• Home
  • Features
  • Pricing
  • Docs
  • Announcements
  • Sign In
Warning: This build has drifted.
The coverage report for this pull request build may be inaccurate because its base commit is no longer the HEAD of its target branch.
This means it includes changes from outside the original pull request, including, potentially, unrelated coverage changes.

    • Learn more: For more information on this, see Tracking coverage changes for pull request builds.
    • Fix now: For a quick fix, rebase this PR at GitHub. Your next report should be accurate.
    • Prevent going forward: To avoid this issue with future PRs, see these Recommended CI Configurations.
New Repo Setting:
INCLUDE COVERAGE % WITH WARNINGS ABOUT DRIFTED BUILDS?

Enabling this setting will include a (potentially inaccurate) coverage % with warning messages in status updates for drifted builds.

Adjust setting

99designs / gqlgen / 29649114412
0%
master: 0%

Build:
Build:
LAST BUILD BRANCH: fix/batch-resolver-value-slices
DEFAULT BRANCH: master
Ran 18 Jul 2026 05:53PM UTC
Jobs 1
Files 0
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

18 Jul 2026 02:59PM UTC coverage: 0.0%. Remained the same
29649114412

Pull #4262

github

webel-crew
fix(codegen): mask stale generated outputs from the loader instead of unlinking them

Root cause of the check-generate CI failures on this PR, found by running
`go generate ./...` locally: base and master pass, every commit of this
branch failed deterministically (the "different type each run" in CI was
map-iteration order picking which missing type to report first — the
failure itself was 100% reproducible).

The old upfront `syscall.Unlink` of the exec + model outputs was not
redundant — it was LOAD-BEARING for every config that autobinds the
package its models are generated into (all codegen/testserver fixtures,
the federation examples):

  1. old: unlink first → stale models file GONE at schema-load time →
     autobind finds nothing → modelgen generates all types → OK.
  2. this branch (before this commit): stale models file VISIBLE at
     load → autobind binds the previously-generated types as if they
     were user-written models → modelgen SKIPS them → the freshly
     written models file loses them → the exec build reload fails with
     "unable to find type: <pkg>.<Type>".

But the unlink is also exactly what #2345/#3505 are about: delete-then-
interrupted leaves the user with no generated file at all. Both
properties need to hold, so replace the destructive unlink with a
NON-destructive loader mask:

  - internal/code.Packages gains an Overlay (packages.Config.Overlay)
    threaded into every Load, plus MaskFile/UnmaskFile.
  - api.generate masks each existing output (exec + model) with a
    package-clause-only stub read from the file's own package clause
    (empty bytes would be a parse error and break loading the rest of
    the package). Missing/unparseable files are left alone — nothing
    stale to bind, matching the old unlink's no-op there.
  - config.Config owns the overlay map (MaskGeneratedFile) because
    LoadSchema RECREATES c.Packages mid-generation — an overlay set
    only on the first instance would silently vanish. The ... (continued)
Pull Request #4262: fix(codegen): write generated files atomically to survive interrupted generation

0 of 0 relevant lines covered (NaN%)

0.0 hits per line

Jobs
ID Job ID Ran Files Coverage
1 29649114412.1 18 Jul 2026 05:53PM UTC 0
0.0
GitHub Action Run
Source Files on build 29649114412
Detailed source file information is not available for this build.
  • Back to Repo
  • Pull Request #4262
  • PR Base - master (#29026859304)
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