通过处理程序 ASP.NET 将文本流式传输到客户端
为了解决 Twitter 流 API 没有跨域文件来从客户端访问它的问题（在本例中为 Silverlight），我在一个 Web 项目中创建了一个通用处理程序文件，该文件基本上从 twitter 下载流，并在读取流时将其写入客户。 这是处理程序代码：
context.Response.Buffer = false;
context.Response.ContentType = "text/plain";
WebRequest request = WebRequest.Create("http://stream.twitter.com/1/statuses/filter.json?locations=-180,-90,180,90");
request.Credentials = new NetworkCredential("username", "password");
StreamReader responseStream = new StreamReader(request.GetResponse().GetResponseStream(), Encoding.GetEncoding("utf-8"));
while (!responseStream.EndOfStream)
{
string line = "(~!-/" + responseStream.ReadLine() + "~!-/)";
context.Response.BinaryWrite((Encoding.UTF8.GetBytes(line)));}
这确实有效，但问题是一旦客户端断开连接，处理程序就会继续下载。那么如何判断客户端是否仍在忙于接收请求，如果不是，则结束 while 循环？
另外，我的第二个问题是，在客户端执行“ReadLine()”不会执行任何操作，大概是因为它将整个流计为一行，因此永远不会得到完整的响应。为了解决这个问题，我逐字节读取它，当它看到“(~！-/”周围的东西时，它知道这是一行。我知道，非常hacky。
谢谢！
To get around twitters streaming API not having a crossdomain file to access it from client side( in this case Silverlight) I have made a Generic Handler file in a web project which basically downloads the stream from twitter and as it reads it, writes it to the client.
Here is the handler code:
context.Response.Buffer = false;
context.Response.ContentType = "text/plain";
WebRequest request = WebRequest.Create("http://stream.twitter.com/1/statuses/filter.json?locations=-180,-90,180,90");
request.Credentials = new NetworkCredential("username", "password");
StreamReader responseStream = new StreamReader(request.GetResponse().GetResponseStream(), Encoding.GetEncoding("utf-8"));
while (!responseStream.EndOfStream)
{
string line = "(~!-/" + responseStream.ReadLine() + "~!-/)";
context.Response.BinaryWrite((Encoding.UTF8.GetBytes(line)));}
And this does work, but the problem is that once the client disconnects the handler just carry's on downloading. So how do I tell if the client is still busy receiving the request and if not, end the while loop?
Also, my second problem is that on the client side doing a "ReadLine()" does nothing, presumably because it is counting the entire stream as one line so never gets the full response. To work around that I read it byte by byte and when it sees "(~!-/" around something it know that is one line. VERY hacky, I know.
Thanks!
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。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
找到答案了！
:)
Found the answer!
:)