The real physical path of the File is abstracted in Mendix 5. You can only obtain an InputStream of the contents. It is described at the Moving from 4 to 5 section of the documentation here.
You can retrieve the filetype from the extension of the name. For example: If the name ends with .pdf it is an acrobat file. As an alternative you can use Apache Tika.
It is not recommended to use or know the physical path. Instead use
Core.getFileDocumentContent(context, fileDocument)
Using a webservice call is easier if allowed.