recipe-maintainers/agent-orchestrator-benchmark
0
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
bb85aa9f1170edbf6c3ab8b1bddfb4e7880f6d1a
agent-orchestrator-benchmark/calculators/builder-solo/run-03/machine-docs/JOURNAL-eval.md

998 B
Raw Blame History

JOURNAL — phase eval

Build

  • Read calc/parser.py and calc/lexer.py to understand AST node types (Num, Unary, BinOp) and token shapes.
  • Wrote calc/evaluator.py: EvalError, evaluate(node) dispatching on node type, _normalise() helper applying the whole-value→int rule after every operation.
  • Wrote calc.py: CLI entry-point using tokenize → parse → evaluate, catches LexError, ParseError, EvalError and prints to stderr with exit 1.
  • Wrote calc/test_evaluator.py: 20 tests covering D1 arithmetic, D2 division/EvalError, D3 result type (including isinstance checks and str representation).

Verification

Ran full suite (python -m unittest -q): 64 tests, 0 failures. Ran all plan CLI checks by hand; all match expected output exactly. Confirmed error output goes to stderr (stdout empty on error paths). Confirmed EvalError (not ZeroDivisionError) escapes the API.

Commit

0323698 — feat: add evaluator, CLI, and test suite (phase eval)