SQLKeywords¶

Overview¶

The SQLKeywords expression is a generator function that produces a table of SQL keywords and their reservation status. It returns all SQL keywords supported by Spark SQL along with a boolean flag indicating whether each keyword is reserved or non-reserved.

Syntax¶

SELECT * FROM sql_keywords()

// DataFrame API usage
spark.sql("SELECT * FROM sql_keywords()")

Arguments¶

Argument	Type	Description
None	N/A	This expression takes no arguments

Return Type¶

Returns a table with the following schema:

keyword (StringType, non-nullable): The SQL keyword string
reserved (BooleanType, non-nullable): Whether the keyword is reserved (true) or non-reserved (false)

Supported Data Types¶

This expression does not accept input data types as it is a parameterless generator function that produces its own data.

Algorithm¶

Retrieves the complete list of SQL keywords from Spark's internal keyword registry
Obtains the reservation status for each keyword using getReservedList()
Combines keywords with their corresponding reservation status using zip operation
Generates InternalRow objects for each keyword-reservation pair
Returns an iterable collection of rows representing the complete keyword catalog

Partitioning Behavior¶

As a generator function:

Does not preserve input partitioning (generates new data)
The output is typically collected to a single partition since it's a metadata operation
No shuffle is required as this is a local data generation operation

Edge Cases¶

Always returns a consistent set of keywords regardless of session state
Keywords are returned as UTF8String objects to ensure proper encoding
Both keyword and reserved columns are marked as non-nullable in the schema
The function generates the same output deterministically across calls

Code Generation¶

This expression extends CodegenFallback, meaning it does not support Tungsten code generation and always falls back to interpreted evaluation mode for compatibility and simplicity.

Examples¶

-- Get all SQL keywords
SELECT * FROM sql_keywords();

-- Filter for reserved keywords only  
SELECT keyword FROM sql_keywords() WHERE reserved = true;

-- Count total keywords
SELECT COUNT(*) as total_keywords FROM sql_keywords();

-- Find non-reserved keywords containing 'ADD'
SELECT keyword FROM sql_keywords() 
WHERE reserved = false AND keyword LIKE '%ADD%';

// DataFrame API usage
import org.apache.spark.sql.SparkSession
val spark = SparkSession.builder().getOrCreate()

// Get all keywords
val keywords = spark.sql("SELECT * FROM sql_keywords()")
keywords.show()

// Filter and collect reserved keywords
val reserved = spark.sql("SELECT keyword FROM sql_keywords() WHERE reserved = true")
reserved.collect()