prs_per_repo script should optimize interactions with GitHub

[chessbyte]

By using GitHub search

[chessbyte]

I think that most of the logic of getting PRs for a particular sprint or set of sprints should be pushed back into the SprintStatisics class (or something named more appropriately). This would make the SprintStatistics class the single communicator with GitHub and, thereby, the single source of truth for all the other data massaging.

[Fryguy]

I think that most of the logic of getting PRs for a particular sprint or set of sprints should be pushed back into the SprintStatisics class (or something named more appropriately). This would make the SprintStatistics class the single communicator with GitHub and, thereby, the single source of truth for all the other data massaging.

I think we need to look at our multiple repos and maybe come up with something to bring then all together. This repo is starting to resemble managiq-release which is my go to repo for anything that calls out to github/git for meta reasons. As such, I'd like to avoid duplicating effort across them.