Closed dvstans closed 4 years ago
This is out of scope. In many cases it is impossible to detect the filename and extension from a URL. The only options are to let the server (wget) decide what to name the file, or to force the filename on download. Since this feature is only for DOIs, it's probably better to force the filename to be the doi number.
Testing with published data that has a URL without a filename, the extension/filename detection failed resulting in downloading to a cryptic filename (determined by wget). HTTP downloads must conform to the standard filename of the DataFed ID with an optional extension.