LengthOfJsonArray¶

Overview¶

The LengthOfJsonArray expression returns the number of elements in a JSON array string. It is implemented as a runtime-replaceable expression that delegates to the JsonExpressionUtils utility class for actual JSON parsing and length calculation.

Syntax¶

json_array_length(json_string)

// DataFrame API
import org.apache.spark.sql.functions.expr
df.select(expr("json_array_length(json_column)"))

Arguments¶

Argument	Type	Description
json_string	String (with collation support)	A JSON string representing an array whose length is to be calculated

Return Type¶

Returns IntegerType representing the number of elements in the JSON array.

Supported Data Types¶

String types with collation support (specifically supports trim collation)
Input must be a valid JSON array string format

Algorithm¶

Accepts a string expression containing JSON array data
Delegates actual processing to JsonExpressionUtils.lengthOfJsonArray() method via StaticInvoke
Parses the JSON string to identify array structure and count elements
Returns integer count of array elements
Handles malformed JSON according to JsonExpressionUtils implementation

Partitioning Behavior¶

Preserves partitioning as it operates on individual rows without requiring data movement
Does not require shuffle operations
Can be pushed down to individual partitions for parallel execution

Edge Cases¶

Returns NULL when input is NULL (expression is nullable)
Behavior with malformed JSON depends on JsonExpressionUtils implementation
Empty JSON array [] should return 0
Non-array JSON values (objects, primitives) handling depends on underlying utility
Invalid JSON strings may return NULL or throw exceptions based on implementation

Code Generation¶

This expression uses RuntimeReplaceable pattern and delegates to StaticInvoke, which supports whole-stage code generation. The actual code generation is handled by the StaticInvoke mechanism calling the static method in JsonExpressionUtils.

Examples¶

-- Basic usage
SELECT json_array_length('[1, 2, 3, 4]') AS length;
-- Returns: 4

-- Empty array
SELECT json_array_length('[]') AS length;  
-- Returns: 0

-- NULL input
SELECT json_array_length(NULL) AS length;
-- Returns: NULL

-- With table data
SELECT id, json_array_length(json_data) AS array_size 
FROM table_with_json 
WHERE json_data IS NOT NULL;

// DataFrame API usage
import org.apache.spark.sql.functions._

val df = spark.createDataFrame(Seq(
  ("1", "[1, 2, 3]"),
  ("2", "[]"),
  ("3", null)
)).toDF("id", "json_array")

df.select(
  col("id"),
  expr("json_array_length(json_array)").as("length")
).show()