From Test Debugging to a Reviewable Change

Goal: complete a reviewable bug fix

Use the small project from the TypeScript, Go, or Python guide, or an equivalent existing Java service. Successful editor installation is not the goal: reproduce a failure, inspect it in a debugger, fix it, refactor, and submit a checked change.

The total function returns 3600 for price=1200 and quantity=3. A negative quantity must be rejected instead of returning a negative amount. The tutorial implementations deliberately lack that validation.

Add this failing test to the Python project's tests:

import pytest

from main import total


def test_negative_quantity_is_rejected() -> None:
    with pytest.raises(ValueError):
        total(1200, -1)

Run it before changing total. Break inside total and inspect quantity=-1. Add validation, preserve the positive test, and rerun. If renaming total to subtotal, use Rename Symbol and inspect the diff. Go's return-type change also requires updating callers and tests.

Complete the workflow

1. Record Git state, tool versions, and baseline tests.
2. Add a regression test and confirm its intended failure.
3. Inspect definitions, references, and the call hierarchy.
4. Debug one test and observe inputs and stack.
5. Add validation and widen tests gradually.
6. Rename or extract, then review the diff.
7. Compare editor and CLI transformations.
8. Run non-mutating format, lint, type, and test checks.
9. Stage intended files and record rationale and validation.
10. Reproduce the workflow in another profile or remote environment.

Use the debugger documentation for controls and the language guide for launch details.

Prove convergence

Save main.py in the Python learning project, then run this POSIX-shell comparison. Fixed temporary filenames are for this single-user exercise; use a unique directory for concurrent runs.

cp main.py /tmp/vscode-lab-editor.py
uv run ruff check main.py --fix
uv run ruff format main.py
cmp main.py /tmp/vscode-lab-editor.py
cp main.py /tmp/vscode-lab-first.py
uv run ruff check main.py --fix
uv run ruff format main.py
cmp main.py /tmp/vscode-lab-first.py

The first comparison checks editor versus CLI; the second checks repeated CLI transformations. Reintroduce a harmless space and save again. PowerShell users can use Copy-Item and Get-FileHash. A repository-wide Git diff is unsuitable when unrelated edits already exist.

Decide whether the workflow is ready

If a framework refactoring remains unsupported, record that task's cost and keep an appropriate tool for it. Evaluate completed work rather than an abstract claim of IDE feature parity.

Record evidence

Project and commit:
OS and local/SSH/WSL/container environment:
VS Code and extension versions:
Runtime, compiler, formatter, and linter versions:
Regression test and initial failure:
Values observed in debugger:
Fix and refactoring scope:
Editor/CLI/repeated-run comparison:
Final commands and results:
Remaining framework or tooling limitations:

Earlier CLI verification record

The source guide recorded a 2026-09-06 Linux amd64 verification in temporary projects: JSON/TOML parsing; TypeScript formatting, lint, typecheck, build, and output 3600; Go tests, vet, and race checks; Python pytest, Ruff, Pyright, output 3600, and a negative-input regression; repeated-format convergence.

Its recorded versions were Node 26.8.1, TypeScript 6.0.3, ESLint 10.10.0, Prettier 3.9.6, typescript-eslint 8.69.0, Go 1.27.1, Python 3.12.3, pytest 9.1.1, Ruff 0.16.6, and Pyright 1.1.411. These are historical verification versions, not a latest-version recommendation or the repository's TS 7 contract.

Java execution lacked a JDK. GUI save actions, breakpoints, refactoring, Spring, and remote end-to-end behavior were not verified. This migration does not claim to have repeated those experiments. Follow the exercise in your own environment; the separate formatting lab record describes its own CLI checks.