We Keep Coding

ASP.NET attempting to load streamed video source into HTML5

Over the last few weeks I have been working on an ASP.NET WebAPI that was designed to stream video from one of the company servers and play it on an HTML5 <video> element. Following a guide on C# Corner, we got the API published and now when the link for one of our videos is pasted into a browser, it starts to download (which, by the way, I'm not sure if it's supposed to do that when all we're trying to do is stream).
The files we need to stream are mp4 and are going to be used largely on iOS devices through Safari. And before anyone asks: srcVid is programmed and confirmed to be able to encode .mp4 files successfully, as we have managed to hard-code videos into this element with no issue. With that said, this is how the page handles its HTML5 elements:
<video autoplay muted id="trainVid" style="width: 75%; height: auto;" controls>
<source id="srcVid" runat="server"
type='video/mp4; codecs*="avc1.424085, mp4a.40.2"' />
</video>
On the API side, here are how the videos are processed, largely following the example set by the C# article, as well as some help from a Stephen Cleary article:
public class VidService
{
public async void WriteVidBytes(Stream outputStream,
HttpContent content, TransportContext tc)
{
try
{
var filePath = "\\\\server\\link\\to\\file.mp4";
int bufferSize = 1000;
byte[] buffer = new byte[bufferSize];
using (var fileStream = new FileStream(
filePath, FileMode.Open, FileAccess.Read, FileShare.Read))
{
int fileSize = (int)fileStream.Length;
while (fileSize > 0)
{
int cnt = fileSize > bufferSize ? bufferSize : fileSize,
readBufferSize = fileStream.Read(buffer, 0, cnt);
await outputStream.WriteAsync(buffer, 0, readBufferSize);
fileSize -= readBufferSize;
}
}
}
catch (HttpException ex) { return; }
finally { outputStream.Close(); }
}
public HttpResponseMessage GetVidContent()
{
// NOTE: please see the Edit on 6/10
var httpResponse = new HttpResponseMessage(HttpStatusCode.OK)
{
Content = new PushStreamContent(
(Action<Stream, HttpContent, TransportContext>)WriteVidBytes
)
};
return httpResponse;
}
}
public class VidController : ApiController
{
private static readonly VidService vs = new VidService();
[HttpGet]
public HttpResponseMessage GetVid(int id)
{
return vs.GetVidContent(id);
}
}
*Note that the actual program dynamically fetches video links through a Video.cs object
And finally on the C# side:
protected void LoadVideo(int vidId)
{
HttpWebRequest req = (HttpWebRequest)WebRequest.Create(
string.Format("http://mobileAPI.website.com/Vid/GetVid/" + vidId.ToString()));
req.Method = "GET";
HttpWebResponse resp = (HttpWebResponse)req.GetResponse();
string jsonString;
using (Stream stream = resp.GetResponseStream())
{
StreamReader r = new StreamReader(stream, System.Text.Encoding.UTF8);
jsonString = r.ReadToEnd();
}
srcVid.Src = jsonString;
}
When opening the page, LoadVideo() seems to execute with no errors -- but after this, the page goes blank and hangs forever. I'm thinking this may be because I'm putting the wrong value into srcVid.Src, but if I don't put in the jsonString, then what do I put in for the source?
As always, any help would be greatly appreciated! If I missed anything obvious, please let me know, as this is the first time I have worked with WebAPI.
UPDATE 1 (6/10)
I made a secondary method that took WriteVidBytes and turned it into a Task -- and other than turning it into a Task, the code inside is exactly the same. Another difference, also, is how GetVidContent fetches the data:
public HttpResponseMessage GetVidContent(int vId)
{
var httpResponse = new HttpResponseMessage(HttpStatusCode.OK)
{
Content = new PushStreamContent(async
(outputStream, httpContext, transportContext) =>
{
await WriteVidTask(outputStream, httpContext, transportContext);
}),
};
return httpResponse;
}
However, the page still hangs even though there is no problem getting the file through Postman or Fiddler.

Following an example from Robert Huang on CodeProject, I was able to get the video to stream through API without any issue.
The first thing I changed was the way that the <video> reads the source. Rather than a JSON string, the video loads the API link.
srcVid.Src = "http://api.website.com/Vid/GetVid?id=" + vidId.ToString();
Following the CodeProject link, I created an HttpResponseMessage, very similar to the one provided -- only this one supports asynchronous loading:
public HttpResponseMessage GetVidContent(RangeHeaderValue rh, FileInfo fi)
{
HttpResponseMessage response = new HttpResponseMessage();
response.Headers.AcceptRanges.Add("bytes");
long totalLength = fi.Length;
if (rh == null || !rh.Ranges.Any())
{ // treat request normally if there is no range header
response.StatusCode = HttpStatusCode.OK;
response.Content = new PushStreamContent(async (outputStream,
httpContent, transpContext) =>
{
using (outputStream) // copy file to output stream straightforward
using (Stream inputStream = fi.OpenRead())
{
try
{
await inputStream.CopyToAsync(outputStream, ReadStreamBufferSize);
}
catch (Exception ex)
{
Debug.WriteLine(ex);
}
}
}, GetMimeNameFromExt(fi.Extension));
return response;
}
long start = 0, end = 0;
if (rh.Unit != "bytes" || rh.Ranges.Count > 1 || TryReadRangeItem(rh.Ranges.First(),
totalLength, out start, out end))
{
response.StatusCode = HttpStatusCode.RequestedRangeNotSatisfiable;
response.Content = new StreamContent(Stream.Null);
response.Content.Headers.ContentRange = new ContentRangeHeaderValue(totalLength);
response.Content.Headers.ContentType = GetMimeNameFromExt(fi.Extension);
return response;
}
var contentRange = new ContentRangeHeaderValue(start, end, totalLength);
response.StatusCode = HttpStatusCode.PartialContent;
response.Content = new PushStreamContent(async (outputStream, httpContent, transpContext) =>
{
using (outputStream)
using (Stream inputStream = fi.OpenRead())
await CreatePartialContent(inputStream, outputStream, start, end);
}, GetMimeNameFromExt(fi.Extension));
response.Content.Headers.ContentLength = end - start + 1;
response.Content.Headers.ContentRange = contentRange;
return response;
}
Any methods from the CodeProject article used in await areas were marked with the async expression.

getting the request body inside HttpContext from a Middleware in asp.net core 2.0

I am having a simple middleware which fetches the body of the request and store it in a string. It is reading fine the stream, but the issue is it wont call my controller which called just after I read the stream and throw the error
A non-empty request body is required
. Below is my code.
public async Task Invoke(HttpContext httpContext)
{
var timer = Stopwatch.StartNew();
ReadBodyFromHttpContext(httpContext);
await _next(httpContext);
timer.Stop();
}
private string ReadBodyFromHttpContext(HttpContext httpContext)
{
return await new StreamReader(httpContext.Request.Body).ReadToEndAsync();
}

You need to convert HttpContext.Request.Body from a forward only memory stream to a seekable stream, shown below.
// Enable seeking
context.Request.EnableBuffering();
// Read the stream as text
var bodyAsText = await new System.IO.StreamReader(context.Request.Body).ReadToEndAsync();
// Set the position of the stream to 0 to enable rereading
context.Request.Body.Position = 0;

when it comes to capturing the body of an HTTP request and/or response, this is no trivial effort. In ASP .NET Core, the body is a stream – once you consume it (for logging, in this case), it’s gone, rendering the rest of the pipeline useless.
Ref:http://www.palador.com/2017/05/24/logging-the-body-of-http-request-and-response-in-asp-net-core/
public async Task Invoke(HttpContext httpContext)
{
var timer = Stopwatch.StartNew();
string bodyAsText = await new StreamReader(httpContext.Request.Body).ReadToEndAsync();
var injectedRequestStream = new MemoryStream();
var bytesToWrite = Encoding.UTF8.GetBytes(bodyAsText);
injectedRequestStream.Write(bytesToWrite, 0, bytesToWrite.Length);
injectedRequestStream.Seek(0, SeekOrigin.Begin);
httpContext.Request.Body = injectedRequestStream;
await _next(httpContext);
timer.Stop();
}

Few things are crucial here:
enable buffering
last flag leaveOpen in StreamReader
reset request body stream position (SeekOrigin.Begin)
public void UseMyMiddleware(IApplicationBuilder app)
{
app.Use(async (context, next) =>
{
context.Request.EnableBuffering();
using (var reader = new StreamReader(context.Request.Body, Encoding.UTF8, false, 1024, true))
{
var body = await reader.ReadToEndAsync();
context.Request.Body.Seek(0, SeekOrigin.Begin);
}
await next.Invoke();
});
}

using (var mem = new MemoryStream())
using (var reader = new StreamReader(mem))
{
Request.Body.CopyTo(mem);
var body = reader.ReadToEnd();
//and this you can reset the position of the stream.
mem.Seek(0, SeekOrigin.Begin);
body = reader.ReadToEnd();
}
Here you are can read how it works. https://gunnarpeipman.com/aspnet-core-request-body/

You can try this
public async Task Invoke(HttpContext context)
{
var request = context.Request;
request.EnableBuffering();
var buffer = new byte[Convert.ToInt32(request.ContentLength)];
await request.Body.ReadAsync(buffer, 0, buffer.Length);
var requestContent = Encoding.UTF8.GetString(buffer);
request.Body.Position = 0; //rewinding the stream to 0
}

Large File download from SQL via WebApi after custom MultipartFormDataStreamProvider upload

This is a follow up to a question I had asked previously that was closed for being too broad.Previous Question
In that question I explained that I needed upload a large file (1-3GB) to the database by storing chunks as individual rows. I did this by overriding the MultipartFormDataStreamProvider.GetStream method. That method returned a custom stream that wrote the buffered chunks to the database.
The problem is that the overriden GetStream method is writing the entire request to the database (including the headers). It is successfully writing that data while keeping the Memory levels flat but when I download the file, in addition to the file contents, it's returning all the header information in the downloaded file contents so the file can't be opened.
Is there a way to, in the overriden GetStream method, write just the contents of the file to the database without writing the headers?
API
[HttpPost]
[Route("file")]
[ValidateMimeMultipartContentFilter]
public Task<HttpResponseMessage> PostFormData()
{
var provider = new CustomMultipartFormDataStreamProvider();
// Read the form data and return an async task.
var task = Request.Content.ReadAsMultipartAsync(provider).ContinueWith<HttpResponseMessage>(t =>
{
if (t.IsFaulted || t.IsCanceled)
{
Request.CreateErrorResponse(HttpStatusCode.InternalServerError, t.Exception);
}
return Request.CreateResponse(HttpStatusCode.OK);
});
return task;
}
[HttpGet]
[Route("file/{id}")]
public async Task<HttpResponseMessage> GetFile(string id)
{
var result = new HttpResponseMessage()
{
Content = new PushStreamContent(async (outputStream, httpContent, transportContext) =>
{
await WriteDataChunksFromDBToStream(outputStream, httpContent, transportContext, id);
}),
StatusCode = HttpStatusCode.OK
};
result.Content.Headers.ContentType = new MediaTypeHeaderValue("application/zipx");
result.Content.Headers.ContentDisposition = new ContentDispositionHeaderValue("attachment") { FileName = "test response.zipx" };
return result;
}
return new HttpResponseMessage(HttpStatusCode.BadRequest);
}
private async Task WriteDataChunksFromDBToStream(Stream responseStream, HttpContent httpContent, TransportContext transportContext, string fileIdentifier)
{
// PushStreamContent requires the responseStream to be closed
// for signaling it that you have finished writing the response.
using (responseStream)
{
using (var myConn = new SqlConnection(System.Configuration.ConfigurationManager.ConnectionStrings["TestDB"].ConnectionString))
{
await myConn.OpenAsync();
using (var myCmd = new SqlCommand("ReadAttachmentChunks", myConn))
{
myCmd.CommandType = System.Data.CommandType.StoredProcedure;
var fileName = new SqlParameter("#Identifier", fileIdentifier);
myCmd.Parameters.Add(fileName);
// Read data back from db in async call to avoid OutOfMemoryException when sending file back to user
using (var reader = await myCmd.ExecuteReaderAsync(CommandBehavior.SequentialAccess))
{
while (await reader.ReadAsync())
{
if (!(await reader.IsDBNullAsync(3)))
{
using (var data = reader.GetStream(3))
{
// Asynchronously copy the stream from the server to the response stream
await data.CopyToAsync(responseStream);
}
}
}
}
}
}
}// close response stream
}
Custom MultipartFormDataStreamProvider GetStream method implementation
public override Stream GetStream(HttpContent parent, HttpContentHeaders headers)
{
// For form data, Content-Disposition header is a requirement
ContentDispositionHeaderValue contentDisposition = headers.ContentDisposition;
if (contentDisposition != null)
{
// If we have a file name then write contents out to AWS stream. Otherwise just write to MemoryStream
if (!String.IsNullOrEmpty(contentDisposition.FileName))
{
var identifier = Guid.NewGuid().ToString();
var fileName = contentDisposition.FileName;// GetLocalFileName(headers);
if (fileName.Contains("\\"))
{
fileName = fileName.Substring(fileName.LastIndexOf("\\") + 1).Replace("\"", "");
}
// We won't post process files as form data
_isFormData.Add(false);
var stream = new CustomSqlStream();
stream.Filename = fileName;
stream.Identifier = identifier;
stream.ContentType = headers.ContentType.MediaType;
stream.Description = (_formData.AllKeys.Count() > 0 && _formData["description"] != null) ? _formData["description"] : "";
return stream;
//return new CustomSqlStream(contentDisposition.Name);
}
// We will post process this as form data
_isFormData.Add(true);
// If no filename parameter was found in the Content-Disposition header then return a memory stream.
return new MemoryStream();
}
throw new InvalidOperationException("Did not find required 'Content-Disposition' header field in MIME multipart body part..");
#endregion
}
Implemented Write method of Stream called by CustomSqlStream
public override void Write(byte[] buffer, int offset, int count)
{
//write buffer to database
using (var myConn = new SqlConnection(System.Configuration.ConfigurationManager.ConnectionStrings["TestDB"].ConnectionString)) {
using (var myCmd = new SqlCommand("WriteAttachmentChunk", myConn)) {
myCmd.CommandType = System.Data.CommandType.StoredProcedure;
var pContent = new SqlParameter("#Content", buffer);
myCmd.Parameters.Add(pContent);
myConn.Open();
myCmd.ExecuteNonQuery();
if (myConn.State == System.Data.ConnectionState.Open)
{
myConn.Close();
}
}
}
((ManualResetEvent)_dataAddedEvent).Set();
}
The "ReadAttachmentChunks" stored procedure gets the rows respective to the file from the db ordered by the time they are inserted into the database. So, the way the code works is it pulls those chunks back and then async writes it back to the PushStreamContent to go back to the user.
So my question is:
Is there a way to write ONLY the content of the file being uploaded as opposed to the headers in addition to the content?
Any help would be greatly appreciated. Thank you.

I finally figured it out. I over-complicated the write process which brought about most of the struggle. Here is my solution to my initial issue:
To keep .net from buffering the file in memory (so that you can handle large file uploads), you first need to override the WebHostBufferPolicySelector so that it doesnt buffer the input stream for your controller and then replace the BufferPolicy Selector.
public class NoBufferPolicySelector : WebHostBufferPolicySelector
{
public override bool UseBufferedInputStream(object hostContext)
{
var context = hostContext as HttpContextBase;
if (context != null)
{
if (context.Request.RequestContext.RouteData.Values["controller"] != null)
{
if (string.Equals(context.Request.RequestContext.RouteData.Values["controller"].ToString(), "upload", StringComparison.InvariantCultureIgnoreCase))
return false;
}
}
return true;
}
public override bool UseBufferedOutputStream(HttpResponseMessage response)
{
return base.UseBufferedOutputStream(response);
}
}
then for replacing the BufferPolicy Selector
GlobalConfiguration.Configuration.Services.Replace(typeof(IHostBufferPolicySelector), new NoBufferPolicySelector());
Then to avoid the default behavior of having the file stream written to disk, you need to provide a stream provider that will write to the database instead. To do this you inherit MultipartStreamProvider and override the GetStream method to return the stream that will write to your database.
public override Stream GetStream(HttpContent parent, HttpContentHeaders headers)
{
// For form data, Content-Disposition header is a requirement
ContentDispositionHeaderValue contentDisposition = headers.ContentDisposition;
if (contentDisposition != null && !String.IsNullOrEmpty(contentDisposition.FileName))
{
// We won't post process files as form data
_isFormData.Add(false);
//create unique identifier for this file upload
var identifier = Guid.NewGuid();
var fileName = contentDisposition.FileName;
var boundaryObj = parent.Headers.ContentType.Parameters.SingleOrDefault(a => a.Name == "boundary");
var boundary = (boundaryObj != null) ? boundaryObj.Value : "";
if (fileName.Contains("\\"))
{
fileName = fileName.Substring(fileName.LastIndexOf("\\") + 1).Replace("\"", "");
}
//write parent container for the file chunks that are being stored
WriteLargeFileContainer(fileName, identifier, headers.ContentType.MediaType, boundary);
//create an instance of the custom stream that will write the chunks to the database
var stream = new CustomSqlStream();
stream.Filename = fileName;
stream.FullFilename = contentDisposition.FileName.Replace("\"", "");
stream.Identifier = identifier.ToString();
stream.ContentType = headers.ContentType.MediaType;
stream.Boundary = (!string.IsNullOrEmpty(boundary)) ? boundary : "";
return stream;
}
else
{
// We will post process this as form data
_isFormData.Add(true);
// If no filename parameter was found in the Content-Disposition header then return a memory stream.
return new MemoryStream();
}
}
The custom stream you create needs to inherit Stream and override the Write method. This is where I overthought the problem and thought I needed to parse out the boundary headers that were passed via the buffer parameter. But this is actually done for you by leveraging the offset and count parameters.
public override void Write(byte[] buffer, int offset, int count)
{
//no boundary is inluded in buffer
byte[] fileData = new byte[count];
Buffer.BlockCopy(buffer, offset, fileData, 0, count);
WriteData(fileData);
}
From there, it's just plugging in the api methods for upload and download.
For upload:
public Task<HttpResponseMessage> PostFormData()
{
var provider = new CustomMultipartLargeFileStreamProvider();
// Read the form data and return an async task.
var task = Request.Content.ReadAsMultipartAsync(provider).ContinueWith<HttpResponseMessage>(t =>
{
if (t.IsFaulted || t.IsCanceled)
{
Request.CreateErrorResponse(HttpStatusCode.InternalServerError, t.Exception);
}
return Request.CreateResponse(HttpStatusCode.OK);
});
return task;
}
For download, and in order to keep the memory footprint low, I leveraged the PushStreamContent to push the chunks back to the user:
[HttpGet]
[Route("file/{id}")]
public async Task<HttpResponseMessage> GetFile(string id)
{
string mimeType = string.Empty;
string filename = string.Empty;
if (!string.IsNullOrEmpty(id))
{
//get the headers for the file being sent back to the user
using (var myConn = new SqlConnection(System.Configuration.ConfigurationManager.ConnectionStrings["PortalBetaConnectionString"].ConnectionString))
{
using (var myCmd = new SqlCommand("ReadLargeFileInfo", myConn))
{
myCmd.CommandType = System.Data.CommandType.StoredProcedure;
var pIdentifier = new SqlParameter("#Identifier", id);
myCmd.Parameters.Add(pIdentifier);
myConn.Open();
var dataReader = myCmd.ExecuteReader();
if (dataReader.HasRows)
{
while (dataReader.Read())
{
mimeType = dataReader.GetString(0);
filename = dataReader.GetString(1);
}
}
}
}
var result = new HttpResponseMessage()
{
Content = new PushStreamContent(async (outputStream, httpContent, transportContext) =>
{
//pull the data back from the db and stream the data back to the user
await WriteDataChunksFromDBToStream(outputStream, httpContent, transportContext, id);
}),
StatusCode = HttpStatusCode.OK
};
result.Content.Headers.ContentType = new MediaTypeHeaderValue(mimeType);// "application/octet-stream");
result.Content.Headers.ContentDisposition = new ContentDispositionHeaderValue("attachment") { FileName = filename };
return result;
}
return new HttpResponseMessage(HttpStatusCode.BadRequest);
}
private async Task WriteDataChunksFromDBToStream(Stream responseStream, HttpContent httpContent, TransportContext transportContext, string fileIdentifier)
{
// PushStreamContent requires the responseStream to be closed
// for signaling it that you have finished writing the response.
using (responseStream)
{
using (var myConn = new SqlConnection(System.Configuration.ConfigurationManager.ConnectionStrings["PortalBetaConnectionString"].ConnectionString))
{
await myConn.OpenAsync();
//stored proc to pull the data back from the db
using (var myCmd = new SqlCommand("ReadAttachmentChunks", myConn))
{
myCmd.CommandType = System.Data.CommandType.StoredProcedure;
var fileName = new SqlParameter("#Identifier", fileIdentifier);
myCmd.Parameters.Add(fileName);
// The reader needs to be executed with the SequentialAccess behavior to enable network streaming
// Otherwise ReadAsync will buffer the entire BLOB into memory which can cause scalability issues or even OutOfMemoryExceptions
using (var reader = await myCmd.ExecuteReaderAsync(CommandBehavior.SequentialAccess))
{
while (await reader.ReadAsync())
{
//confirm the column that has the binary data of the file returned is not null
if (!(await reader.IsDBNullAsync(0)))
{
//read the binary data of the file into a stream
using (var data = reader.GetStream(0))
{
// Asynchronously copy the stream from the server to the response stream
await data.CopyToAsync(responseStream);
await data.FlushAsync();
}
}
}
}
}
}
}// close response stream
}

Ugh. This is nasty. With the upload, you have to make sure to
separate the headers from the content portion - you must follow the requirements RFC documents for HTTP.
Allow for chunked transfers
Of course, the content portion (unless you are transmitting text) will be binary encoded into strings.
Allow for transfers that are compressed, i.e. GZIP or DEFLATE.
Maybe - just maybe - take the encoding into account (ASCII, Unicode, UTF8, etc).
You can't really ensure that you're persisting the right info to the database without looking at all of these. For the latter items, all of your metadata as to what to do will be somewhere in the header, so it's not just a throwaway.

WriteToStreamAsync cancel does not work

I am running a Task, which copies from one stream to another. This works without problems, including progress reporting. But i cant cancel the task. If i fire the CancellationToken, the copy progress runs till its completion, then the task is cancelled, but this is of course to late. Here is my actual code
private async Task Download(Uri uriToWork, CancellationToken cts)
{
HttpClient httpClient = new HttpClient();
HttpRequestMessage requestAction = new HttpRequestMessage();
requestAction.Method = new HttpMethod("GET");
requestAction.RequestUri = uriToWork;
HttpResponseMessage httpResponseContent = await httpClient.SendRequestAsync(requestAction, HttpCompletionOption.ResponseHeadersRead);
using (Stream streamToRead = (await httpResponseContent.Content.ReadAsInputStreamAsync()).AsStreamForRead())
{
string fileToWrite = Path.GetTempFileName();
using (Stream streamToWrite = File.Open(fileToWrite, FileMode.Create))
{
await httpResponseContent.Content.WriteToStreamAsync(streamToWrite.AsOutputStream()).AsTask(cts, progressDownload);
await streamToWrite.FlushAsync();
//streamToWrite.Dispose();
}
await streamToRead.FlushAsync();
//streamToRead.Dispose();
}
httpClient.Dispose();
}
Can someone help me please, or can explain, why it does not work?

Is it this operation that continues until it completes ?
await httpResponseContent.Content.WriteToStreamAsync(streamToWrite.AsOutputStream()).AsTask(cts, progressDownload);
Or is it this one ?
await streamToWrite.FlushAsync();
I think the latter needs probably to have the CancellationToken as well:
await streamToWrite.FlushAsync(cts);
Unfortunately I cannot answer why this cancel does not occur. However, a solution that consists in writing the Stream in chunks may help.
Here is something very dirty that works:
private async Task Download(Uri uriToWork, CancellationToken cts) {
using(HttpClient httpClient = new HttpClient()) {
HttpRequestMessage requestAction = new HttpRequestMessage();
requestAction.Method = new HttpMethod("GET");
requestAction.RequestUri = uriToWork;
HttpResponseMessage httpResponseContent = await httpClient.SendRequestAsync(requestAction, HttpCompletionOption.ResponseHeadersRead);
string fileToWrite = Path.GetTempFileName();
using(Stream streamToWrite = File.Open(fileToWrite, FileMode.Create)) {
// Disposes streamToWrite to force any write operation to fail
cts.Register(() => streamToWrite.Dispose());
try {
await httpResponseContent.Content.WriteToStreamAsync(streamToWrite.AsOutputStream()).AsTask(cts, p);
}
catch(TaskCanceledException) {
return; // "gracefully" exit when the token is cancelled
}
await streamToWrite.FlushAsync();
}
}
}
I enclosed the httpClient in a using so a return disposes it properly.
I removed the streamToRead which was not used at all
Now here is the horror: I added a delegate that executes when the token is cancelled: it disposes streamToWrite while it is written to (ughhhh), which triggers an TaskCancelledException when WriteToStreamAsync cannot longer write in this disposed stream.
Please dont throw a puke bag at me yet, I am not experienced enough in this "Universal" Framework which looks very different as the usual one.
Here is a chunked stream solution that looks more acceptable. I shortened a bit the original code and added the IProgress as a parameter.
async Task Download(Uri uriToWork, CancellationToken cts, IProgress<int> progress) {
using(HttpClient httpClient = new HttpClient()) {
var chunkSize = 1024;
var buffer = new byte[chunkSize];
int count = 0;
string fileToWrite = Path.GetTempFileName();
using(var inputStream = await httpClient.GetInputStreamAsync(uriToWork)) {
using(var streamToRead = inputStream.AsStreamForRead()) {
using(Stream streamToWrite = File.OpenWrite(fileToWrite)) {
int size;
while((size = await streamToRead.ReadAsync(buffer, 0, chunkSize, cts).ConfigureAwait(false)) > 0) {
count += size;
await Dispatcher.RunAsync(CoreDispatcherPriority.Normal, () => progress.Report(count));
// progress.Report(count);
await streamToWrite.WriteAsync(buffer, 0, size, cts).ConfigureAwait(false);
}
}
}
}
}
}

The blocking operation is most probably not WriteToStreamAsync() but FlushAsync(), so #Larry's assumption should be right, the FlushAsync method needs the cancellation token as well.

Modify OWIN/Katana PhysicalFileSystem page on request

I have a self-hosted app that's using OWIN to provide a basic web server. The key part of the configuration is the following line:
appBuilder.UseFileServer(new FileServerOptions {
FileSystem = new PhysicalFileSystem(filePath)
});
This provides the static files listed in the filePath for browsing, and this much is working as expected.
However I've run into a case where I want to slightly modify one of the files on a request-by-request basis. In particular, I want to load the "normal" version of the file from the filesystem, alter it slightly based on the incoming web request's headers, and then return the altered version to the client instead of the original. All other files should remain unmodified.
How do I go about doing this?

Well, I don't know if this is a good way to do this, but it seems to work:
internal class FileReplacementMiddleware : OwinMiddleware
{
public FileReplacementMiddleware(OwinMiddleware next) : base(next) {}
public override async Task Invoke(IOwinContext context)
{
MemoryStream memStream = null;
Stream httpStream = null;
if (ShouldAmendResponse(context))
{
memStream = new MemoryStream();
httpStream = context.Response.Body;
context.Response.Body = memStream;
}
await Next.Invoke(context);
if (memStream != null)
{
var content = await ReadStreamAsync(memStream);
if (context.Response.StatusCode == 200)
{
content = AmendContent(context, content);
}
var contentBytes = Encoding.UTF8.GetBytes(content);
context.Response.Body = httpStream;
context.Response.ETag = null;
context.Response.ContentLength = contentBytes.Length;
await context.Response.WriteAsync(contentBytes, context.Request.CallCancelled);
}
}
private static async Task<string> ReadStreamAsync(MemoryStream stream)
{
stream.Seek(0, SeekOrigin.Begin);
using (var reader = new StreamReader(stream, Encoding.UTF8))
{
return await reader.ReadToEndAsync();
}
}
private bool ShouldAmendResponse(IOwinContext context)
{
// logic
}
private string AmendContent(IOwinContext context, string content)
{
// logic
}
}
Add this to the pipeline somewhere before the static files middleware.