Input · Files
Parquet Reader
Reads a Parquet file and keeps the column types stored in it.
Also known as: Columnar, Arrow
How it works
The file is read through a drone with the types stored in it, so nothing is sniffed and nothing is re-typed. For a GeoParquet file the drone reports the coordinate reference system and the node carries it downstream, which is what the metric spatial nodes such as Buffer need. The path is typed in File path or bound to a column of the Loop's Current item, so a loop over a Directory reads one file per iteration.
Inputs and outputs
Inputs
None — this node starts a workflow.
Outputs
data— Table
Settings
The Attributes panel of Parquet Reader, as in the app: the controls work, but nothing runs — there is no data behind it.
Example
A folder of monthly extracts, sales_2026-01.parquet onwards. Drop a Parquet Reader into a Loop fed by a Directory, bind File path to the item's path column, and the loop collects every month into one table.