We have a few "expected" errors:
- We killed the execution of the command because it timed out (in one of 3 ways)
- We were given a string to parse that is too long
In these cases, we currently return an empty command which is implicitly "failed" since it has no arguments.
However, in these situations it is likely more helpful to throw an exception. This should be handled such that:
- When running the CLI, this message is returned to the user (likely via logging)
- When running via the Python API, this can be caught
- When running in the Automation repo, we can store a structured error
We have a few "expected" errors:
In these cases, we currently return an empty command which is implicitly "failed" since it has no arguments.
However, in these situations it is likely more helpful to throw an exception. This should be handled such that: