我正在编写一个用于消费一些数据的Web服务.发送它的第三方正在使用多部分请求,当我在WireShark中查看请求时,它会被分块.
当我尝试通过fiddler作为标准(unchunked)请求运行完全相同的请求时,它工作正常,但是,我似乎无法读取分块请求.
我尝试了两种不同的方法.第一:
public HttpResponseMessage ImportEstimate(HttpRequestMessage request) { var data = request.Content.ReadAsMultipartAsync().Result; IEnumerable<HttpContent> parts = data.Contents; return request.CreateResponse(HttpStatusCode.OK,parts.Count()); }
这不会返回任何东西.它只是坐着,直到请求超时.
第二种方法是:
public HttpResponseMessage ImportEstimate(HttpRequestMessage request) { IEnumerable<HttpContent> parts = null; Task.Factory .StartNew(() => parts = Request.Content.ReadAsMultipartAsync().Result.Contents,CancellationToken.None,TaskCreationoptions.LongRunning,// guarantees separate thread TaskScheduler.Default) .Wait(); return request.CreateResponse(HttpStatusCode.OK,parts.Count()); }
哪个返回错误:
Unexpected end of MIME multipart stream. MIME multipart message is not complete.
我在这里错过了什么?
编辑:以下是WireShark的请求
POST /myservice/importestimate HTTP/1.1 Host: devapi.mydomain.com Content-Type: multipart/related; type="application/xop+xml"; start="<start.xml>"; start-info="text/xml"; boundary="--MIME_boundary" transfer-encoding: chunked SOAPAction: "importestimate" X-Forwarded-For: xxx.xx.xxx.xxx Connection: close 94 ----MIME_boundary Content-Type: application/xop+xml; type="text/xml; charset=UTF-8" Content-transfer-encoding: 8bit Content-Id: <start.xml> 170 <?xml version='1.0' encoding='UTF-8' ?><soap:Envelope xmlns:soap="http://schemas.xmlsoap.org/soap/envelope/" xmlns:xop="http://www.w3.org/2004/08/xop/include" ><soap:Body><XDOC><XNET_INFO transactionId="001P92V" ><ATTACHMENTS><ATTACHMENT><xop:Include href="cid:[email protected]" /></ATTACHMENT></ATTACHMENTS></XNET_INFO></XDOC></soap:Body></soap:Envelope> B3 ----MIME_boundary Content-Type: application/zip Content-transfer-encoding: binary Content-disposition: attachment; filename="XDOC.ZIP" Content-Id: <[email protected]> 5BC ... lots of data here removed for brevity... 15 ----MIME_boundary-- 0
解决方法
经过大量研究,我发现了问题!
显然我的请求并没有以CRLF结束,.Net必须要求发出请求结束的信号.
我最终阅读了整个请求,添加了一个CRLF,并从MemoryStream创建了自己的ReadAsMultipartAsync.这似乎有效.
private byte[] ProcessInput(HttpRequestMessage request) { List<HttpContent> parts = new List<HttpContent>(); byte[] result; result = request.Content.ReadAsByteArrayAsync().Result; // Stupid chunked requests remove the final CRLF,which makes .Net puke on the request. // So add our own CRLF. List<byte> crlfTemp = result.ToList(); crlfTemp.Add(0x0D); crlfTemp.Add(0x0A); result = crlfTemp.ToArray(); // Convert stream to MIME using (var stream = new MemoryStream(result)) { // note: StreamContent has no Content-Type set by default // set a suitable Content-Type for ReadAsMultipartAsync() var content = new StreamContent(stream); content.Headers.ContentType = System.Net.Http.Headers.MediaTypeHeaderValue.Parse( "multipart/related; boundary=--MIME_boundary"); content.Headers.ContentLength = result.Length; bool isMPC = content.IsMimeMultipartContent(); Task.Factory .StartNew(() => parts = content.ReadAsMultipartAsync().Result.Contents.ToList(),// guarantees separate thread TaskScheduler.Default) .Wait(); } }
版权声明:本文内容由互联网用户自发贡献,该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容, 请发送邮件至 [email protected] 举报,一经查实,本站将立刻删除。