curl download file and name it <custom-name>.<provided-extension>

Question

Say I have a bunch of URLs like this:

https://domain.com/foo/bar/2d9ec0354908e055acf372

From the URL I don't know what the extension is, and I don't have any other "frontend" information to tell me what kind of file it is.

But I do know what the title of the file is on the frontend. So I would like to just hardcode a title in when saving the file, and then append on whatever extension the real file has. So if the title of the file is "File A of N", I would like to call it "file-a-of-n" (I would just write that out). Then I would like the final result to be:

file-a-of-n.zip

Or whatever the extension ended up being. Maybe instead of .zip this particular file was .tar.gz, so it would end up being:

file-a-of-n.tar.gz

Whatever the extension ended up being, that's fine, want to append that to the file. But I don't want this:

2d9ec0354908e055acf372.zip

So wondering how to do this. Something with curl along the lines of:

curl -o file-a-of-n.(extension) https://domain.com/foo/bar/2d9ec0354908e055acf372

And the end result would be a file like:

file-a-of-n.zip

Answer 1

curl doesn't support what you're trying to do directly, you can either pass it a new filename where to store the results, or let it determine the filename (and possibly extension) from the HTTP headers.

So you should do this in two steps. I recommend letting curl determine the name of the file, therefore downloading 2d9ec0354908e055acf372.zip, which you can then rename to file-a-of-n.zip, preserving the ".zip" part of it.

To make your job easier, create a temporary directory in which to work, and then create subdirectories for the a-of-n's and download the file into that subdirectory, that way you'll have a single file to handle in each subdirectory (easy to tell which one you need to process), which you can then move and rename, while preserving the extension.