Generator and table-valued functions

Preview Spark generator functions in SparkSQL and Databricks compatibility modes. Both modes share these functions. Support is one of: Supported, With limitations, or Not supported. Supported means every example in the Spark reference returns what Spark returns; a note under With limitations quotes a call that differs.

Function	Support	Notes
`RANGE`	Supported
`EXPLODE`	With limitations	Supported in a `FROM` clause, as in `SELECT col FROM explode(array(1, 2))`. Not supported in the select list.
`POSEXPLODE`	With limitations	Supported in a `FROM` clause, as in `SELECT pos, col FROM posexplode(array('a', 'b'))`. Not supported in the select list.
`STACK`	With limitations	Supported in a `FROM` clause, as in `SELECT * FROM stack(2, 1, 2)`. Not supported in the select list.
`EXPLODE_OUTER`, `INLINE`, `INLINE_OUTER`, `POSEXPLODE_OUTER`	Not supported

Limitations

A generator in the select list, such as SELECT explode(arr) FROM t, is not supported. Call it in the FROM clause instead.
LATERAL VIEW and LATERAL table functions, such as FROM t, LATERAL explode(t.arr), are not supported, so a generator cannot read a column of another table in the same FROM clause.
IDENTIFIER('t') in a FROM clause names a table, as in Spark.

Overview

Performance and Observability

Security

Self-Managed

Managed service

Guides

SQL reference

Release notes

API reference

Legal

Limitations

​Limitations

Limitations