BL_FORMAT= Data Set Option

Specifies the format of the external or web table data.

Valid in: DATA and PROC steps (when accessing DBMS data using SAS/ACCESS software)
Categories: Bulk Loading
Data Set Control
Default: TEXT [Greenplum, Hadoop, HAWQ, Spark in HDFS, Yellowbrick]
CSV [PostgreSQL, Spark in Databricks]
Requirement: To specify this option, you must first specify BULKLOAD=YES.
Data source: Greenplum, Hadoop, HAWQ, PostgreSQL, Spark, Yellowbrick
Notes: Support for Yellowbrick was added in SAS 9.4M7.
Support for Hadoop was added in SAS 9.4M8.
Support for Spark was added in SAS 9.4M9.
The Parquet format was added in SAS 9.4M9.
See: BL_DELIMITER= data set option, BL_FORCE_NOT_NULL= data set option, BL_NULL= data set option, BL_QUOTE= data set option, BULKLOAD= data set option

Table of Contents

Syntax

BL_FORMAT=TEXT | CSV | ORC | PARQUET

Syntax Description

CSV

specifies a comma-separated value format.

ORC

specifies the Apache ORC (Optimized Row Columnar) open-source file format.

Restriction ORC format is supported only for Hadoop.

PARQUET

specifies the Apache Parquet open-source file format.

Note Parquet is supported only for Hadoop.

TEXT

specifies plain text format.

Details

In Hadoop and Spark, this option controls the format of the bulk load staging file. The Hadoop engine and the Spark engine (when bulk loading to HDFS) do not support other bulk loading data set options for the staging file.

Spark: This option is used when the Spark engine bulk loads or bulk unloads data in Databricks. For more information, see Bulk Loading and Unloading to Databricks in Azure.

Last updated: February 3, 2026