Adds support for SQL standard PERCENTILE_CONT and PERCENTILE_DISC aggregate functions, leveraging Calcite's existing parser support.

Implementation

Rust sqllib additions:

• percentile_collect<P, V>() - Accumulates values into Tup2<Option<P>, Vec<V>> where P is percentile, V is data type
• PercentileSemigroup - Combines accumulators by taking first non-null percentile and concatenating vectors

SQL compiler extensions:

• NonLinearAggregate pattern in AggregateCompiler for percentile functions
• Post-processing calls appropriate percentile_cont/percentile_disc functions with nullable variants
• Type casting chain: SQL DECIMAL → F64 wrapper → primitive f64 via into_inner()

Example

SELECT 
  PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY salary) AS median,
  PERCENTILE_DISC(0.9) WITHIN GROUP (ORDER BY salary) AS p90
FROM employees;

The accumulator maintains a single percentile value across all groups while collecting all data points for final computation.

Testing

Java Unit Tests (AggregateTests.java):

• testPercentileCont - Validates PERCENTILE_CONT with mixed data types (integers, doubles)
• testPercentileDisc - Validates PERCENTILE_DISC with various percentiles (0.0, 0.5, 0.9, 1.0)
• testPercentileDescending - Tests descending order support with both functions

Python Runtime Tests (aggregate_tests6.py):

• test_percentile_cont_basic - Basic PERCENTILE_CONT functionality with integer data
• test_percentile_disc_basic - Basic PERCENTILE_DISC functionality with integer data
• test_percentile_multiple - Multiple percentile calculations in single query
• test_percentile_null_handling - NULL value handling in input data
• test_percentile_empty_group - Empty group behavior
• test_percentile_single_value - Single value edge case
• test_percentile_with_groupby - GROUP BY clause integration
• test_percentile_floating_point - Floating point data with PERCENTILE_CONT

All tests follow existing patterns in their respective test suites and validate correct incremental computation behavior.

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.