테스트를 다음 명령어로 실행하세요.

<div class="termy">

```console
$ pytest

================ test session starts ================
platform linux -- Python 3.6.9, pytest-5.3.5, py-1.8.1, pluggy-0.13.1
rootdir: /home/user/code/superawesome-cli/app
plugins: forked-1.1.3, xdist-1.31.0, cov-2.8.1
collected 6 items

---> 100%

예를 들어, Pydantic 모델을 사용하고 커스텀 `description`을 가진 상태 코드 `404` 응답을 선언할 수 있습니다.

또한 `response_model`을 사용하는 상태 코드 `200` 응답을 선언하되, 커스텀 `example`을 포함할 수도 있습니다:

{* ../../docs_src/additional_responses/tutorial003_py310.py hl[20:31] *}

이 모든 내용은 OpenAPI에 결합되어 포함되고, API 문서에 표시됩니다:

<img src="/img/tutorial/additional-responses/image01.png">

## 미리 정의된 응답과 커스텀 응답 결합하기 { #combine-predefined-responses-and-custom-ones }

例如，你可以声明一个状态码为 `404` 的响应，它使用一个 Pydantic 模型并带有自定义的 `description`。

以及一个状态码为 `200` 的响应，它使用你的 `response_model`，但包含自定义的 `example`：

{* ../../docs_src/additional_responses/tutorial003_py310.py hl[20:31] *}

所有这些都会被合并并包含到你的 OpenAPI 中，并显示在 API 文档里：

<img src="/img/tutorial/additional-responses/image01.png">

## 组合预定义响应和自定义响应 { #combine-predefined-responses-and-custom-ones }

            return 0;

        // Adjust based on error rate
        double errorRate = getErrorRate();
        if (errorRate > 0.1)
            score -= 50;
        else if (errorRate > 0.01)
            score -= 20;

        // Adjust based on interface capabilities
        score += localInterface.getScore() / 100;
        score += remoteInterface.getScore() / 100;

73.961746215820312,40.800731658935547,1,1.25,7,1,0.5,0,0,,,8.5,2,1,75,24,green,0.00,1.2,0.0,51,36,5.37,1267,168,1,Manhattan,016800,1016800,E,MN33,East Harlem South,3804,1902,193,1,Manhattan,019300,1019300,E,MN09,Morningside Heights,3802^3389241,2,2014-03-10 19:29:40,2014-03-10 19:53:09,N,1,-73.950843811035156,40.791805267333984,-73.938186645507812,40.843559265136719,1,4.57,19,1,0.5,0,0,,,20.5,2,1,75,244,green,0.00,1.2,0.0,51,36,5.37,1267,168,1,Manhattan,016800,1016800,E,MN33,East Harlem South,38...

    assertEquals('q', in.readChar());
    assertEquals('e', in.readChar());
    assertEquals('d', in.readChar());
  }

  public void testNewDataInput_readUnsignedShort() {
    byte[] data = {0, 0, 0, 1, (byte) 0xFF, (byte) 0xFF, 0x12, 0x34};
    ByteArrayDataInput in = ByteStreams.newDataInput(data);
    assertEquals(0, in.readUnsignedShort());
    assertEquals(1, in.readUnsignedShort());

  /**
   * Maximum allowed false positive probability of detecting a hash flooding attack given random
   * input.
   */
  @VisibleForTesting static final double HASH_FLOODING_FPP = 0.001;

  /**
   * Maximum allowed length of a hash table bucket before falling back to a j.u.HashMap based
   * implementation. Experimentally determined.
   */
  static final int MAX_HASH_BUCKET_LENGTH = 8;

    assertEquals('q', in.readChar());
    assertEquals('e', in.readChar());
    assertEquals('d', in.readChar());
  }

  public void testNewDataInput_readUnsignedShort() {
    byte[] data = {0, 0, 0, 1, (byte) 0xFF, (byte) 0xFF, 0x12, 0x34};
    ByteArrayDataInput in = ByteStreams.newDataInput(data);
    assertEquals(0, in.readUnsignedShort());
    assertEquals(1, in.readUnsignedShort());

    assertFalse(numberOrdering.isOrdered(asList(0, 5, 3, 9)));
    assertTrue(numberOrdering.isOrdered(asList(0, 3, 5, 9)));
    assertTrue(numberOrdering.isOrdered(asList(0, 0, 3, 3)));
    assertTrue(numberOrdering.isOrdered(asList(0, 3)));
    assertTrue(numberOrdering.isOrdered(singleton(1)));
    assertTrue(numberOrdering.isOrdered(Collections.<Integer>emptyList()));
  }