Original files that contain encoded characters don't load on default site paths

[ericras]

To reproduce bug:

• Be on a site with a default path and no domain like https://cms.unl.edu/sample/site/
• Upload a file with a name like Morrill%20Hall.jpg that contains an encoded space
• Try to access the original file.

Like so:
https://cms.unl.edu/ianr/east-campus-discovery/sites/unl.edu.ianr.east-campus-discovery/files/2024-05/Morrill%2520Hall.jpg?itok=0GnZVcmW

Workarounds:

• Have a domain directed at the site like sample.unl.edu
• If outputting an image, use an image style.

[beckyaiken]

Just wanted to note that this seems to be true of PDFs as well: https://cms.unl.edu/ianr/sites/unl.edu.ianr/files/media/file/IANR%2520Liaison%2520Committee%2520Notes%2520December%252015%25202017.pdf