webshop-scraper/scripts/install.sh

#!/bin/bash

set -e

echo "==================================="
echo "Web Scraper Installation Script"
echo "==================================="
echo ""

# Check if running as root
if [ "$EUID" -ne 0 ]; then
    echo "Error: Please run as root (use sudo)"
    exit 1
fi

# Get the actual user who ran sudo
ACTUAL_USER=${SUDO_USER:-$USER}
ACTUAL_HOME=$(eval echo ~$ACTUAL_USER)

# Installation directory
INSTALL_DIR="/opt/webshop-scraper"
SERVICE_NAME="webshop-scraper"

# Get the script directory and project directory
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
PROJECT_DIR="$(dirname "$SCRIPT_DIR")"
SOURCE_ENV_FILE="$PROJECT_DIR/.env"

echo "Installation directory: $INSTALL_DIR"
echo "Service name: $SERVICE_NAME"
echo "Source directory: $PROJECT_DIR"
echo ""

# Check for source .env file first
if [ -f "$SOURCE_ENV_FILE" ]; then
    echo "✓ Found .env file in source directory: $SOURCE_ENV_FILE"
    echo "  Using configuration from source .env file"
    echo "  All environment variables will be automatically configured"
    echo ""

    # Load the source .env to show summary
    source "$SOURCE_ENV_FILE" 2>/dev/null || true

    echo "Configuration preview:"
    [ -n "$API_KEY" ] && echo "  ✓ API_KEY: ${API_KEY:0:8}... (configured)"
    [ -n "$HOST" ] && echo "  ✓ HOST: $HOST"
    [ -n "$PORT" ] && echo "  ✓ PORT: $PORT"
    [ -n "$QDRANT_ENABLED" ] && echo "  ✓ QDRANT_ENABLED: $QDRANT_ENABLED"
    [ -n "$QDRANT_API_URL" ] && echo "  ✓ QDRANT_API_URL: $QDRANT_API_URL"
    [ -n "$OPENROUTER_API_KEY" ] && echo "  ✓ OPENROUTER_API_KEY: ${OPENROUTER_API_KEY:0:15}... (configured)"
    [ -n "$MCP_ENABLED" ] && echo "  ✓ MCP_ENABLED: $MCP_ENABLED"
    echo ""

    USE_SOURCE_ENV=true
    SKIP_CONFIG_PROMPTS=true
else
    echo "⚠ No .env file found in source directory"
    echo "  Expected location: $SOURCE_ENV_FILE"
    echo ""
    echo "Please create a .env file in the project directory before running the installer."
    echo "You can copy .env.example and configure it:"
    echo "  cp .env.example .env"
    echo "  nano .env"
    echo ""
    exit 1
fi

# Create installation directory
echo "Creating installation directory..."
mkdir -p "$INSTALL_DIR"

# Check if Node.js is installed
if ! command -v node &> /dev/null; then
    echo "Error: Node.js is not installed. Please install Node.js first."
    exit 1
fi

# Check if npm is installed
if ! command -v npm &> /dev/null; then
    echo "Error: npm is not installed. Please install npm first."
    exit 1
fi

# Build application in the source directory first
echo "Building application..."
cd "$PROJECT_DIR"

# Install all dependencies (including dev dependencies for build)
echo "Installing build dependencies..."
npm install

# Build the application
echo "Compiling TypeScript and building web assets..."
npm run build

# Check if build was successful
if [ ! -f "$PROJECT_DIR/dist/index.js" ]; then
    echo "Error: Build failed - dist/index.js not found"
    exit 1
fi

# Copy application files
echo "Copying application files..."
cp -r package.json "$INSTALL_DIR/"
cp -r dist "$INSTALL_DIR/"

# Change to installation directory
cd "$INSTALL_DIR"

# Install only production dependencies
echo "Installing production dependencies..."
npm install --production

# Copy environment file from source
echo "Copying environment file..."
cp "$SOURCE_ENV_FILE" "$INSTALL_DIR/.env"
echo "✓ Environment file copied from source to $INSTALL_DIR/.env"

# Set permissions
chmod 600 "$INSTALL_DIR/.env"
chown -R $ACTUAL_USER:$ACTUAL_USER "$INSTALL_DIR"

# Create systemd service file
echo "Creating systemd service..."
cat > "/etc/systemd/system/$SERVICE_NAME.service" <<EOF
[Unit]
Description=Web Scraper Service
After=network.target

[Service]
Type=simple
User=$ACTUAL_USER
WorkingDirectory=$INSTALL_DIR
EnvironmentFile=$INSTALL_DIR/.env
ExecStart=/usr/bin/node $INSTALL_DIR/dist/index.js
Restart=on-failure
RestartSec=10

# Logging
StandardOutput=journal
StandardError=journal
SyslogIdentifier=$SERVICE_NAME

# Security
NoNewPrivileges=true
PrivateTmp=true

[Install]
WantedBy=multi-user.target
EOF

# Reload systemd
echo "Reloading systemd..."
systemctl daemon-reload

# Enable service
echo "Enabling service..."
systemctl enable "$SERVICE_NAME"

# Handle service startup
if systemctl is-active --quiet "$SERVICE_NAME"; then
    echo "Service is currently running. Restarting with new configuration..."
    systemctl restart "$SERVICE_NAME"
else
    echo "Starting service..."
    systemctl start "$SERVICE_NAME"
fi

# Check status
echo ""
echo "==================================="
echo "Installation completed successfully!"
echo "==================================="
echo ""
echo "Service status:"
systemctl status "$SERVICE_NAME" --no-pager || true
echo ""
echo "Useful commands:"
echo "  Start service:   sudo systemctl start $SERVICE_NAME"
echo "  Stop service:    sudo systemctl stop $SERVICE_NAME"
echo "  Restart service: sudo systemctl restart $SERVICE_NAME"
echo "  View logs:       sudo journalctl -u $SERVICE_NAME -f"
echo "  Check status:    sudo systemctl status $SERVICE_NAME"
echo ""

# Load .env to get actual values for display
source "$INSTALL_DIR/.env" 2>/dev/null || true

# Display appropriate URL based on host configuration
DISPLAY_HOST="${HOST:-localhost}"
if [ "$DISPLAY_HOST" = "0.0.0.0" ]; then
    DISPLAY_HOST="localhost"
fi

echo "Configuration loaded from: $SOURCE_ENV_FILE"
echo "API is available at: http://$DISPLAY_HOST:${PORT:-3000}"
echo "Health check: http://$DISPLAY_HOST:${PORT:-3000}/health"

# Show Qdrant status if enabled
if [ "$QDRANT_ENABLED" = "true" ]; then
    echo ""
    echo "Qdrant Integration: ENABLED"
    echo "  Qdrant URL: $QDRANT_API_URL"
    echo "  OpenRouter: ${OPENROUTER_API_KEY:+CONFIGURED}"
    echo "  MCP Server: ${MCP_ENABLED}"
fi
echo ""