Skip to content
BifrostQL
Docs

Drag-and-Drop Pivot Table Designer

Building a cross tabulation in the pivot wells — Category into Rows, status into Columns, a counted measure into Values. The server cross-tabs; only the summary comes back.

The pivot UI is a drag-and-drop pivot table designer over the server pivot surface — the table’s <table>Pivot root field. The server does all cross-tabbing; the UI never groups fetched rows itself. A pivot configuration saves as a saved object.

Field wells

Section titled “Field wells”
  • Rows — one or more row-key columns (a functional multi-field well; all selected keys go in one request).
  • Columns — a single pivot column.
  • Values — a column plus an aggregate op (count, sum, avg, min, max).

The field list is the schema-derived columns for the chosen source. Moving a field between wells triggers exactly one re-query after a debounce interval (a debounce regression is caught by test).

Sources, results, and errors

Section titled “Sources, results, and errors”
  • The same configuration works against a table source and a saved-query source, producing identical results for equivalent data.
  • The result grid renders dynamic pivot-value columns against a fixed row-key column group; NULL pivot values appear as an explicitly labeled category.
  • When a later pivot request returns the server cardinality-guard error, the UI surfaces a message naming the offending column and suggesting a filter — the previous grid stays visible rather than blanking.

Export uses the shared export utility.

Section titled “Related”