Map node (legacy)
The Map node pairs upstream columns to target columns via drag-drop, without the expression editor Transform provides. Use it when the shaping is a straight rename or a subset selection and you don't need derived fields. The Configure modal treats "map" and "transform" as siblings for backwards compatibility.
When to use
- You want a low-friction "pick the columns" node for a straightforward projection.
- The target schema is known and you'd like Map's auto-suggest to propose matches by column name similarity.
- You don't need type coercion or derived fields.
Ports
| Port | Direction | Kind | Notes |
|---|---|---|---|
in | in | data | Rows to pair. |
out | out | data | Rows containing only the mapped columns. |
error | out | control | Fires on validation failures. |
notify | out | control | Notification policy port. |
Configuration
- Field mappings. An ordered
{ from, to }list. Each row picks one upstream column and one target column. - Auto-suggest. The Configure modal proposes matches by column name similarity when the target schema is declared.
Configure Action walkthrough
- Modal opens with two columns — upstream on the left, target on the right — sourced from the connected node's declared schemas.
- Drag from an upstream column to a target column to pair them. Or double-click a target column to auto-pick the highest-similarity upstream column.
- Unpaired target columns emit
nullif the target expects them; drop them from the pair list to omit entirely.
Runtime behaviour
- Per-row projection: the output row contains only the paired target columns with their upstream values.
- No expression evaluation; the value passes through unchanged.
Failure modes
- Missing target column. Falls back to
null. - Missing upstream column. The paired target column emits
nullfor that row.