• Home
  • Features
  • Pricing
  • Docs
  • Announcements
  • Sign In

MapColonies / shigola / 85842fb54
48%

Build:
DEFAULT BRANCH: master
Ran 26 Aug 2026 01:15PM UTC
Jobs 1
Files 204
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst

26 Aug 2026 01:11PM UTC coverage: 48.173% (+0.4%) from 47.744%
85842fb54

push

github

web-flow
test(ci): record the coverage baseline and gate CI on a floor (MAPCO-11480) (#7)

* test(ci): add the coverage floor tool

First half of the coverage gate: the profile parser, the baseline format and
the floor comparison, with tests. The measured baseline, the CI wiring and the
CONTRIBUTING entry are still to come, so nothing enforces anything yet.

The floor lives in the baseline file rather than in the workflow, so the number
and the measurement it came from cannot drift apart. Repeated blocks in a
merged profile count their statements once -- `go test ./...` can emit the same
block from several test binaries, and adding those up would inflate the
denominator and quietly deflate every percentage.

MAPCO-11480

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

* test(ci): leave headroom below the measured coverage floor

The floor was rounded down to a whole percent, which is only a margin when the
measurement is not already just above one. The first real baseline measured
44.01% and would have recorded a floor of 44.00 -- under two statements of
headroom once BelowFloor's rounding tolerance is taken off. A gate that thin
fires on ordinary statement movement rather than on a regression, and a gate
that fires on noise gets edited rather than read.

floorFor now drops a further whole percent when truncation alone leaves less
than half a point, so the floor is derived rather than typed at a terminal and
regenerating reproduces the same reasoning.

Also covers two properties the package relied on but never asserted: that
renderBaseline's output parses back through ParseBaseline unchanged, and that
regenerating on a machine with fewer integration gates keeps the recorded floor
instead of quietly lowering it.

* test(ci): record the coverage baseline and gate CI on the floor

Coverage was measured and uploaded to Coveralls but gated nothing, so a
regression was only visible if somebody went looking.

ci/coverage-baseline.txt records the m... (continued)

227 of 292 new or added lines in 3 files covered. (77.74%)

9823 of 20391 relevant lines covered (48.17%)

203.79 hits per line

Uncovered Changes

Lines Coverage ∆ File
48
65.71
ci/coverage/main.go
9
89.66
ci/coverage/coverage.go
8
87.69
ci/coverage/baseline.go
Jobs
ID Job ID Ran Files Coverage
1 85842fb54.1 26 Aug 2026 01:15PM UTC 204
48.17
GitHub Action Run
Source Files on build 85842fb54
  • Tree
  • List 204
  • Changed 0
  • Source Changed 0
  • Coverage Changed 0
Coverage ∆ File Lines Relevant Covered Missed Hits/Line
  • Back to Repo
  • 85842fb5 on github
  • Prev Build on master (#CAFC4F691)
  • Next Build on master (#57A221C73)
  • Delete
STATUS · Troubleshooting · Open an Issue · Sales · Support · CAREERS · ENTERPRISE · START FREE TRIAL · SCHEDULE DEMO
ANNOUNCEMENTS · TWITTER · TOS & SLA · Supported CI Services · What's a CI service? · Automated Testing

© 2026 Coveralls, Inc