System.Text.Encoding.Default.GetBytes fails

Question

Here is my sample code:

CodeSnippet 1: This code executes in my file repository server and returns the file as encoded string using the WCF Service:

byte[] fileBytes = new byte[0];
using (FileStream stream = System.IO.File.OpenRead(@"D:\PDFFiles\Sample1.pdf"))
{
    fileBytes = new byte[stream.Length];
    stream.Read(fileBytes, 0, fileBytes.Length);
    stream.Close();
}

string retVal = System.Text.Encoding.Default.GetString(fileBytes);  // fileBytes size is 209050

Code Snippet 2: Client box, which demanded the PDF file, receives the encoded string and converts to PDF and save to local.

byte[] encodedBytes = System.Text.Encoding.Default.GetBytes(retVal); /// GETTING corrupted here

string pdfPath = @"C:\DemoPDF\Sample2.pdf";
using (FileStream fileStream = new FileStream(pdfPath, FileMode.Create))  //encodedBytes is 327279
{
    fileStream.Write(encodedBytes, 0, encodedBytes.Length);
    fileStream.Close();
}

Above code working absolutely fine Framework 4.5 , 4.6.1

When I use the same code in Asp.Net Core 2.0, it fails to convert to Byte Array properly. I am not getting any runtime error but, the final PDF is not able to open after it is created. Throws error as pdf file is corrupted.

I tried with Encoding.Unicode and Encoding.UTF-8 also. But getting same error for final PDF.

Also, I have noticed that when I use Encoding.Unicode, atleast the Original Byte Array and Result byte array size are same. But other encoding types are mismatching with bytes size also.

So, the question is, System.Text.Encoding.Default.GetBytes broken in .NET Core 2.0 ?

I have edited my question for better understanding. Sample1.pdf exists on a different server and communicate using WCF to transmit the data to Client which stores the file encoded stream and converts as Sample2.pdf

Hopefully my question makes some sense now.

Answer 1

1: the number of times you should ever use Encoding.Default is essentially zero; there may be a hypothetical case, but if there is one: it is elusive

2: PDF files are not text, so trying to use an Encoding on them is just... wrong; you aren't "GETTING corrupted here" - it just isn't text.

You may wish to see Extracting text from PDFs in C# or Reading text from PDF in .NET

If you simply wish to copy the content without parsing it: File.Copy or Stream.CopyTo are good options.