ponder-lab / Hybridize-Functions-Refactoring / #2331
80%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 32
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
coverage: 79.588%. Remained the same
#2331

push

github

web-flow 
Run the six issue-374 tests behind an Assume gate on the wala/ML#687 binding (#731)

* Convert the six issue-374 suppressions to inverted pins on wala/ML#687

The six multi-file tests suppressed under #374 (`testPythonSideEffects59`-`61`, `testClassInDifferentFile`, `testClassInDifferentFile4`, `testModule6`) reduce to one upstream defect: a plain `import B` never binds to the sibling script module, while `from B import X` resolves the same module over the same files (minimal pair on `testClassInDifferentFile`/`2`; the import-binding localization and the refuted order-dependence probes are recorded on #374 and wala/ML#687).

Each test now runs and asserts the current broken behavior directly (`assertNull` for the unanalyzed cross-file entities, `assertFalse` for the invisible transitive side-effects), with a `TODO(wala/ML#687)` inversion note, so the suite turns red the moment the upstream fix lands instead of hiding behind `@Ignore`. The orphaned map-based `testPythonSideEffects` helper is removed. If CI diverges from local on any pin, that is wala/ML#687's unconfirmed 2024 environment-sensitivity observation surfacing and belongs on that issue as evidence, not back under a suppression.

Skips drop from 11 to 5 (the remaining five are the #726 re-enable candidates). Full suite green: 586 tests, 0 failures.

Refs #374.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01S1U9TQRo3Wmhfu86DRcBNW

* Re-enable the six issue-374 tests with their original assertions

The pin conversion attempt doubled as the environment experiment and resolved wala/ML#687's open observation with a reversal: on CI (GitHub Actions) the plain-import binding works and all six tests pass their original assertions; the failure is per-machine (this desktop fails deterministically; 2024's provider polarity is inverted), pointing at directory-enumeration order in the loader's script-module resolution (evidence on wala/ML#687).

Acc... (continued)

1700 of 2136 relevant lines covered (79.59%)

0.8 hits per line

Jobs
ID Job ID Ran Files Coverage
1 #2331.1 32
79.59
Source Files on build #2331