Skip to content
Nembrix

Result grid

The result grid uses TanStack Virtual under the hood, so a million rows costs the same as 200.

Result grid

Things you can do

Section titled “Things you can do”
  • Click a cell to select; ⌘C copies the text.
  • Right-click a cell for Copy / Copy as JSON / Copy row (TSV) / Edit / Set NULL. See Editing cells.
  • Click a column header for a column summary (distribution, uniques, min/max, top values). Each top value is a one-click “filter by this”.
  • Drag the right edge of a column header to resize; widths persist.
  • Click the pin in the gutter to pin a row to the top.
  • Quick search input in the toolbar — substring search across the rows already loaded (not a SQL filter).

Row count

Section titled “Row count”

The toolbar shows N of M (est.) rows:

  • N — rows currently rendered (capped by Limit).
  • M — total in the table. Two-phase fetch: first a planner estimate from pg_class.reltuples, then an exact COUNT(*) wrapped in SET LOCAL statement_timeout = 2500 so huge tables don’t hang the UI. The “(est.)” suffix disappears when the exact count lands.

Empty results render too

Section titled “Empty results render too”

If a query returns zero rows but the columns are known, the grid still draws its header so you can see the schema. A centered “No rows match — the table or query returned 0 results” hint floats over the empty body.