Parallel Processing

Configuring Workers

The workers parameter controls how many files are processed concurrently:

// Auto-detect CPU cores (default)
$config = AnalyzerConfig::make()
    ->paths(['src', 'tests'])
    ->workers(0);  // 0 = auto-detect

// Specify exact worker count
$config->workers(2);   // 2 workers (smaller machines)
$config->workers(4);   // 4 workers (balanced)
$config->workers(8);   // 8 workers (large codebases)

Auto-Detection

When workers is set to 0 or 'auto', the analyzer automatically detects your CPU core count:

• macOS: Uses sysctl -n hw.ncpu
• Linux: Uses nproc
• Fallback: 4 cores if detection fails

use Cline\Analyzer\Actions\DetectCoreCount;

$cores = (new DetectCoreCount())();  // Returns detected core count

Command-Line Override

# Auto-detect cores
php artisan analyzer:analyze --workers=auto

# Specify worker count
php artisan analyzer:analyze --workers=8

# Use serial processing (1 worker)
php artisan analyzer:analyze --serial

Serial Processing

Use SerialProcessor for single-threaded execution:

use Cline\Analyzer\Processors\SerialProcessor;

$config = AnalyzerConfig::make()
    ->processor(new SerialProcessor());

This processes files sequentially, which can be useful for:

• Debugging issues
• Consistent output ordering
• Memory-constrained systems
• Small codebases (<50 files)

Performance Considerations

When to Use Parallel Processing

• Large codebases (100+ files)
• Modern multi-core systems
• Production CI/CD pipelines
• Regular development workflow

When to Use Single-Threaded

• Small codebases (<50 files)
• Debugging analysis issues
• Memory-limited environments
• Systems without process support

Benchmarks

Example performance on a 1000-file codebase:

Workers	Time
Single-threaded	~45 seconds
2 workers	~25 seconds
4 workers	~15 seconds
8 workers	~12 seconds

Memory Usage

Parallel processing uses more memory:

Workers	Memory
Single-threaded	~50MB baseline
4 workers	~200MB total
8 workers	~400MB total

Plan worker count based on available system memory.

Adaptive Configuration

// Adaptive worker count based on environment
$workers = match (true) {
    getenv('CI') === 'true' => 0,        // Auto-detect in CI
    PHP_OS_FAMILY === 'Windows' => 2,    // Conservative for Windows
    default => 0,                         // Auto-detect for development
};

$config = AnalyzerConfig::make()
    ->workers($workers);

Custom Core Detection

use Cline\Analyzer\Actions\DetectCoreCount;

// Get detected core count
$detector = new DetectCoreCount();
$cores = $detector();

// Use 50% of available cores
$workers = max(1, (int) ($cores / 2));

$config = AnalyzerConfig::make()
    ->workers($workers);