query(sql, options?)
Execute a SQL query against the oleander lake using DuckDB. Returns structured results with columns, types, and rows.
const result = await client.query(
"SELECT * FROM public.iris_dataset LIMIT 10"
);
const columns = result.results?.columns; // ["sepal_length", "sepal_width", ...]
const rows = result.results?.rows; // [[5.1, 3.5, ...], ...]
const count = result.row_count; // 10
const elapsed = result.execution_time; // "42ms"
Parameters
The SQL query to execute. Supports all DuckDB SQL syntax.
When true, persists query results as a table. The table name is returned in saved_table_name.
Saving results
Use save: true to persist query results as a table for later use.
const result = await client.query(
"SELECT * FROM public.iris_dataset LIMIT 10",
{ save: true }
);
if (result.saved_table_name) {
// Run a follow-up query against the saved table
const followUp = await client.query(
`SELECT avg(sepal_length) FROM ${result.saved_table_name}`
);
}
Iterating over results
const result = await client.query("SELECT name, score FROM public.rankings");
const { columns, rows } = result.results!;
const nameIdx = columns.indexOf("name");
const scoreIdx = columns.indexOf("score");
for (const row of rows) {
const name = row[nameIdx] as string;
const score = row[scoreIdx] as number;
// process each record ...
}
Return type: LakeQueryResult
| Field | Type | Description |
|---|
success | boolean | Whether the query executed successfully |
results | object | Query results with columns, column_types, and rows |
row_count | number | Number of rows returned |
execution_time | string | Query execution time (e.g., "42ms") |
saved_table_name | string | Table name if save: true was used |
query | string | The original SQL query |
