pub fn render_result_table(records: &[VerifierCellRecord]) -> Option<String>Expand description
Render the per-cell records into a summary grid: one row per topology preset, one column per scheduler, aggregating across kernels. Each cell is:
- ✅ every kernel that ran this (topology, scheduler) passed,
- ❌ every kernel that ran it failed,
- 🇽 mixed — at least one kernel passed AND at least one failed,
-no kernel ran this (topology, scheduler) (e.g. the scheduler’s constraints rejected the preset).
After the grid, the specific failing (scheduler, kernel, topology)
combinations are listed so a ❌ or 🇽 cell can be drilled to the exact
kernel(s) that failed. Returns None for an empty record set (the
caller prints nothing).
Rows, columns, and the failing list are BTreeSet-sorted so the same run renders the same output (shell-pipeline stable). The header line carries a ✅/❌/🇽 tally counting grid cells.