Problem
Project-wide test coverage is 90.91%, but several modules are significantly below that average. The Codecov project threshold is already catching small drops on PRs, and these low-coverage files are the primary drag.
Files below 85% coverage
| File |
Lines |
Covered |
Missed |
Coverage |
vera/browser/emit.py |
23 |
0 |
23 |
0.00% |
vera/wasm/markdown.py |
289 |
134 |
155 |
46.37% |
vera/parser.py |
45 |
29 |
16 |
64.44% |
vera/cli.py |
503 |
344 |
159 |
68.39% |
vera/wasm/inference.py |
546 |
408 |
138 |
74.73% |
vera/wasm/operators.py |
303 |
236 |
67 |
77.89% |
vera/checker/resolution.py |
98 |
79 |
19 |
80.61% |
vera/checker/control.py |
238 |
200 |
38 |
84.03% |
Files in the 85-88% range (lower priority)
| File |
Lines |
Covered |
Missed |
Coverage |
vera/tester.py |
335 |
285 |
50 |
85.07% |
vera/markdown.py |
413 |
353 |
60 |
85.47% |
vera/smt.py |
550 |
472 |
78 |
85.82% |
vera/codegen/closures.py |
163 |
140 |
23 |
85.89% |
vera/verifier.py |
455 |
399 |
56 |
87.69% |
Notes
vera/browser/emit.py (0%) is the browser bundle emitter — it writes files to disk and is only exercised by vera compile --target browser, which the test suite doesn't call directly. Likely needs a targeted integration test or pragma: no cover if it's inherently side-effectful.
vera/wasm/markdown.py (46%) is the WASM host binding for Markdown operations — many branches handle edge cases in the Markdown AST that existing conformance tests don't exercise.
vera/cli.py (68%) has many subcommands and error paths that aren't tested. The CLI tests exist but don't cover all flags and error conditions.
vera/wasm/inference.py (75%) and vera/wasm/operators.py (78%) have many type-specific branches in the WASM codegen that aren't hit by current test programs.
vera/checker/resolution.py and vera/checker/control.py likely have uncovered error-reporting branches.
Approach
For each file, assess whether the uncovered code is:
- Testable — write targeted unit tests
- Defensive/unreachable — add
pragma: no cover with rationale
- Dead code — remove it
Goal: bring all files above 85%, overall project coverage above 91%.
Codecov dashboard
https://app.codecov.io/gh/aallan/vera?displayType=list
Problem
Project-wide test coverage is 90.91%, but several modules are significantly below that average. The Codecov project threshold is already catching small drops on PRs, and these low-coverage files are the primary drag.
Files below 85% coverage
vera/browser/emit.pyvera/wasm/markdown.pyvera/parser.pyvera/cli.pyvera/wasm/inference.pyvera/wasm/operators.pyvera/checker/resolution.pyvera/checker/control.pyFiles in the 85-88% range (lower priority)
vera/tester.pyvera/markdown.pyvera/smt.pyvera/codegen/closures.pyvera/verifier.pyNotes
vera/browser/emit.py(0%) is the browser bundle emitter — it writes files to disk and is only exercised byvera compile --target browser, which the test suite doesn't call directly. Likely needs a targeted integration test orpragma: no coverif it's inherently side-effectful.vera/wasm/markdown.py(46%) is the WASM host binding for Markdown operations — many branches handle edge cases in the Markdown AST that existing conformance tests don't exercise.vera/cli.py(68%) has many subcommands and error paths that aren't tested. The CLI tests exist but don't cover all flags and error conditions.vera/wasm/inference.py(75%) andvera/wasm/operators.py(78%) have many type-specific branches in the WASM codegen that aren't hit by current test programs.vera/checker/resolution.pyandvera/checker/control.pylikely have uncovered error-reporting branches.Approach
For each file, assess whether the uncovered code is:
pragma: no coverwith rationaleGoal: bring all files above 85%, overall project coverage above 91%.
Codecov dashboard
https://app.codecov.io/gh/aallan/vera?displayType=list