Store job binary paths in job execution objects

Sahara should store the list of binary urls used during a job execution in the job execution object, so that a user can easily see what binaries were used in a previously run job.

Similar storage of datasource urls is part of this blueprint https://blueprints.launchpad.net/sahara/+spec/edp-datasource-placeholders.

Storing the job binary urls is good for two reasons:

1) It will be consistent with the storage of data source urls, and a user may look in one place to see all the details of a previously run job

2) It allows job binary objects to be edited even if an existing job references them. Currently the only way to find the urls for job binaries used in an execution is to trace back from the job execution object to the job object and then to the job binary objects. So, editing a job binary object to change a path will make it impossible to know for certain the path used in a job execution. For that reason, editing job binaries referenced by a job execution object would not be allowed (but this change would remove that restriction)