webshop-scraper/scripts/install.sh

#!/bin/bash

set -e

echo "==================================="
echo "Webshop 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"

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

# Check if application is already installed
EXISTING_INSTALLATION=false
EXISTING_ENV_FILE="$INSTALL_DIR/.env"

if [ -f "$EXISTING_ENV_FILE" ]; then
    EXISTING_INSTALLATION=true
    echo "Existing installation detected!"
    echo "Found existing configuration at: $EXISTING_ENV_FILE"
    echo ""

    # Load existing configuration
    source "$EXISTING_ENV_FILE" 2>/dev/null || true

    # Set defaults from existing config or fallback to hardcoded defaults
    EXISTING_API_KEY="$API_KEY"
    EXISTING_HOST="${HOST:-0.0.0.0}"
    EXISTING_PORT="${PORT:-3000}"
    EXISTING_MAX_CONCURRENT_JOBS="${MAX_CONCURRENT_JOBS:-3}"

    echo "Current configuration:"
    echo "  API Key: ${EXISTING_API_KEY:0:8}... (masked)"
    echo "  Host: $EXISTING_HOST"
    echo "  Port: $EXISTING_PORT"
    echo "  Max Concurrent Jobs: $EXISTING_MAX_CONCURRENT_JOBS"
    echo ""

    # Ask if user wants to keep existing configuration
    echo "Do you want to keep the existing configuration? [Y/n]:"
    read KEEP_CONFIG
    KEEP_CONFIG=${KEEP_CONFIG:-Y}

    if [[ "$KEEP_CONFIG" =~ ^[Yy]([Ee][Ss])?$ ]]; then
        echo "Using existing configuration..."
        API_KEY="$EXISTING_API_KEY"
        HOST="$EXISTING_HOST"
        PORT="$EXISTING_PORT"
        MAX_CONCURRENT_JOBS="$EXISTING_MAX_CONCURRENT_JOBS"
        SKIP_CONFIG_PROMPTS=true
    else
        echo "Will prompt for new configuration..."
        SKIP_CONFIG_PROMPTS=false
    fi
    echo ""
else
    echo "No existing installation found. Starting fresh installation..."
    echo ""
    SKIP_CONFIG_PROMPTS=false
fi

# Configuration prompts (skip if using existing config)
if [ "$SKIP_CONFIG_PROMPTS" != "true" ]; then
    # Prompt for API key
    if [ "$EXISTING_INSTALLATION" = "true" ]; then
        echo "Enter new API key for authentication (current: ${EXISTING_API_KEY:0:8}...):"
        echo "Press Enter to keep current API key, or enter new one:"
        read -s NEW_API_KEY
        if [ -n "$NEW_API_KEY" ]; then
            API_KEY="$NEW_API_KEY"
        else
            API_KEY="$EXISTING_API_KEY"
        fi
    else
        echo "Please enter the API key for authentication:"
        read -s API_KEY
    fi
    echo ""

    if [ -z "$API_KEY" ]; then
        echo "Error: API key cannot be empty"
        exit 1
    fi

    # Prompt for host/IP address
    echo "Enter the host/IP address to listen on:"
    echo "  - 0.0.0.0 for all interfaces (accessible from anywhere)"
    echo "  - 127.0.0.1 for localhost only (accessible only from this machine)"
    echo "  - or enter a specific IP address"
    if [ "$EXISTING_INSTALLATION" = "true" ]; then
        echo "Current: $EXISTING_HOST, Default: $EXISTING_HOST"
        read HOST
        HOST=${HOST:-$EXISTING_HOST}
    else
        echo "Default: 0.0.0.0"
        read HOST
        HOST=${HOST:-0.0.0.0}
    fi

    # Prompt for port (optional)
    if [ "$EXISTING_INSTALLATION" = "true" ]; then
        echo "Enter the HTTP port (current: $EXISTING_PORT):"
        read PORT
        PORT=${PORT:-$EXISTING_PORT}
    else
        echo "Enter the HTTP port (default: 3000):"
        read PORT
        PORT=${PORT:-3000}
    fi

    # Prompt for max concurrent jobs (optional)
    if [ "$EXISTING_INSTALLATION" = "true" ]; then
        echo "Enter the maximum number of concurrent jobs (current: $EXISTING_MAX_CONCURRENT_JOBS):"
        read MAX_CONCURRENT_JOBS
        MAX_CONCURRENT_JOBS=${MAX_CONCURRENT_JOBS:-$EXISTING_MAX_CONCURRENT_JOBS}
    else
        echo "Enter the maximum number of concurrent jobs (default: 3):"
        read MAX_CONCURRENT_JOBS
        MAX_CONCURRENT_JOBS=${MAX_CONCURRENT_JOBS:-3}
    fi
fi

echo ""
echo "Configuration:"
echo "  Host: $HOST"
echo "  Port: $PORT"
echo "  Max Concurrent Jobs: $MAX_CONCURRENT_JOBS"
echo ""

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

# Get the script directory (where install.sh is located)
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
PROJECT_DIR="$(dirname "$SCRIPT_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

# Create environment file
echo "Creating environment file..."

# Check if source .env file exists
SOURCE_ENV_FILE="$PROJECT_DIR/.env"

if [ -f "$SOURCE_ENV_FILE" ]; then
    echo "Found .env file in source directory, using it..."
    # Copy the entire .env file from source
    cp "$SOURCE_ENV_FILE" "$INSTALL_DIR/.env"
    echo ".env file copied from source directory"
else
    # Fallback to creating from prompted values
    cat > "$INSTALL_DIR/.env" <<EOF
API_KEY=$API_KEY
HOST=$HOST
PORT=$PORT
MAX_CONCURRENT_JOBS=$MAX_CONCURRENT_JOBS
EOF
    echo ".env file created with basic configuration"
    echo ""
    echo "NOTE: For Qdrant and OpenRouter integration, please add these variables to $INSTALL_DIR/.env:"
    echo "  QDRANT_ENABLED=true"
    echo "  QDRANT_API_URL=http://your-qdrant-server:6333"
    echo "  QDRANT_API_KEY=your-qdrant-api-key"
    echo "  OPENROUTER_API_KEY=your-openrouter-api-key"
    echo "  MCP_ENABLED=true"
    echo "  MCP_PORT=3001"
    echo ""
fi

# 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=Webshop 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 based on existing installation
if [ "$EXISTING_INSTALLATION" = "true" ]; then
    # Check if service is already running
    if systemctl is-active --quiet "$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
else
    # Fresh installation - start service
    echo "Starting service..."
    systemctl start "$SERVICE_NAME"
fi

# Check status
echo ""
echo "==================================="
if [ "$EXISTING_INSTALLATION" = "true" ]; then
    echo "Update completed successfully!"
else
    echo "Installation completed successfully!"
fi
echo "==================================="
echo ""
echo "Service status:"
systemctl status "$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 $SERVICE_NAME"
echo ""
# Display appropriate URL based on host configuration
if [ "$HOST" = "0.0.0.0" ]; then
    DISPLAY_HOST="localhost"
else
    DISPLAY_HOST="$HOST"
fi

echo "API is available at: http://$DISPLAY_HOST:$PORT"
echo "Health check: http://$DISPLAY_HOST:$PORT/health"
echo ""