`ServerSentEvent`는 `fastapi.sse`에서 임포트합니다:

{* ../../docs_src/server_sent_events/tutorial002_py310.py hl[4,26] *}

`data` 필드는 항상 JSON으로 인코딩됩니다. Pydantic 모델을 포함해 JSON으로 직렬화할 수 있는 값을 모두 전달할 수 있습니다.

## 원시 데이터 { #raw-data }

JSON 인코딩 없이 데이터를 보내야 한다면, `data` 대신 `raw_data`를 사용하세요.

미리 포맷된 텍스트, 로그 라인, 또는 `[DONE]`과 같은 특수한 <dfn title="특수한 조건이나 상태를 나타내는 데 사용되는 값">"센티널"</dfn> 값을 보낼 때 유용합니다.

    ArtifactResolutionResult resolve(ArtifactResolutionRequest request);

    // Install

    // Deploy

    // Map types of artifacts

    //
    // Raw file transfers
    //
    void publish(
            ArtifactRepository repository, File source, String remotePath, ArtifactTransferListener transferListener)
            throws ArtifactTransferFailedException;

    /** 32-bit status codes flag */
    int FLAGS2_STATUS32 = 0x4000;
    /** Unicode strings flag */
    int FLAGS2_UNICODE = 0x8000;

    /** No capabilities */
    int CAP_NONE = 0x0000;
    /** Raw mode capability */
    int CAP_RAW_MODE = 0x0001;
    /** MPX mode capability */
    int CAP_MPX_MODE = 0x0002;
    /** Unicode capability */
    int CAP_UNICODE = 0x0004;
    /** Large files capability */

Write throughput is typically at least 500MB/s per CPU core,
and scales with the number of available CPU cores.
Decompression speed is typically at least 1GB/s.

This means that in cases where raw IO is below these numbers
compression will not only reduce disk usage but also help increase system throughput.
Typically, enabling compression on spinning disk systems
will increase speed when the content can be compressed.

        }
        checkArgument(
            fromParameterizedType.getRawType().equals(toParameterizedType.getRawType()),
            "Inconsistent raw type: %s vs. %s",
            fromParameterizedType,
            to);
        Type[] fromArgs = fromParameterizedType.getActualTypeArguments();
        Type[] toArgs = toParameterizedType.getActualTypeArguments();

			readTestZip(t, zt)
		})
	}
}

func readTestZip(t *testing.T, zt ZipTest) {
	var z *Reader
	var err error
	var raw []byte
	if zt.Source != nil {
		rat, size := zt.Source()
		z, err = NewReader(rat, size)
		raw = make([]byte, size)
		if _, err := rat.ReadAt(raw, 0); err != nil {
			t.Errorf("ReadAt error=%v", err)
			return
		}
	} else {
		path := filepath.Join("testdata", zt.Name)

	}

	for _, op := range g.ops {
		db = op(db)
	}
	return db
}

func (c *g[T]) Raw(sql string, values ...interface{}) ExecInterface[T] {
	return execG[T]{g: &g[T]{
		db: c.db,
		ops: append(c.ops, func(db *DB) *DB {
			var r T
			return db.Model(r).Raw(sql, values...)
		}),
	}}
}

func (c *g[T]) Exec(ctx context.Context, sql string, values ...interface{}) error {

    val value = headers.value(i)

    when {
      name.equals("Cache-Control", ignoreCase = true) -> {
        if (headerValue != null) {
          // Multiple cache-control headers means we can't use the raw value.
          canUseHeaderValue = false
        } else {
          headerValue = value
        }
      }

      name.equals("Pragma", ignoreCase = true) -> {

从 `fastapi.sse` 导入 `ServerSentEvent`：

{* ../../docs_src/server_sent_events/tutorial002_py310.py hl[4,26] *}

`data` 字段始终会被编码为 JSON。你可以传入任何可被序列化为 JSON 的值，包括 Pydantic 模型。

## 原始数据 { #raw-data }

如果你需要发送**不**进行 JSON 编码的数据，请使用 `raw_data` 而不是 `data`。

这对于发送预格式化文本、日志行或特殊的 <dfn title="用于指示特殊条件或状态的值">"哨兵"</dfn> 值（例如 `[DONE]`）很有用。

{* ../../docs_src/server_sent_events/tutorial003_py310.py hl[17] *}

        });
    }

    /**
     * Gets the HTTP response containing the content of a dictionary file.
     * This method retrieves the raw file content from ConfigSync storage.
     *
     * @param dictFile the dictionary file to retrieve content for
     * @return a CurlResponse containing the file content
     */