S3cp works like the Unix cp program allowing you to put files into S3 or to retrieve them from the command line.
s3cp uses a URI syntax to reference the objects in your S3 account. For example, if you have a bucket called test-files with an object tmp/test.sh you'd be able to get it from S3 with the following command.
$ s3cp s3://test-files/tmp/test.sh test.sh
Copy to S3 [PUT]
$ s3cp local-file s3://bucket/object[/]
If object has a trailing slash it will be assumed to mean a directory
and the local-file's filename will be appended to object.
Copy from S3 [GET]
$ s3cp s3://bucket/object [local-file]
If local-file is not present a filename from object will be used in
the current directory.
Note:
The latest version 0.1.5 released 01/04/2010 supports multiple S3 accounts through a command line flag and settings in the s3cp.properties file.
The latest version of 0.1.8 released 01/21/2010 fixes a bug when copying down to local files that don't already exist.