Hello,

I'm on Ubuntu 11.04. I have read around about how to use curl to download a list of URLs from a text file, and everyone says to use . This is what the curl man page says as well. However, for even a simple file with one URL, this command outputs a bunch of weird symbols for me instead of downloading the file.

For example, I have a text file "test.txt" with one line in the following format:

I use the curl command to download this file:

...and get something similar to the following output for a long time before the command finally exits:

No downloaded file shows up after the command is complete. I have no idea why the curl command is not working correctly here. Any help would be appreciated.

Curl is not a download mgr.
The -K switch allows curl to read from a config file.

If you want a download mgr use
wget
gwget
firefox extension
fetch
axle

or a host of others.

You can use curl to see where a file is.
http://linux.die.net/man/1/curl

Instead of curl, you can use wget. Like this:

The file must simply contain one URL on each line.

I know I could use wget, but the idea here was that curl can avoid re-connecting to the FTP server for every single file in a list of URLs, thereby saving a whole ton of time during the download compared to wget, which goes through the trouble of reconnecting for every file. I have easily downloaded lists of images with curl before as well using its sequence option if the URLs are similar, so I know that it is possible, even if that is not its main purpose.

"curl -K" doesn't create files, it writes the downloaded data to stdout.

The data might have not been text, which explains the funny characters.

You specify an URL, but no output file. So curl writes the downloaded file to the standard output, which will obviously produce weird symbols in case of a JPEG-file.

I found that curl has the "--remote-name-all" option, which causes it to create files named like the remote files and download to them.

Try this:

Excellent. This is what I was looking for. Thanks a bunch.