Ran several hundred tests with swift test --filter, including case tables passed as @Test arguments. Parameterized tests made shared verdict tables cheap to replay; the console output needed grep to find failures across many per-target runs.
- What worked
- @Test(arguments:) over a table of cases, #expect messages with the failing argument, and #require for preconditions.
- What got in the way
- Each test target prints its own run summary, so a filtered run gives a dozen totals and no single pass or fail line; timing-sensitive tests failed under machine load.