Decide the coverage threshold gate once the post-merge figure is published (follow-up to #3940)
- Dominant language
- Go
- Stars
- 20.9k
- Forks
- 1.8k
- Avg merge
- 5d 18h
- Merged PRs (30d)
- 39
Description
**Context**
#3940 folds the datasource and metrics-exporter submodules into the reported coverage
figure. Before it, `parse_coverage` and `upload_coverage` merged only the Example and PKG
profiles, so the published percentage excluded every module under
`pkg/gofr/datasource/*` and `pkg/gofr/metrics/exporters/*` while presenting as a
whole-repo number.
A threshold gate used to live in `parse_coverage` as a commented-out 92% check. #3940
deliberately did not restore it, and review agreed the reasoning was right but asked that
the decision live in the queue rather than in a YAML comment. This is that issue.
**The decision**
Once a run on `development` has published the post-merge figure, decide:
1. **Restore the gate** at a floor picked from that number, or
2. **Delete** the commented-out block and keep `parse_coverage` as reporting only.
A floor cannot be chosen before the number exists — #3940 moves the reported figure for
the first time in a long while, and picking 92% (or anything else) beforehand is a guess.
**What the number now includes**
- Example, PKG and submodule profiles, merged identically in `parse_coverage` and
`upload_coverage` so the PR figure and the qlty figure cannot drift.
- Generated mocks are stripped (`grep -v '/mock_'`) in both merges, matching the filter
the PKG job already applies before uploading its profile. Without it the two halves of
one number were computed under different rules — 66 `mock_*.go` files live under the
submodules.
**Where the gate goes**
`.github/workflows/go.yml`, in `parse_coverage`, after the `Parse code-coverage value`
step. `CODE_COVERAGE` is already exported to `$GITHUB_ENV` and the figure is written to
`$GITHUB_STEP_SUMMARY`.
Contributor guide
Research direction
Inspect .github/workflows/go.yml, especially parse_coverage after the “Parse code-coverage value” step, and review the development run after #3940 publishes its figure. Decide whether the post-merge number supports restoring a threshold or removing the commented-out block; done means the chosen policy is reflected in the workflow and the coverage result remains consistent with the published figure.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- github-actions, go
- Domain
- ci-cd, testing-qa
- Issue type
- Feature
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 48/100