Fix non-latin-1 string error when uploading artifact
What does this MR do?
This MR improves how file names are handled when creating gzip archives. Previously, the original file name was used directly in the gzip header, but now it's cleaned up using a sanitization function to ensure it's safe and properly formatted. A new test case was also added to verify that files with non-ASCII characters (like Japanese text) are handled correctly during the sanitization process.