github-actions[bot]
commented
7 months ago Coverage report
St.:grey_question: |
Category | Percentage | Covered / Total |
|---|---|---|---|
| 🟢 | Statements | 86.39% (+0.02% 🔼) |
2373/2747 |
| 🟡 | Branches | 73.77% (+0.1% 🔼) |
2208/2993 |
| 🟢 | Functions | 88.89% | 424/477 |
| 🟢 | Lines | 86.72% (+0.02% 🔼) |
2293/2644 |
Test suite run success
451 tests passing in 31 suites.
Report generated by 🧪jest coverage report action from ceb446f9288a0dbe977b8bb8cb5200c4a6a8fcfd
Summary
287 contains CMS69 which includes a
!~(not equivalent) operation in the CQL that has some undesirable coverage calculation behavior due to the way Not Equivalent and Not Equal operations are translated from CQL to ELM json. This PR provides a workaround so that Not Equivalent and Not Equal clauses behave like Equal and Equivalent clauses.New behavior
Before, in order to reach 100% clause coverage with CMS69, test cases where
BMIEncounter.classWAS equivalent to 'virtual' AND whereBMIEncounter.classwas NOT equivalent to 'virtual' needed to be provided. This behavior did not match the clause coverage highlighting where only the not equivalent test case was needed for the clause to be highlighted as covered. Now, the coverage calculation matches the coverage highlighting. For CMS69 to reach 100% clause coverage, ONLY a test case whereBMIEncounter.classis NOT equivalent (!~) to 'virtual' is needed.Code changes
ClauseResultsHelpers.ts- a workaround was added to for this specific case infindAllLocalIdsInStatementso that the result of the equal or equivalent is now the result of the not.Testing guidance
npm run checknpm run cli -- detailed -m <path-to-CMS69> --patients-directory <path-to-CMS69-patients-directory> --trust-meta-profile true --debugon the master branch with the test cases provided in the should result in 96% clause coverage.!~clause is now included in the calculation.npm linkthis branch to fqm-testify to play around with different test cases.