Challenge 2: 100% Coverage, Still Buggy — Possible Solution
====================================================================

CONCRETE EXAMPLE:

function calculateDiscount(price, isMember) {
  if (isMember) {
    return price * 0.8;   // BUG: should be 0.9 per the actual 10% member discount spec
  }
  return price;
}

it("applies a discount for members", () => {
  const result = calculateDiscount(100, true);
  expect(result).toBeGreaterThan(0); // a real, but nearly meaningless assertion
});

WHY THIS ACHIEVES 100% COVERAGE WHILE MISSING THE BUG
------------------------------
Calling calculateDiscount(100, true) executes BOTH the isMember branch
AND the multiplication line — every line and branch inside the
function runs at least once during this single test, which is exactly
what a line/branch coverage tool measures. From a pure coverage
report's perspective, this function is "fully tested."

But the assertion — expect(result).toBeGreaterThan(0) — doesn't check
the ACTUAL VALUE returned, only that it's a positive number. The real
bug (multiplying by 0.8 instead of the correct 0.9, a 20% discount
applied instead of the intended 10%) produces a result (80) that is
still greater than 0 — so this hollow assertion passes regardless of
whether the discount math is correct or wrong.

This directly matches this chapter's own point: coverage measures
whether a line EXECUTED, never whether the test's ASSERTIONS were
strong enough to actually catch a wrong result. A stronger assertion —
expect(result).toBe(90) — would have caught this exact bug immediately,
using the SAME lines of code, with the SAME 100% coverage number, just
a meaningfully different assertion.