Stub out API test cases related to Reports

[kdelee]

This is another area where we need to stub out tests for functionality that does not exist much yet, so they need to be relatively high level.

Tests for all reports:

• Test that any fact included in any report may be obfuscated/redacted/forward encrypted upon "export" (whatever that may be)
• Test that I may “tag” hosts (ideally via DNS name or network IP address) with “exclude” attribute to communicate to SEAP that these hosts should be considered as paid for through some other arrangement (Example: AWS instance paid through royalty by Amazon itself)

Test by Report type:

• "Raw Report" -- host by host breakdown of all facts collected, rho-like
  • general correctness test like rho
  • unlike rho, it seems we want some determination of what is there (not just contents of /etc/redhat-release which is a string, but a boolean or ternary fact that says YES/NO/MAYBE this is RHEL 4)
• "Density Reports"
  • Test that for each product, I can generate a “density report” that includes the following data about about hosts found with product including:
    • # hosts
    • # of virtual cores
    • # of physical cores
    • # of sockets
    • # of socket pairs
    • Version of product ?
    • Other detail about how we ascertained that the product was present?
    • Report includes sum of virtual cores
    • Report includes sum of physical cores
    • Report includes sum of sockets
    • Report includes sum of socket pairs
    • Report correctly differentiates between redhat and community versions of product
    • Report deduces from raw facts definitive version of rhel and when it cannot deduce definitive version, it indicates this in some way (unknown version?)
    • Report indicates what fact collections this report is sourced from
• "Time series" reports
  • Test that for each product (will we regard EAP 6.y and 7.y the same product?), I can generate a that takes in many fact collections given a start date and end date and source and produces a report that shows how many {hosts, physical cores, sockets, socket pairs, virtual cores} were found with that product at each scan. Must allow some definition of “step value”, i.e. period between fact collections to include.

One outstanding question regarding Density and Raw reports if they will also aggregate data from multiple scans and do any deduplication. Time series by definition must take in multiple fact collections, but these would obstesibly be from different points in time. It is unclear if they must be able to merge fact collections from the same "time step"

[kdelee]

We should suggest that time series reports get punted to the future, because this does not make sense without scheduling

[kdelee]

Sample "density report" as I imagine it:

JBOSS EAP DENISTY REPORT
hosts	cores	sockets
one.com	27.5	1
two.com	6	2
three.com	8	3
summary	41.5	6

Obviously the 27.5 cores is silly

[kdelee]

• See Update OS Count endpoint quipucords#339 for OS count report work progress
• See RHEL Deployment Report quipucords#340 for RHEL report progress
  Also related to RHEL report, to be able to use network and vcenter scan for report, must de-duplicate
• Deduplication between systems from multiple data source inspection  quipucords#338.
• De-duplication must occur for network sources as well Deduplication within systems from an Network (Ansible) data source inspection quipucords#337

[kdelee]

Also add test stub (functionality not implemented yet) for redacting fields of reports at some "export" stage