Corrupt PDF files When File is written to the response.

22 Aug 2006 4:12 PM

eventuranza

For my application, there users can upload and download files to the a
webserver. Straightforward enough. However, when they upload a PDF
file then try to download it, the file seems to be corrupted. The had
not been a problem before but it seems to have been introduced when I
refactored the writing to the response to a buffered version to prevent
potential OutOfMemoryExceptions (you wouldn't believe the size of some
of these files...).

Here's the code (pretty much jacked from a Microsoft knowledge base
article):

private void WriteToResponse(string filepath, string contentType,
NameValueCollection extraHeaders)
        {
            Stream iStream = null;

            // Buffer to read 10K bytes in chunk:
            byte[] buffer = new Byte[10000];

            // Length of the file:
            int length;

            // Total bytes to read:
            long dataToRead;

            // Identify the file name.
            string filename = Path.GetFileName(filepath);

            try
            {
                // Open the file.
                iStream = new FileStream(filepath, FileMode.Open,
                    FileAccess.Read, FileShare.Read);

                // Total bytes to read:
                dataToRead = iStream.Length;

                Response.ContentType = contentType;
                if(extraHeaders != null)
                {
                    for (int i = 0; i < extraHeaders.Count; i++)
                    {
                        Trace.Write("adding Key value: " +
extraHeaders.GetKey(i)
                                    + ". Adding value: " +
extraHeaders[i]);
                        Response.AddHeader(extraHeaders.GetKey(i),
extraHeaders[i]);
                    }
                }

                // Read the bytes.
                while (dataToRead > 0)
                {
                    // Verify that the client is connected.
                    if (Response.IsClientConnected)
                    {
                        // Read the data in buffer.
                        length = iStream.Read(buffer, 0, 10000);

                        // Write the data to the current output stream.
                        Response.OutputStream.Write(buffer, 0, length);

                        // Flush the data to the HTML output.
                        Response.Flush();

                        buffer = new Byte[10000];
                        dataToRead = dataToRead - length;
                    }
                    else
                    {
                        //prevent infinite loop if user disconnects
                        dataToRead = -1;
                    }
                }
            }
            catch(IOException ioEx)
            {
                errorRow.Visible = true;
                lblErrorMsg.Text = ioEx.Message;
                log.Error(ioEx.ToString());
            }
            catch (Exception ex)
            {
                errorRow.Visible = true;
                // Trap the error, if any.
                log.Error(ex.ToString());
            }
            finally
            {
                if (iStream != null)
                {
                    //Close the file.
                    iStream.Close();
                }
            }
        }

Thanks in advance for any help.

22 Aug 2006 5:03 PM

Patrice

Trace not enabled ? All HTML stripped ? (using Response.Clear allows to make
sure you won't have a character coming from your APSX markup).

IMO The easiest path is to save the file (possibly using a content
disposition header) so that you can check the length. If not the length you
can do a file comparison to see where the file is corrupted.

--
Patrice

<eventura***@gmail.com> a écrit dans le message de news:
1156263142.620290.68***@74g2000cwt.googlegroups.com...

Show quote

> For my application, there users can upload and download files to the a
> webserver. Straightforward enough. However, when they upload a PDF
> file then try to download it, the file seems to be corrupted. The had
> not been a problem before but it seems to have been introduced when I
> refactored the writing to the response to a buffered version to prevent
> potential OutOfMemoryExceptions (you wouldn't believe the size of some
> of these files...).
>
> Here's the code (pretty much jacked from a Microsoft knowledge base
> article):
>
> private void WriteToResponse(string filepath, string contentType,
> NameValueCollection extraHeaders)
>        {
>            Stream iStream = null;
>
>            // Buffer to read 10K bytes in chunk:
>            byte[] buffer = new Byte[10000];
>
>            // Length of the file:
>            int length;
>
>            // Total bytes to read:
>            long dataToRead;
>
>            // Identify the file name.
>            string filename = Path.GetFileName(filepath);
>
>            try
>            {
>                // Open the file.
>                iStream = new FileStream(filepath, FileMode.Open,
>                    FileAccess.Read, FileShare.Read);
>
>
>                // Total bytes to read:
>                dataToRead = iStream.Length;
>
>                Response.ContentType = contentType;
>                if(extraHeaders != null)
>                {
>                    for (int i = 0; i < extraHeaders.Count; i++)
>                    {
>                        Trace.Write("adding Key value: " +
> extraHeaders.GetKey(i)
>                                    + ". Adding value: " +
> extraHeaders[i]);
>                        Response.AddHeader(extraHeaders.GetKey(i),
> extraHeaders[i]);
>                    }
>                }
>
>                // Read the bytes.
>                while (dataToRead > 0)
>                {
>                    // Verify that the client is connected.
>                    if (Response.IsClientConnected)
>                    {
>                        // Read the data in buffer.
>                        length = iStream.Read(buffer, 0, 10000);
>
>                        // Write the data to the current output stream.
>                        Response.OutputStream.Write(buffer, 0, length);
>
>                        // Flush the data to the HTML output.
>                        Response.Flush();
>
>                        buffer = new Byte[10000];
>                        dataToRead = dataToRead - length;
>                    }
>                    else
>                    {
>                        //prevent infinite loop if user disconnects
>                        dataToRead = -1;
>                    }
>                }
>            }
>            catch(IOException ioEx)
>            {
>                errorRow.Visible = true;
>                lblErrorMsg.Text = ioEx.Message;
>                log.Error(ioEx.ToString());
>            }
>            catch (Exception ex)
>            {
>                errorRow.Visible = true;
>                // Trap the error, if any.
>                log.Error(ex.ToString());
>            }
>            finally
>            {
>                if (iStream != null)
>                {
>                    //Close the file.
>                    iStream.Close();
>                }
>            }
>        }
>
> Thanks in advance for any help.
>

22 Aug 2006 5:15 PM

Tim_Mac

hi,
did you try using Response.BufferOutput = true;
i thought that was supposed to buffer the output to the client in smaller
chunks.
..Net 2.0 has a method called Response.TransmitFile() which according to the
docs says that it sends a file straight to the client without loading the
contents into memory. it must do a direct stream transfer from the file to
the http response.

the thought of doing tracing in the middle of a binary write makes me a bit
suspicious, surely this would corrupt the download? not sure though if the
trace strings actually get written out with the response.

thanks
tim

<eventura***@gmail.com> wrote in message
Show quote

news:1156263142.620290.68290@74g2000cwt.googlegroups.com...
> For my application, there users can upload and download files to the a
> webserver. Straightforward enough. However, when they upload a PDF
> file then try to download it, the file seems to be corrupted. The had
> not been a problem before but it seems to have been introduced when I
> refactored the writing to the response to a buffered version to prevent
> potential OutOfMemoryExceptions (you wouldn't believe the size of some
> of these files...).
>
> Here's the code (pretty much jacked from a Microsoft knowledge base
> article):
>
> private void WriteToResponse(string filepath, string contentType,
> NameValueCollection extraHeaders)
>        {
>            Stream iStream = null;
>
>            // Buffer to read 10K bytes in chunk:
>            byte[] buffer = new Byte[10000];
>
>            // Length of the file:
>            int length;
>
>            // Total bytes to read:
>            long dataToRead;
>
>            // Identify the file name.
>            string filename = Path.GetFileName(filepath);
>
>            try
>            {
>                // Open the file.
>                iStream = new FileStream(filepath, FileMode.Open,
>                    FileAccess.Read, FileShare.Read);
>
>
>                // Total bytes to read:
>                dataToRead = iStream.Length;
>
>                Response.ContentType = contentType;
>                if(extraHeaders != null)
>                {
>                    for (int i = 0; i < extraHeaders.Count; i++)
>                    {
>                        Trace.Write("adding Key value: " +
> extraHeaders.GetKey(i)
>                                    + ". Adding value: " +
> extraHeaders[i]);
>                        Response.AddHeader(extraHeaders.GetKey(i),
> extraHeaders[i]);
>                    }
>                }
>
>                // Read the bytes.
>                while (dataToRead > 0)
>                {
>                    // Verify that the client is connected.
>                    if (Response.IsClientConnected)
>                    {
>                        // Read the data in buffer.
>                        length = iStream.Read(buffer, 0, 10000);
>
>                        // Write the data to the current output stream.
>                        Response.OutputStream.Write(buffer, 0, length);
>
>                        // Flush the data to the HTML output.
>                        Response.Flush();
>
>                        buffer = new Byte[10000];
>                        dataToRead = dataToRead - length;
>                    }
>                    else
>                    {
>                        //prevent infinite loop if user disconnects
>                        dataToRead = -1;
>                    }
>                }
>            }
>            catch(IOException ioEx)
>            {
>                errorRow.Visible = true;
>                lblErrorMsg.Text = ioEx.Message;
>                log.Error(ioEx.ToString());
>            }
>            catch (Exception ex)
>            {
>                errorRow.Visible = true;
>                // Trap the error, if any.
>                log.Error(ex.ToString());
>            }
>            finally
>            {
>                if (iStream != null)
>                {
>                    //Close the file.
>                    iStream.Close();
>                }
>            }
>        }
>
> Thanks in advance for any help.
>

23 Aug 2006 2:34 AM

eventuranza

Thanks for all your input. What it turned to be was my neglecting to
put in "Response.End()" at the end of this method. The files that were
being downloaded actually had the emitted HTML response appended to the
end of the file. It turns out that out of out of a whole cross section
of files, ranging form movie files to Office and images, the PDF was
most sensitive to this extra information after the EOF and was gettin
gcorrupted.

Tim_Mac wrote:
Show quote

> hi,
> did you try using Response.BufferOutput = true;
> i thought that was supposed to buffer the output to the client in smaller
> chunks.
> .Net 2.0 has a method called Response.TransmitFile() which according to the
> docs says that it sends a file straight to the client without loading the
> contents into memory. it must do a direct stream transfer from the file to
> the http response.
>
> the thought of doing tracing in the middle of a binary write makes me a bit
> suspicious, surely this would corrupt the download? not sure though if the
> trace strings actually get written out with the response.
>
> thanks
> tim
>
> <eventura***@gmail.com> wrote in message
> news:1156263142.620290.68290@74g2000cwt.googlegroups.com...
> > For my application, there users can upload and download files to the a
> > webserver. Straightforward enough. However, when they upload a PDF
> > file then try to download it, the file seems to be corrupted. The had
> > not been a problem before but it seems to have been introduced when I
> > refactored the writing to the response to a buffered version to prevent
> > potential OutOfMemoryExceptions (you wouldn't believe the size of some
> > of these files...).
> >
> > Here's the code (pretty much jacked from a Microsoft knowledge base
> > article):
> >
> > private void WriteToResponse(string filepath, string contentType,
> > NameValueCollection extraHeaders)
> >        {
> >            Stream iStream = null;
> >
> >            // Buffer to read 10K bytes in chunk:
> >            byte[] buffer = new Byte[10000];
> >
> >            // Length of the file:
> >            int length;
> >
> >            // Total bytes to read:
> >            long dataToRead;
> >
> >            // Identify the file name.
> >            string filename = Path.GetFileName(filepath);
> >
> >            try
> >            {
> >                // Open the file.
> >                iStream = new FileStream(filepath, FileMode.Open,
> >                    FileAccess.Read, FileShare.Read);
> >
> >
> >                // Total bytes to read:
> >                dataToRead = iStream.Length;
> >
> >                Response.ContentType = contentType;
> >                if(extraHeaders != null)
> >                {
> >                    for (int i = 0; i < extraHeaders.Count; i++)
> >                    {
> >                        Trace.Write("adding Key value: " +
> > extraHeaders.GetKey(i)
> >                                    + ". Adding value: " +
> > extraHeaders[i]);
> >                        Response.AddHeader(extraHeaders.GetKey(i),
> > extraHeaders[i]);
> >                    }
> >                }
> >
> >                // Read the bytes.
> >                while (dataToRead > 0)
> >                {
> >                    // Verify that the client is connected.
> >                    if (Response.IsClientConnected)
> >                    {
> >                        // Read the data in buffer.
> >                        length = iStream.Read(buffer, 0, 10000);
> >
> >                        // Write the data to the current output stream.
> >                        Response.OutputStream.Write(buffer, 0, length);
> >
> >                        // Flush the data to the HTML output.
> >                        Response.Flush();
> >
> >                        buffer = new Byte[10000];
> >                        dataToRead = dataToRead - length;
> >                    }
> >                    else
> >                    {
> >                        //prevent infinite loop if user disconnects
> >                        dataToRead = -1;
> >                    }
> >                }
> >            }
> >            catch(IOException ioEx)
> >            {
> >                errorRow.Visible = true;
> >                lblErrorMsg.Text = ioEx.Message;
> >                log.Error(ioEx.ToString());
> >            }
> >            catch (Exception ex)
> >            {
> >                errorRow.Visible = true;
> >                // Trap the error, if any.
> >                log.Error(ex.ToString());
> >            }
> >            finally
> >            {
> >                if (iStream != null)
> >                {
> >                    //Close the file.
> >                    iStream.Close();
> >                }
> >            }
> >        }
> >
> > Thanks in advance for any help.
> >

23 Aug 2006 7:33 AM

Damien

eventura***@gmail.com wrote:

> Thanks for all your input. What it turned to be was my neglecting to
> put in "Response.End()" at the end of this method. The files that were
> being downloaded actually had the emitted HTML response appended to the
> end of the file. It turns out that out of out of a whole cross section
> of files, ranging form movie files to Office and images, the PDF was
> most sensitive to this extra information after the EOF and was gettin
> gcorrupted.
>
>

Glad you sorted it, but can I just ask, why do you continually
reallocate the buffer in your loop? Surely that would increase the
memory pressure?

Damien

23 Aug 2006 4:39 PM

Tim_Mac

hi,
you may prefer to use Response.Close()
in .net 1.1 i ran into ThreadAbortExceptions when using Response.End.
changing this to Response.Close() solved the problem.

tim

30 Aug 2006 10:32 AM

lmcpferreira

Oh thankyou thankyou thankyou

.... same problem!...

22 Aug 2006 5:19 PM

Peter Bromberg [C# MVP]

Try using the (new in ASP.NET 2.0) Response.TransmitFile method.
Peter

--
Co-founder, Eggheadcafe.com developer portal:
http://www.eggheadcafe.com
UnBlog:
http://petesbloggerama.blogspot.com

Show quote

"eventura***@gmail.com" wrote:

> For my application, there users can upload and download files to the a
> webserver. Straightforward enough. However, when they upload a PDF
> file then try to download it, the file seems to be corrupted. The had
> not been a problem before but it seems to have been introduced when I
> refactored the writing to the response to a buffered version to prevent
> potential OutOfMemoryExceptions (you wouldn't believe the size of some
> of these files...).
>
> Here's the code (pretty much jacked from a Microsoft knowledge base
> article):
>
> private void WriteToResponse(string filepath, string contentType,
> NameValueCollection extraHeaders)
>         {
>             Stream iStream = null;
>
>             // Buffer to read 10K bytes in chunk:
>             byte[] buffer = new Byte[10000];
>
>             // Length of the file:
>             int length;
>
>             // Total bytes to read:
>             long dataToRead;
>
>             // Identify the file name.
>             string filename = Path.GetFileName(filepath);
>
>             try
>             {
>                 // Open the file.
>                 iStream = new FileStream(filepath, FileMode.Open,
>                     FileAccess.Read, FileShare.Read);
>
>
>                 // Total bytes to read:
>                 dataToRead = iStream.Length;
>
>                 Response.ContentType = contentType;
>                 if(extraHeaders != null)
>                 {
>                     for (int i = 0; i < extraHeaders.Count; i++)
>                     {
>                         Trace.Write("adding Key value: " +
> extraHeaders.GetKey(i)
>                                     + ". Adding value: " +
> extraHeaders[i]);
>                         Response.AddHeader(extraHeaders.GetKey(i),
> extraHeaders[i]);
>                     }
>                 }
>
>                 // Read the bytes.
>                 while (dataToRead > 0)
>                 {
>                     // Verify that the client is connected.
>                     if (Response.IsClientConnected)
>                     {
>                         // Read the data in buffer.
>                         length = iStream.Read(buffer, 0, 10000);
>
>                         // Write the data to the current output stream.
>                         Response.OutputStream.Write(buffer, 0, length);
>
>                         // Flush the data to the HTML output.
>                         Response.Flush();
>
>                         buffer = new Byte[10000];
>                         dataToRead = dataToRead - length;
>                     }
>                     else
>                     {
>                         //prevent infinite loop if user disconnects
>                         dataToRead = -1;
>                     }
>                 }
>             }
>             catch(IOException ioEx)
>             {
>                 errorRow.Visible = true;
>                 lblErrorMsg.Text = ioEx.Message;
>                 log.Error(ioEx.ToString());
>             }
>             catch (Exception ex)
>             {
>                 errorRow.Visible = true;
>                 // Trap the error, if any.
>                 log.Error(ex.ToString());
>             }
>             finally
>             {
>                 if (iStream != null)
>                 {
>                     //Close the file.
>                     iStream.Close();
>                 }
>             }
>         }
>
> Thanks in advance for any help.
>
>