Test Runner

Requirements

• Export TestRunner as a named and default class. test(name, callback) registers a test; skip(name) registers a skipped test. Both return the same runner for chaining. Names must be nonempty strings after trimming; preserve their original text. Invalid names or non-function callbacks throw TypeError without registering anything. Duplicate names are allowed.
• run() returns a Promise of { total, passed, failed, skipped, results }. Each result has name and status: passed, failed, or skipped. Only failed results have an own error property, containing the exact thrown value or rejection reason, including undefined.
• Run active callbacks serially in registration order, awaiting each. Any returned or resolved value means passed. A throw or rejection means failed, but later tests must still run. Skipped registrations do not call any callback.
• Take a snapshot of registrations at run start. Tests registered during a run are included only in future runs. Each run executes the full snapshot again and returns fresh report/result objects. Empty runners produce zero counts.
• Reject overlapping run() calls on the same runner with Error. They must not invoke callbacks or disturb the active run. Independent runner instances may run concurrently. Release the running flag after completion so later runs work.
• Implement scheduling and aggregation yourself; do not use a test-runner library. Timeouts, parallel scheduling, hooks, assertions, retries, and duration measurement are outside scope. TypeScript callbacks return unknown or PromiseLike<unknown>; error values are unknown.

Examples

Input: const r = new TestRunner(); r.test("ok", () => 1).test("bad", () => { throw "oops"; }).skip("later"); await r.run();
Output: { total: 3, passed: 1, failed: 1, skipped: 1, results: [{ name: "ok", status: "passed" }, { name: "bad", status: "failed", error: "oops" }, { name: "later", status: "skipped" }] }

Input: await new TestRunner().run();
Output: { total: 0, passed: 0, failed: 0, skipped: 0, results: [] }