优化 HttpWebResponse - GetResponse

发布于 2024-08-21 05:47:42 字数 813 浏览 4 评论 0原文

我使用以下代码行来读取异步 HttpWebRequest 的响应。这似乎是特定操作中花费的最大时间。这里有什么我可以优化的吗？

System.Net.HttpWebResponse oResp =(System.Net.HttpWebResponse)oReq.EndGetResponse(oResult);
oResp = (HttpWebResponse)oReq.GetResponse();
StreamReader oStreamReader = new StreamReader(oResp.GetResponseStream());
string sResponse = oStreamReader.ReadToEnd();

...继续创建 XmlDocument，向其中附加更多 XML，然后执行 XSL 转换。

创建连接：

HttpWebRequest oReq;
oReq = (HttpWebRequest)WebRequest.Create(sUrl + sQueryString);
oReq.ContentType = sContentType;
oReq.Method = "POST";
oReq.ContentLength = aBytes.Length;
Stream oStream = oReq.GetRequestStream();
oStream.Write(aBytes, 0, aBytes.Length);
oStream.Close();
AsyncState oState = new AsyncState(oReq);
return oReq.BeginGetResponse(fCallBack, oState);

原文

I'm using the following lines of code to read the response of an asynchronous HttpWebRequest. This seems to be the largest amount of time spent in a particular operation. Is there anything I can optimize here?

System.Net.HttpWebResponse oResp =(System.Net.HttpWebResponse)oReq.EndGetResponse(oResult);
oResp = (HttpWebResponse)oReq.GetResponse();
StreamReader oStreamReader = new StreamReader(oResp.GetResponseStream());
string sResponse = oStreamReader.ReadToEnd();

...goes on to make an XmlDocument, append some more XML to it, then perform an XSL transform.

Creating the Connections:

HttpWebRequest oReq;
oReq = (HttpWebRequest)WebRequest.Create(sUrl + sQueryString);
oReq.ContentType = sContentType;
oReq.Method = "POST";
oReq.ContentLength = aBytes.Length;
Stream oStream = oReq.GetRequestStream();
oStream.Write(aBytes, 0, aBytes.Length);
oStream.Close();
AsyncState oState = new AsyncState(oReq);
return oReq.BeginGetResponse(fCallBack, oState);

分享到QQ

分享到微博

如果你对这篇内容有疑问，欢迎到本站社区发帖提问参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。

发布评论

需要登录才能够评论，你可以免费注册一个本站的账号。

魔 2024-08-28 05:47:42

我发现我正在使用的方案有一个重大改进。而不是使用 StreamReader 和 ReadToEnd 将 Stream 转换为字符串，然后才将其转换为 XmlDocument。我跳过了中间人，直接将 Stream 转换为 XmlDocument。

但这给我带来了另一个问题，我必须更改 XmlDocument 的父级以适合我的 Xslt（有很多，它们都期望我拥有的结构）。请参阅如何将新根元素添加到 C# XmlDocument ？进行修复。

这使我处理 Web 服务调用结果所需的时间减少了大约 2/3，并且使用的内存量也大大减少。在以前的版本中，响应 xml 在内存中存在两次不同的时间（如果流计数的话，可能是三次）！

此外，删除额外的 GetResponse 似乎也有帮助。

using (HttpWebResponse oResp = (HttpWebResponse)oReq.EndGetResponse(oResult))
{
oXml.Load(oResp.GetResponseStream());
XmlNode oApiResult = oXml.RemoveChild(oXml.DocumentElement);
oXml.LoadXml(sOtherXml);
oXml.DocumentElement.AppendChild(oApiResult);
}

I found one major improvement to the scheme I was using. Rather than using the StreamReader and ReadToEnd to get the Stream into a string, only then to convert it into an XmlDocument. I skipped the middle man and converted the Stream directly into a XmlDocument.

This left me with another problem though, I had to change the parent of the XmlDocument to fit my Xslt (there are a great many and they all expect the structure I had). See How can I add new root element to a C# XmlDocument? for that fix.

This has given me a roughly 2/3 decrease in the time taken to process the results of the webservice call, and a great decrease in the amount of memory used. In the previous version, the response xml was in memory two different times (maybe three if the stream counts)!

In addition, removing the extra GetResponse seemed to help.

using (HttpWebResponse oResp = (HttpWebResponse)oReq.EndGetResponse(oResult))
{
oXml.Load(oResp.GetResponseStream());
XmlNode oApiResult = oXml.RemoveChild(oXml.DocumentElement);
oXml.LoadXml(sOtherXml);
oXml.DocumentElement.AppendChild(oApiResult);
}

回复收藏 0 原文

~没有更多了~