name: Xerv Crayon Production Build

# ============================================================================
# TRIGGER CONDITIONS
# ============================================================================
on:
  push:
    branches: [ "main", "dev" ]
  pull_request:
    branches: [ "main" ]

jobs:
  # ==========================================================================
  # JOB 1: INTEL/AMD CPU ENGINE (AVX2/AVX-512 Check)
  # ==========================================================================
  build-cpu:
    name: 🔵 Build CPU (Intel/AMD)
    runs-on: ubuntu-latest
    
    steps:
      - name: Checkout Repository
        uses: actions/checkout@v4
      
      - name: Set up Python 3.10
        uses: actions/setup-python@v5
        with:
          python-version: "3.10"
          
      - name: Install Dependencies
        run: |
          python -m pip install --upgrade pip
          pip install pytest setuptools wheel build
          
      - name: Compile Crayon (CPU Mode)
        run: |
          # This triggers setup.py to build CPU extensions
          pip install -v . --no-build-isolation
          
      - name: Verify CPU Extension
        run: |
          python -c "from crayon.c_ext import crayon_cpu; print('✅ CPU Engine Loaded')"
          python -c "from crayon.c_ext import crayon_cpu; print(f'Hardware: {crayon_cpu.get_hardware_info()}')"
          
      - name: Verify Trainer Extension
        run: |
          python -c "from crayon.c_ext import crayon_trainer; print('✅ Trainer Engine Loaded')"
          python -c "from crayon.c_ext import crayon_trainer; print(f'Version: {crayon_trainer.get_version()}')"
          python -c "from crayon.c_ext import crayon_trainer; print(f'Algorithm: {crayon_trainer.get_algorithm_info()}')"
          
      - name: Run Basic Tokenization Test
        run: |
          python -c "
          from crayon import CrayonVocab
          v = CrayonVocab(device='cpu')
          v.load_profile('lite') # LOAD PROFILE FIRST
          result = v.tokenize('Hello Cloud! Testing CRAYON on GitHub Actions.')
          print(f'✅ Tokenized to {len(result)} tokens')
          print(f'   Tokens: {result[:10]}...')
          "
          
      - name: Run Trainer Test
        run: |
          python -c "
          from crayon.c_ext import crayon_trainer
          
          # Test with minimal corpus
          corpus = b'The quick brown fox jumps over the lazy dog. ' * 100
          merges = crayon_trainer.train_fast(corpus, 300, min_freq=2, verbose=0)
          
          print(f'✅ Trainer generated {len(merges)} merge rules')
          print(f'   First 3 merges: {merges[:3]}')
          "
          
      - name: Run pytest (Unit Tests)
        run: |
          pytest tests/ -v --tb=short || true

  # ==========================================================================
  # JOB 2: NVIDIA CUDA ENGINE (Compilation Verification)
  # ==========================================================================
  build-cuda:
    name: 🟢 Build NVIDIA (CUDA 12)
    runs-on: ubuntu-latest
    
    # Use NVIDIA's official CUDA development container
    container: nvidia/cuda:12.2.0-devel-ubuntu22.04
    
    steps:
      - name: Checkout Repository
        uses: actions/checkout@v4
      
      - name: Install Python & Dependencies
        run: |
          apt-get update
          apt-get install -y python3 python3-pip python3-venv python3-dev git
          python3 -m pip install --upgrade pip setuptools wheel
          
      - name: Install PyTorch (CUDA)
        run: |
          # Install PyTorch with CUDA support for CUDAExtension
          pip install torch --index-url https://download.pytorch.org/whl/cu121
          
      - name: Compile Crayon (CUDA Mode)
        run: |
          # Force CUDA build
          export CRAYON_FORCE_CUDA=1
          pip install -v . --no-build-isolation
          
      - name: Verify CUDA Extension Built
        run: |
          # Check if the CUDA shared object was created
          find . -name "*crayon_cuda*.so" -o -name "*crayon_cuda*.pyd" | grep . && echo "✅ CUDA Binary Built!"
          
      - name: Verify CPU Extension (Sanity Check)
        run: |
          python3 -c "from crayon.c_ext import crayon_cpu; print('✅ CPU Engine Loaded')"
          
      - name: Verify Trainer Extension
        run: |
          python3 -c "from crayon.c_ext import crayon_trainer; print('✅ Trainer Engine Loaded')"

  # ==========================================================================
  # JOB 3: AMD ROCm ENGINE (Compilation Verification)
  # ==========================================================================
  build-rocm:
    name: 🔴 Build AMD (ROCm 6.0)
    runs-on: ubuntu-latest
    
    # Use AMD's official ROCm development container
    container: rocm/dev-ubuntu-22.04:6.0
    
    steps:
      - name: Checkout Repository
        uses: actions/checkout@v4
      
      - name: Install Python & Dependencies
        run: |
          apt-get update
          apt-get install -y python3 python3-pip python3-venv python3-dev git
          python3 -m pip install --upgrade pip setuptools wheel
          
      - name: Verify ROCm Installation
        run: |
          hipcc --version
          echo "ROCM_HOME=${ROCM_HOME:-/opt/rocm}"
          ls -la /opt/rocm/bin/ | head -20
          
      - name: Compile Crayon (ROCm Mode)
        run: |
          # Force ROCm build
          export CRAYON_FORCE_ROCM=1
          export ROCM_HOME=/opt/rocm
          pip install -v . --no-build-isolation
          
      - name: Verify ROCm Extension Built
        run: |
          # Check if the ROCm shared object was created
          find . -name "*crayon_rocm*.so" | grep . && echo "✅ ROCm Binary Built!"
          
      - name: Verify CPU Extension (Sanity Check)
        run: |
          python3 -c "from crayon.c_ext import crayon_cpu; print('✅ CPU Engine Loaded')"
          
      - name: Verify Trainer Extension
        run: |
          python3 -c "from crayon.c_ext import crayon_trainer; print('✅ Trainer Engine Loaded')"

  # ==========================================================================
  # JOB 4: WINDOWS CPU BUILD
  # ==========================================================================
  build-windows:
    name: 🪟 Build Windows (CPU)
    runs-on: windows-latest
    
    steps:
      - name: Checkout Repository
        uses: actions/checkout@v4
      
      - name: Set up Python 3.10
        uses: actions/setup-python@v5
        with:
          python-version: "3.10"
          
      - name: Install Dependencies
        run: |
          python -m pip install --upgrade pip
          pip install pytest setuptools wheel build
          
      - name: Compile Crayon (Windows CPU)
        run: |
          pip install -v . --no-build-isolation
          
      - name: Verify Extensions
        run: |
          python -c "from crayon.c_ext import crayon_cpu; print('✅ CPU Engine Loaded')"
          python -c "from crayon.c_ext import crayon_trainer; print('✅ Trainer Engine Loaded')"
          
      - name: Run Basic Test
        run: |
          python -c "from crayon import CrayonVocab; v = CrayonVocab(device='cpu'); v.load_profile('lite'); print(v.tokenize('Hello Windows!'))"

  # ==========================================================================
  # JOB 5: BENCHMARK (CPU Performance Validation)
  # ==========================================================================
  benchmark:
    name: 📊 Benchmark Performance
    runs-on: ubuntu-latest
    needs: [build-cpu]  # Only run after CPU build succeeds
    
    steps:
      - name: Checkout Repository
        uses: actions/checkout@v4
      
      - name: Set up Python 3.10
        uses: actions/setup-python@v5
        with:
          python-version: "3.10"
          
      - name: Install Crayon
        run: |
          pip install --upgrade pip setuptools wheel
          pip install -v . --no-build-isolation
          
      - name: Run Trainer Benchmark
        run: |
          python -c "
          import time
          from crayon.c_ext import crayon_trainer
          
          # Generate test corpus
          corpus = b'The quick brown fox jumps over the lazy dog. ' * 10000
          corpus_mb = len(corpus) / (1024 * 1024)
          
          print(f'Corpus Size: {corpus_mb:.2f} MB')
          
          # Warmup
          _ = crayon_trainer.train_fast(corpus[:10000], 300, verbose=0)
          
          # Benchmark
          start = time.perf_counter()
          merges = crayon_trainer.train_fast(corpus, 1000, verbose=1)
          elapsed = time.perf_counter() - start
          
          print(f'\\n=== BENCHMARK RESULTS ===')
          print(f'Merge Rules:   {len(merges):,}')
          print(f'Time:          {elapsed:.2f}s')
          print(f'Speed:         {corpus_mb / elapsed:.2f} MB/s')
          print(f'Merges/sec:    {len(merges) / elapsed:,.0f}')
          
          # Performance gate
          if elapsed > 30:
              print('⚠️  Warning: Training took longer than expected')
          else:
              print('✅ Performance acceptable')
          "
          
      - name: Run Tokenization Benchmark
        run: |
          python -c "
          import time
          from crayon import CrayonVocab
          
          v = CrayonVocab(device='cpu')
          v.load_profile('lite')
          
          # Generate test text
          text = 'The quick brown fox jumps over the lazy dog. ' * 10000
          text_mb = len(text.encode('utf-8')) / (1024 * 1024)
          
          # Warmup
          _ = v.tokenize(text[:1000])
          
          # Benchmark
          iterations = 5
          total_time = 0
          total_tokens = 0
          
          for _ in range(iterations):
              start = time.perf_counter()
              tokens = v.tokenize(text)
              elapsed = time.perf_counter() - start
              total_time += elapsed
              total_tokens += len(tokens)
          
          avg_time = total_time / iterations
          avg_tokens = total_tokens / iterations
          
          print(f'=== TOKENIZATION BENCHMARK ===')
          print(f'Text Size:     {text_mb:.2f} MB')
          print(f'Avg Tokens:    {avg_tokens:,.0f}')
          print(f'Avg Time:      {avg_time * 1000:.2f} ms')
          print(f'Tokens/sec:    {avg_tokens / avg_time:,.0f}')
          print(f'MB/sec:        {text_mb / avg_time:.2f}')
          print('✅ Benchmark complete')
          "