refactor: rename benchmark_cherrypick to benchmark_di, update paths, pubspec, imports, and documentation

benchmark_di/lib/cli/runner.dart

@@ -0,0 +1,96 @@
+import 'dart:io';
+import 'dart:math';
+import 'package:benchmark_di/benchmarks/universal_chain_benchmark.dart';
+import 'package:benchmark_di/benchmarks/universal_chain_async_benchmark.dart';
+
+/// Holds the results for a single benchmark execution.
+class BenchmarkResult {
+  /// List of timings for each run (in microseconds).
+  final List<num> timings;
+  /// Difference in memory (RSS, in KB) after running.
+  final int memoryDiffKb;
+  /// Difference between peak RSS and initial RSS (in KB).
+  final int deltaPeakKb;
+  /// Peak RSS memory observed (in KB).
+  final int peakRssKb;
+  BenchmarkResult({
+    required this.timings,
+    required this.memoryDiffKb,
+    required this.deltaPeakKb,
+    required this.peakRssKb,
+  });
+  /// Computes a BenchmarkResult instance from run timings and memory data.
+  factory BenchmarkResult.collect({
+    required List<num> timings,
+    required List<int> rssValues,
+    required int memBefore,
+  }) {
+    final memAfter = ProcessInfo.currentRss;
+    final memDiffKB = ((memAfter - memBefore) / 1024).round();
+    final peakRss = [...rssValues, memBefore].reduce(max);
+    final deltaPeakKb = ((peakRss - memBefore) / 1024).round();
+    return BenchmarkResult(
+      timings: timings,
+      memoryDiffKb: memDiffKB,
+      deltaPeakKb: deltaPeakKb,
+      peakRssKb: (peakRss / 1024).round(),
+    );
+  }
+}
+
+/// Static methods to execute and time benchmarks for DI containers.
+class BenchmarkRunner {
+  /// Runs a synchronous benchmark ([UniversalChainBenchmark]) for a given number of [warmups] and [repeats].
+  /// Collects execution time and observed memory.
+  static Future<BenchmarkResult> runSync({
+    required UniversalChainBenchmark benchmark,
+    required int warmups,
+    required int repeats,
+  }) async {
+    final timings = <num>[];
+    final rssValues = <int>[];
+    for (int i = 0; i < warmups; i++) {
+      benchmark.setup();
+      benchmark.run();
+      benchmark.teardown();
+    }
+    final memBefore = ProcessInfo.currentRss;
+    for (int i = 0; i < repeats; i++) {
+      benchmark.setup();
+      final sw = Stopwatch()..start();
+      benchmark.run();
+      sw.stop();
+      timings.add(sw.elapsedMicroseconds);
+      rssValues.add(ProcessInfo.currentRss);
+      benchmark.teardown();
+    }
+    return BenchmarkResult.collect(timings: timings, rssValues: rssValues, memBefore: memBefore);
+  }
+
+  /// Runs an asynchronous benchmark ([UniversalChainAsyncBenchmark]) for a given number of [warmups] and [repeats].
+  /// Collects execution time and observed memory.
+  static Future<BenchmarkResult> runAsync({
+    required UniversalChainAsyncBenchmark benchmark,
+    required int warmups,
+    required int repeats,
+  }) async {
+    final timings = <num>[];
+    final rssValues = <int>[];
+    for (int i = 0; i < warmups; i++) {
+      await benchmark.setup();
+      await benchmark.run();
+      await benchmark.teardown();
+    }
+    final memBefore = ProcessInfo.currentRss;
+    for (int i = 0; i < repeats; i++) {
+      await benchmark.setup();
+      final sw = Stopwatch()..start();
+      await benchmark.run();
+      sw.stop();
+      timings.add(sw.elapsedMicroseconds);
+      rssValues.add(ProcessInfo.currentRss);
+      await benchmark.teardown();
+    }
+    return BenchmarkResult.collect(timings: timings, rssValues: rssValues, memBefore: memBefore);
+  }
+}