Fix copy issues

Bug fixes

Copying files with a different format than utf-8 lead to corrupted destination files.
This was due to reading the file and interpreting it as utf-8 before writing it to the destination.
Now the copying is done with a performant stream solution.

This solution also has two performance benefits:

• Less memory usage, since the read data is immediately written to the output file.
• Faster, because we start writing before the whole source file is read out.