DuckDB¶
All integration guides. Fetch Arrow data from Flight, register it in a local DuckDB connection, and query it there. This path transfers data before DuckDB executes SQL; it does not require a DuckDB Flight extension.
Install and query a SQL result¶
Follow the Python environment and sample table setup, which includes DuckDB. Run:
import duckdb
import adbc_driver_flightsql.dbapi as sql
with sql.connect("grpc://localhost:8815", autocommit=True) as conn, conn.cursor() as cur:
cur.execute("SELECT id, label, amount FROM flight_tool_sample ORDER BY id")
table = cur.fetch_arrow_table()
with duckdb.connect() as local:
local.register("redis_rows", table)
result = local.execute("SELECT COUNT(*), SUM(amount) FROM redis_rows").fetchone()
assert result == (3, 7.5)
print(result)
Apply remote filters and select only needed columns in the ADBC query to limit
the transfer. fetch_arrow_table() materializes that result in client memory.
Subsequent DuckDB SQL runs locally; it does not update Redis.
Query a named dataset¶
Download a dataset using the Python Flight example,
reading every returned endpoint. Inside a DuckDB connection, call
local.register("features", received) and query features. You can join the
registered data to local DuckDB tables or files.
Validation and cleanup¶
Run make clients-smoke against a disposable server. Recorded coverage includes
local aggregates over Flight SQL results and ordered/NULL results over a native
Flight dataset; see versions and limits.
No direct DuckDB Flight connector was tested. The Arrow/ADBC write-back
recipe below passed in the linked live guide run. Remove the
sample table using the shared cleanup.
Write and verify a SQL table¶
Run in the environment installed above. This example writes the tool's Arrow output through ADBC, reads the persisted values back and drops its unique table. The live recipe run passed with the documented client requirements. It does not establish tool-native writer API support.
import uuid
import pyarrow as pa
import adbc_driver_flightsql.dbapi as sql
import duckdb
with duckdb.connect() as local:
output = local.sql("SELECT 1::BIGINT AS id, 2.0::DOUBLE AS amount").to_arrow_table()
name = "guide_" + uuid.uuid4().hex
with sql.connect("grpc://localhost:8815", autocommit=True) as conn, conn.cursor() as cur:
try:
cur.adbc_ingest(name, output, mode="create")
cur.execute(f'SELECT * FROM "{name}" ORDER BY id')
assert cur.fetch_arrow_table().equals(output)
finally:
cur.execute(f'DROP TABLE IF EXISTS "{name}"')
cur.fetch_arrow_table()