Use different return codes to differentiate user from system errors

[pablochacin]

Presently in the event of any error, k6deps will return the same 1 return code.

In some use cases, it would be convenient to differentiate if the error was produced by the user-supplied data (e.g. corrupted script archive) or by other reasons not under the user's control (e.g. lack of disk space for processing input).

The suggested solution is to use different return code. For example

• 1 for errors related to user input
• 2 for other errors