Accessing Iceberg Tables Created by Third-Party Engines

For an Iceberg table generated by Spark, DataArts Fabric SQL accesses Iceberg data on OBS through external tables.

1. Obtain the corresponding OBS path access permissions.

   Log in to the Huawei Cloud LakeFormation console. In the navigation pane on the left, choose Data Permissions > Data Authorization. On the displayed page, click Authorize. In the dialog box that appears, grant READ/WRITE permissions for the file path to the current IAM user.

2. Create an external table.

   1	CREATE EXTERNAL TABLE table_name [column_name type_name,...] location 'fullLocation' store as iceberg;

   Here, fullLocation is the complete path of the Iceberg table on OBS. If specifying column names and types, ensure they match the metadata at the file level. If unspecified, the table will be created using the column names and types from the Iceberg metadata at the file level. The mapping between ICEBERG file-level data types and DataArts Fabric data types, along with the required table creation types, is shown in the table below.

   Table 1 Mapping between DataArts Fabric data types and ICEBERG data types

   Type Name	ICEBERG Data Type	DataArts Fabric Table Creation Type
   2-byte integer	INTEGER	INTEGER
   4-byte integer	INTEGER	INTEGER
   8-byte integer	LONG	BIGINT
   Single-precision floating point	FLOAT	FLOAT4
   Double-precision floating point	DOUBLE	FLOAT8
   Scientific data type	DECIMAL	DECIMAL[p,s]
   Date	DATE	DATE
   Time	TIMESTAMP	TIMESTAMP
   Boolean	BOOLEAN	BOOLEAN
   Char	STRING	TEXT
   VARCHAR	STRING	TEXT
   String (large text object)	STRING	TEXT