Merge pull request #2 from WenyinWei/copilot/fix-1

Enhance parallelism capability with modern C++ features

Author: WenyinWei

.gitignore

@@ -40,6 +40,12 @@
 # debug information files
 *.dwo
 
-# Build directories
-.xmake/
-build/
+# Build directories
+.xmake/
+build/
+
+# Example executables
+simple_standard_parallelism
+standard_parallelism_demo
+test_standard_parallelism
+parallelism_integration_demo

PARALLELISM_USAGE.md

@@ -0,0 +1,85 @@
+# Simplified Parallelism Usage in diffeq
+
+The diffeq library now provides a much simpler parallelism interface alongside the advanced facade system.
+
+## Quick Start (Simple Interface)
+
+For most users, parallel execution is now as simple as:
+
+```cpp
+#include <diffeq.hpp>
+
+// Parallel execution of ODE integration for multiple initial conditions
+std::vector<std::vector<double>> initial_conditions = /* your data */;
+
+diffeq::execution::parallel_for_each(initial_conditions, [](std::vector<double>& state) {
+    auto integrator = diffeq::RK4Integrator<std::vector<double>>(your_system);
+    integrator.step(state, dt);
+});
+```
+
+## Simple API Reference
+
+### Free Functions (Global)
+```cpp
+// Execute function on each element in parallel
+diffeq::execution::parallel_for_each(container, lambda);
+
+// Async execution with future
+auto future = diffeq::execution::parallel_async(lambda);
+
+// Configure global parallel settings
+diffeq::execution::set_parallel_workers(8);
+diffeq::execution::enable_gpu_acceleration();
+diffeq::execution::enable_cpu_only();
+```
+
+### Parallel Class (Custom Instances)
+```cpp
+// Create custom parallel executor
+auto parallel = diffeq::execution::Parallel(4); // 4 workers
+
+// Execute on each element
+parallel.for_each(container, lambda);
+
+// Async execution
+auto future = parallel.async(lambda);
+
+// Configuration
+parallel.set_workers(8);
+parallel.use_gpu();
+parallel.use_cpu();
+
+// Information
+size_t workers = parallel.worker_count();
+bool gpu_available = parallel.gpu_available();
+```
+
+## Advanced Usage
+
+For complex scenarios requiring fine-grained control, real-time priorities, hardware-specific optimizations, etc., the full ParallelismFacade and builder pattern are still available:
+
+```cpp
+auto facade = diffeq::execution::parallel_execution()
+    .target_gpu()
+    .use_thread_pool()
+    .workers(1024)
+    .realtime_priority()
+    .enable_load_balancing()
+    .build();
+```
+
+## Migration Guide
+
+**Before (complex):**
+```cpp
+diffeq::execution::ParallelismFacade facade;
+facade.parallel_for_each(container, lambda);
+```
+
+**Now (simple):**
+```cpp
+diffeq::execution::parallel_for_each(container, lambda);
+```
+
+The simplified interface makes parallelism accessible to all users while maintaining full backward compatibility.