Question

我需要在不对对象进行序列化的情况下从Azure函数流式传输JSON，因为序列化开销太高。代码可以归结为这样：

[FunctionName("AzureFunction")]
public static async Task<IActionResult> 
Run([HttpTrigger(AuthorizationLevel.Anonymous, "get", Route = "v1/function")], HttpRequest req, ILogger log)
{
    MemoryStream outputStream = new MemoryStream();
    StreamWriter writer = new StreamWriter(outputStream);
    JsonWriter jsonWriter = new JsonTextWriter(writer);

    jsonWriter.WriteStartObject();
    jsonWriter.WritePropertyName("Property");
    jsonWriter.WriteValue("Value");
    jsonWriter.WriteEndObject();

    outputStream.Seek(0, SeekOrigin.Begin);
    return new FileStreamResult(outputStream, "application/json");
}

但是，此JSON在某些时候被击中之前已被转换为字符串。我已经能够将来自另一个HTTP调用的流直接传递到FileStreamResult对象，并且该对象不会转换为字符串，因此我确定这是可能的。

Answer 1

我知道了，返回之前需要刷新Stream。正确的代码如下所示：

[FunctionName("AzureFunction")]
public static async Task<IActionResult> 
Run([HttpTrigger(AuthorizationLevel.Anonymous, "get", Route = "v1/function")], HttpRequest req, ILogger log)
{
    MemoryStream outputStream = new MemoryStream();
    StreamWriter writer = new StreamWriter(outputStream);
    JsonWriter jsonWriter = new JsonTextWriter(writer);

    jsonWriter.WriteStartObject();
    jsonWriter.WritePropertyName("Property");
    jsonWriter.WriteValue("Value");
    jsonWriter.WriteEndObject();

    jsonWriter.Flush(); // Flush the json before returning.
    outputStream.Seek(0, SeekOrigin.Begin);
    return new FileStreamResult(outputStream, "application/json");
}

.NET Azure函数-如何从流返回JSON对象？

1 个答案: