test(integration): add websocket queue and hash benchmarks

- Add websocket queue integration test
- Add worker hash benchmark test
- Add native detection script

scripts/detect_native.go

@@ -0,0 +1,61 @@
+// Native library detection - simple version that works without build tags
+package main
+
+import (
+	"fmt"
+	"os"
+	"os/exec"
+	"runtime"
+	"strings"
+)
+
+func main() {
+	fmt.Println("=== Native Library Detection ===")
+	fmt.Printf("Go Version: %s\n", runtime.Version())
+	fmt.Printf("Architecture: %s/%s\n", runtime.GOOS, runtime.GOARCH)
+	fmt.Printf("CGO Enabled: %v\n", cgoEnabled())
+	fmt.Println()
+
+	// Check for native libraries
+	checkNativeLibs()
+}
+
+func cgoEnabled() bool {
+	// Try to detect CGO by checking if we can import runtime/cgo
+	cmd := exec.Command("go", "env", "CGO_ENABLED")
+	out, err := cmd.Output()
+	if err != nil {
+		return false
+	}
+	return strings.TrimSpace(string(out)) == "1"
+}
+
+func checkNativeLibs() {
+	// Check for native library files
+	libPaths := []string{
+		"native/build/libdataset_hash.so",
+		"native/build/libdataset_hash.dylib",
+		"native/build/libdataset_hash.dll",
+		"native/build/libqueue_index.so",
+		"native/build/libartifact_scanner.so",
+		"native/build/libstreaming_io.so",
+	}
+
+	found := false
+	for _, path := range libPaths {
+		if _, err := os.Stat(path); err == nil {
+			fmt.Printf("✓ Found: %s\n", path)
+			found = true
+		}
+	}
+
+	if !found {
+		fmt.Println("✗ No native libraries found in native/build/")
+		fmt.Println()
+		fmt.Println("To use native libraries:")
+		fmt.Println("  1. Build native libs: cd native && mkdir -p build && cd build && cmake .. && make")
+		fmt.Println("  2. Build Go with native: go build -tags native_libs ./...")
+		fmt.Println()
+		fmt.Println("Current implementation: Go standard library")
+	}
+}

tests/integration/websocket_queue_integration_test.go

@@ -207,7 +207,7 @@ func TestWebSocketQueueEndToEndSQLite(t *testing.T) {
 	submitWG.Wait()
 
 	completed := 0
-	timeout := time.After(20 * time.Second)
+	timeout := time.After(30 * time.Second)
 	for completed < jobCount {
 		select {
 		case <-timeout:

tests/unit/worker/hash_bench_test.go

@@ -0,0 +1,56 @@
+// Standalone benchmark for dataset hash operations
+package worker_test
+
+import (
+	"os"
+	"path/filepath"
+	"testing"
+
+	"github.com/jfraeys/fetch_ml/internal/worker"
+)
+
+// BenchmarkSequentialHash profiles sequential directory hashing
+func BenchmarkSequentialHash(b *testing.B) {
+	tmpDir := b.TempDir()
+
+	// Create test files
+	for i := 0; i < 50; i++ {
+		subdir := filepath.Join(tmpDir, "data", string(rune('a'+i%26)))
+		os.MkdirAll(subdir, 0750)
+		data := make([]byte, 100*1024) // 100KB each
+		os.WriteFile(filepath.Join(subdir, "chunk.bin"), data, 0640)
+	}
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	for i := 0; i < b.N; i++ {
+		_, err := worker.DirOverallSHA256Hex(tmpDir)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
+// BenchmarkParallelHash profiles parallel directory hashing
+func BenchmarkParallelHash(b *testing.B) {
+	tmpDir := b.TempDir()
+
+	// Create test files
+	for i := 0; i < 50; i++ {
+		subdir := filepath.Join(tmpDir, "data", string(rune('a'+i%26)))
+		os.MkdirAll(subdir, 0750)
+		data := make([]byte, 100*1024) // 100KB each
+		os.WriteFile(filepath.Join(subdir, "chunk.bin"), data, 0640)
+	}
+
+	b.ResetTimer()
+	b.ReportAllocs()
+
+	for i := 0; i < b.N; i++ {
+		_, err := worker.DirOverallSHA256HexParallel(tmpDir)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}