Package com.github.rlacher.sortbench.benchmark

Class BenchmarkRunner

java.lang.Object

com.github.rlacher.sortbench.benchmark.BenchmarkRunner

public class BenchmarkRunner
extends Object

Executes and benchmarks sorting algorithms based on a provided configuration and returns raw benchmark results.

Each benchmark run focuses on a single profiling mode, though it may involve multiple sorting strategies, data sizes, and data types.

Constructor Summary

Constructors

Constructor	Description
BenchmarkRunner(Sorter sorter)	Constructs a BenchmarkRunner that delegates sorting to the provided Sorter.

Method Summary

Modifier and Type	Method	Description
protected Map<BenchmarkContext,List<BenchmarkData>>	generateBenchmarkData(Set<String> strategyNames, Set<BenchmarkData.DataType> dataTypes, List<Integer> sizes, int iterations)	Generates benchmark data grouped by context to facilitate batch processing.
protected SortStrategy	getStrategyInstance(Class<? extends SortStrategy> strategyClass, Benchmarker.ProfilingMode mode)	Creates an instance of a SortStrategy based on the provided strategy class and profiling mode.
List<BenchmarkResult>	run(Map<String,Class<? extends SortStrategy>> selectedStrategies, Set<BenchmarkData.DataType> dataTypes, List<Integer> inputSizes, int iterations, Benchmarker.ProfilingMode profilingMode)	Runs benchmarks based on the provided parameters.
protected List<BenchmarkResult>	runIterations(Map<String,SortStrategy> strategies, Map<BenchmarkContext,List<BenchmarkData>> benchmarkDataMap)	Runs the provided sorting strategies on the given benchmark data using the sorter.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

BenchmarkRunner

public BenchmarkRunner(Sorter sorter)

Constructs a BenchmarkRunner that delegates sorting to the provided Sorter.

Parameters:

sorter - The Sorter instance to use.

Throws:

IllegalArgumentException - If sorter is null.

Method Details

run

public List<BenchmarkResult> run(Map<String,Class<? extends SortStrategy>> selectedStrategies,
 Set<BenchmarkData.DataType> dataTypes,
 List<Integer> inputSizes,
 int iterations,
 Benchmarker.ProfilingMode profilingMode)

Runs benchmarks based on the provided parameters.

Parameters:

selectedStrategies - A mapping from strategy names (e.g., "InsertionSort") to their respective SortStrategy class types for dynamic instantiation.

dataTypes - A set of data types for which to generate benchmark data.

inputSizes - A list of input sizes for the benchmark data.

iterations - The number of iterations for each data arrangement.

profilingMode - The profiling mode to use for the benchmark.

Returns:

A list of benchmark results with size equal or greater than 1.

Throws:

IllegalArgumentException - If any of the input parameters are null or invalid:

• selectedStrategies is null, empty, contains null keys or values, or duplicate keys (ignoring case) or values.
• dataTypes is null or empty.
• inputSizes is null or empty or contains non-positive integers.
• iterations is not positive.
• profilingMode is null.

runIterations

protected List<BenchmarkResult> runIterations(Map<String,SortStrategy> strategies,
 Map<BenchmarkContext,List<BenchmarkData>> benchmarkDataMap)

Runs the provided sorting strategies on the given benchmark data using the sorter.

Parameters:

strategies - A map of sorting strategy names to instances.

benchmarkDataMap - The map of BenchmarkContext to lists of benchmark data.

Returns:

A list of benchmark results.

Throws:

IllegalArgumentException - If the strategies or benchmarkDataMap is null or empty.

IllegalStateException - If a required sorting strategy is not found for a benchmark context.

generateBenchmarkData

protected Map<BenchmarkContext,List<BenchmarkData>> generateBenchmarkData(Set<String> strategyNames,
 Set<BenchmarkData.DataType> dataTypes,
 List<Integer> sizes,
 int iterations)

Generates benchmark data grouped by context to facilitate batch processing.

This improves JVM optimisation by enhancing data locality. Deep copies data for each strategy to ensure result comparability.

Parameters:

strategyNames - A set of sorting strategy names.

dataTypes - A set of data types for which to generate data.

sizes - A list of integer data sizes for which to generate data.

iterations - The number of data arrangements per size and data type.

Returns:

A map of BenchmarkContext to a list of benchmark data. The map may be empty if any of the input parameters are empty.

Throws:

NullPointerException - If any of dataTypes, sizes or strategyNames are null. This is unexpected as the caller run(java.util.Map<java.lang.String, java.lang.Class<? extends com.github.rlacher.sortbench.strategies.SortStrategy>>, java.util.Set<com.github.rlacher.sortbench.benchmark.data.BenchmarkData.DataType>, java.util.List<java.lang.Integer>, int, com.github.rlacher.sortbench.benchmark.Benchmarker.ProfilingMode) ensures these arguments are not null.

getStrategyInstance

protected SortStrategy getStrategyInstance(Class<? extends SortStrategy> strategyClass,
 Benchmarker.ProfilingMode mode)

Creates an instance of a SortStrategy based on the provided strategy class and profiling mode.

Parameters:

strategyClass - The class type of the sorting strategy.

mode - The profiling mode to use.

Returns:

An SortStrategy instance of type strategyClass configured with the provided profiling mode.

Throws:

IllegalArgumentException - If strategyClass is null.

IllegalStateException - If instantiation fails for the SortStrategy.