Spark Query Language

Interactive SQL (Structured Query Language) queries are widely used for exploring and analyzing data in the current context by many business intelligence users. WSO2 DAS 3.0.0. ships with the feature of running SQL queries on the underlying datasources as specified in the DAS Data Access Layer (DAL).

It uses Spark SQL as the query engine, which succeeds Apache Hive from WSO2 DAS 3.0.0 onwards. This provides a powerful integration with the rest of the Spark analytics engine. For more information on Spark SQL, see Spark SQL Programming Guide.

Spark SQL queries

Spark SQL follows the standard SQL format. For information on the syntax explanations of the standard SQL format, see SQL Syntax.

Some query types of the standard SQL format are not supported by Spark SQL.

The query types that are supported by the Spark SQL parser are yet to appear in the published docs by the Apache Spark project. For more information on the SparkSQL query syntax, see the SparkSQL parser code, and the SQL Query test suite.

WSO2 DAS SQL guide

WSO2 DAS inherits the query parsing options from the Spark SQL’s native query parser. Click on the relevant tab to view the query formats to be used for the required action.

Reserved words in Spark SQL

The following are the reserved words in Spark SQL by default. These words cannot be used in Data Definition Language (DDL) tasks (e.g., as column names, etc).

The reserved words are case insensitive

ABS
ALL
AND
APPROXIMATE
AS
ASC
AVG
BETWEEN

BY
CASE
CAST
COALESCE
COUNT
DESC
DISTINCT
ELSE

END
EXCEPT
FALSE
FIRST
FROM
FULL
GROUP
HAVING

IF
IN
INNER
INSERT
INTERSECT
INTO
IS
JOIN

LAST
LEFT
LIKE
LIMIT
LOWER
MAX
MIN
NOT

NULL
ON
OR
ORDER
SORT
OUTER
OVERWRITE
REGEXP

RIGHT
RLIKE
SELECT
SEMI
SQRT
SUBSTR
SUBSTRING
SUM

TABLE
THEN
TRUE
UNION
UPPER
WHEN
WHERE
WITH

Reserved words in the WSO2 Carbon environment

The following words are reserved in the WSO2 Carbon environment.

The reserved words are case sensitive

CarbonAnalytics
CarbonJDBC