The upstream source collection code is currently built included in two different files - "upstream-source-collector.py" and "poll-upstream-sources.py". The files reside in "jobs/confs/python-scripts" which in turn mean they are mean for embedding into jobs by JJB.

This situation causes several issues:

1. It is not easy for users to use these scripts locally, and hence its difficult to build or test repos that include upstream source dependencies locally

2. There is code duplication between the two python scripts

3. Adding certain features is harder because it requires one of the scripts to have access to code that resides in the other script (See https://ovirt-jira.atlassian.net/browse/OVIRT-1641#icft=OVIRT-1641 for example)

4. It is impossible to use the code from pipelines (See OVIRT-1745)

5. It is not east to add tests for the code

It is desirable to have the upstream source handling code be concentrated into a single tool that resides in the 'scripts' directory.