#!/bin/bash
set -e

echo "🚀 Starting RAG Pipeline Optimizer..."
echo "================================================"

# Display Python and Streamlit versions
echo "📦 Environment Information:"
python --version
streamlit --version
echo "================================================"

# Check if running on HuggingFace Spaces
if [ -n "$SPACE_ID" ]; then
    echo "🤗 Running on HuggingFace Spaces: $SPACE_ID"
else
    echo "💻 Running in local environment"
fi
echo "================================================"

# Download ChromaDB data if not present
echo "📥 Checking ChromaDB data..."

if [ -f "scripts/download_data.py" ]; then
    python scripts/download_data.py
    
    if [ $? -ne 0 ]; then
        echo "❌ Failed to download ChromaDB data"
        exit 1
    fi
else
    echo "⚠️  download_data.py not found, skipping data download"
    echo "Assuming data is already present in the repository"
fi

echo "================================================"

# Check environment variables
echo "🔍 Checking environment variables..."

# Azure OpenAI
if [ -n "$AZURE_OPENAI_API_KEY" ]; then
    echo "✅ AZURE_OPENAI_API_KEY is set"
else
    echo "⚠️  WARNING: AZURE_OPENAI_API_KEY not set!"
fi

if [ -n "$AZURE_OPENAI_ENDPOINT" ]; then
    echo "✅ AZURE_OPENAI_ENDPOINT is set"
else
    echo "⚠️  WARNING: AZURE_OPENAI_ENDPOINT not set!"
fi

# Azure Embeddings
if [ -n "$AZURE_OPENAI_EMBEDDING_API_KEY" ]; then
    echo "✅ AZURE_OPENAI_EMBEDDING_API_KEY is set"
else
    echo "⚠️  WARNING: AZURE_OPENAI_EMBEDDING_API_KEY not set!"
fi

# Cohere Rerank
if [ -n "$AZURE_COHERE_RERANK_KEY" ]; then
    echo "✅ AZURE_COHERE_RERANK_KEY is set"
else
    echo "⚠️  WARNING: AZURE_COHERE_RERANK_KEY not set!"
fi

# Anthropic
if [ -n "$ANTHROPIC_API_KEY" ]; then
    echo "✅ ANTHROPIC_API_KEY is set"
else
    echo "⚠️  ANTHROPIC_API_KEY not set (optional)"
fi

echo "================================================"

# Check if ChromaDB data exists
echo "🔍 Verifying ChromaDB collections..."

PIPELINES=("a" "b" "c" "d" "e" "f")
DATA_DIR="data/vector_stores"

if [ -d "$DATA_DIR" ]; then
    FOUND_COUNT=0
    for pipeline in "${PIPELINES[@]}"; do
        CORPUS_PATH="$DATA_DIR/pipeline_${pipeline}_corpus"
        if [ -d "$CORPUS_PATH" ]; then
            echo "  ✓ Pipeline ${pipeline^^} corpus found"
            FOUND_COUNT=$((FOUND_COUNT + 1))
        else
            echo "  ✗ Pipeline ${pipeline^^} corpus MISSING"
        fi
    done
    
    echo "Found $FOUND_COUNT out of 6 pipeline collections"
    
    if [ $FOUND_COUNT -eq 0 ]; then
        echo "⚠️  WARNING: No ChromaDB collections found!"
        echo "The app may not function correctly without vector stores."
    fi
else
    echo "⚠️  WARNING: data/vector_stores directory not found!"
fi

echo "================================================"

# Check database
echo "🔍 Checking evaluation database..."
if [ -f "data/evaluation_results.db" ]; then
    DB_SIZE=$(du -h data/evaluation_results.db | cut -f1)
    echo "✅ Evaluation database found (size: $DB_SIZE)"
else
    echo "⚠️  evaluation_results.db not found - will be created on first run"
fi

echo "================================================"
echo "🌐 Starting Streamlit on port 8501..."
echo "================================================"
echo ""
echo "🎯 Access the dashboard at:"
echo "   Local: http://localhost:8501"
if [ -n "$SPACE_ID" ]; then
    echo "   Public: https://huggingface.co/spaces/$SPACE_ID"
fi
echo ""
echo "================================================"

# Start Streamlit with proper configuration
exec streamlit run app/dashboard.py \
    --server.port=8501 \
    --server.address=0.0.0.0 \
    --server.headless=true \
    --browser.gatherUsageStats=false \
    --server.enableCORS=false \
    --server.enableXsrfProtection=false \
    --theme.base=dark \
    --logger.level=info