CODEWITHSUNDEEP
c#silverlightencodingwebclient
Peter Smartt
Peter Smartt

Reputation: 358

Cannot download .pdf file in Silverlight

I am fairly new to Silverlight. I am trying to download a .pdf file (and a couple of other formats) in Silverlight. The user clicks a button, the system goes and gets the URI, then shows a SaveFileDialog to obtain the location to save the file. Here is a code snippet:

WebClient wc = new WebClient();
wc.DownloadStringCompleted += (s, e3) =>
{
    if (e3.Error == null)
    {
        try
        {
            byte[] fileBytes = Encoding.UTF8.GetBytes(e3.Result);
            using (Stream fs = (Stream)mySaveFileDialog.OpenFile())
            {
                fs.Write(fileBytes, 0, fileBytes.Length);
                fs.Close();

                MessageBox.Show("File successfully saved!");
            }
        }
        catch (Exception ex)
        {
            MessageBox.Show("Error getting result: " + ex.Message);
        }
    }
    else
    {
        MessageBox.Show(e3.Error.Message);
};

wc.DownloadStringAsync("myURI", UriKind.RelativeOrAbsolute));

The file gets saved OK, but it is about twice as big as the original and is unreadable. e3.Result looks about the right size (5Mb), but I suspect it contains a lot of extraneous characters. FileBytes seems to be about two times too big (11Mb). I wanted to try DownloadDataAsync instead of DownloadStringAsync (hoping it would resolve any encoding issues), but Silverlight has a very cut-down version of System.Net.WebClient and does not support DownloadDataAsync (it won't compile).

I am fairly sure it is an encoding problem, but I cannot see how to get around it.

Upvotes: 0

Views: 1767

Answers (1)

Martin Liversage
Martin Liversage

Reputation: 106796

PDF files are binary and not encoded using UTF8. To download a PDF file using Silverlight you need to use the OpenReadAsync method of the WebClient class to start downloading the binary data of the file, and not the DownloadStringAsync method as you seem to be doing.

Instead of handling the DownloadStringCompleted event you should handle the OpenReadCompleted event and write the received bytes to the stream of the local PDF file. If you set the AllowReadStreamBuffering to true the OpenReadCompleted event is only fired when the entire file has been downloaded providing you with the same behavior as the DownloadStringCompleted. However, the entire PDF file will be buffered in memory which may be a bad idea if the file is very large.

Upvotes: 1

Related Questions