I noticed I wait quite a while on pull request reviews. Also there is at times a lot of back and forth which slows down delivery.
I wanted to gather some data to support my intuition and help me make this issue more visible and potentially create some KPIs for the team.
The program speaks to GitHub using an API token. It accepts an organization and a team as input. It retrieves a list of repositories for that team and then recent pull requests for every repository (date range is configurable, by default last two weeks).
Then it calculates how many working hours passed till first review (first time somebody found time to have a look), how many reviews there were in total on the pull request and how long it took to approve.
- Checkout the repository.
- Install dependencies
go get. - Build and run:
go build
./ghs stats -org <org-name> -team <team-name> -t <api-token>The token needs read:org and repo rights. You can create one in Settings -> Developer Settings -> Personal Access
Tokens.
Currently more of an MVP. Future improvements:
- Paging get pull requests when last 30 PRs is not enough for specified range.
- Dockerization