Update Dockerfile

Dockerfile

@@ -1,120 +1,177 @@
-# Use Ubuntu as base image for better Docker-in-Docker support
-FROM ubuntu:22.04
+# Dockerfile for Hugging Face Spaces - Non-root setup
+FROM node:18-slim
 
-# Avoid prompts from apt
-ENV DEBIAN_FRONTEND=noninteractive
+# Create non-root user
+RUN groupadd -r firecrawl && useradd -r -g firecrawl firecrawl
 
-# Install system dependencies
+# Install system dependencies as root
 RUN apt-get update && apt-get install -y \
-    curl \
     git \
-    sudo \
+    curl \
+    wget \
+    bash \
     ca-certificates \
-    gnupg \
-    lsb-release \
-    && rm -rf /var/lib/apt/lists/*
-
-# Install Docker
-RUN curl -fsSL https://download.docker.com/linux/ubuntu/gpg | gpg --dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg \
-    && echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable" | tee /etc/apt/sources.list.d/docker.list > /dev/null \
-    && apt-get update \
-    && apt-get install -y docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin \
+    redis-server \
+    postgresql \
+    postgresql-contrib \
+    sudo \
     && rm -rf /var/lib/apt/lists/*
 
-# Create non-root user
-RUN useradd -m -s /bin/bash -u 1001 firecrawl \
-    && usermod -aG docker firecrawl \
-    && echo 'firecrawl ALL=(ALL) NOPASSWD: /usr/bin/dockerd' >> /etc/sudoers
-
-# Set working directory
-WORKDIR /home/firecrawl
+# Create app directory and set permissions
+RUN mkdir -p /home/firecrawl/app && \
+    chown -R firecrawl:firecrawl /home/firecrawl
 
-# Switch to non-root user for cloning
+# Switch to non-root user
 USER firecrawl
+WORKDIR /home/firecrawl/app
 
 # Clone Firecrawl repository
-RUN git clone https://github.com/mendableai/firecrawl.git /home/firecrawl/firecrawl
-
-# Set working directory to firecrawl
-WORKDIR /home/firecrawl/firecrawl
-
-# Copy .env.example to .env in the API directory
-RUN if [ -f apps/api/.env.example ]; then \
-        cp apps/api/.env.example apps/api/.env; \
-    fi
-
-# Create startup script
-USER root
-RUN echo '#!/bin/bash' > /start-firecrawl.sh && \
-    echo 'set -e' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'echo "🔥 Starting Firecrawl Self-Host Setup 🔥"' >> /start-firecrawl.sh && \
-    echo 'echo "==========================================="' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo '# Start Docker daemon in background if not running' >> /start-firecrawl.sh && \
-    echo 'if ! pgrep dockerd > /dev/null; then' >> /start-firecrawl.sh && \
-    echo '    echo "Starting Docker daemon..."' >> /start-firecrawl.sh && \
-    echo '    sudo dockerd --host=unix:///var/run/docker.sock --host=tcp://0.0.0.0:2376 &' >> /start-firecrawl.sh && \
-    echo '    sleep 5' >> /start-firecrawl.sh && \
-    echo 'fi' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo '# Wait for Docker to be ready' >> /start-firecrawl.sh && \
-    echo 'echo "Waiting for Docker to be ready..."' >> /start-firecrawl.sh && \
-    echo 'timeout=30' >> /start-firecrawl.sh && \
-    echo 'while [ $timeout -gt 0 ]; do' >> /start-firecrawl.sh && \
-    echo '    if docker info > /dev/null 2>&1; then' >> /start-firecrawl.sh && \
-    echo '        echo "Docker is ready!"' >> /start-firecrawl.sh && \
-    echo '        break' >> /start-firecrawl.sh && \
-    echo '    fi' >> /start-firecrawl.sh && \
-    echo '    sleep 1' >> /start-firecrawl.sh && \
-    echo '    timeout=$((timeout-1))' >> /start-firecrawl.sh && \
-    echo 'done' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'if [ $timeout -eq 0 ]; then' >> /start-firecrawl.sh && \
-    echo '    echo "Error: Docker failed to start within 30 seconds"' >> /start-firecrawl.sh && \
-    echo '    exit 1' >> /start-firecrawl.sh && \
-    echo 'fi' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'cd /home/firecrawl/firecrawl' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'echo "Prerequisites check:"' >> /start-firecrawl.sh && \
-    echo 'echo "✓ Docker and Docker Compose are installed"' >> /start-firecrawl.sh && \
-    echo 'echo "✓ Repository cloned to /home/firecrawl/firecrawl"' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'if [ -f apps/api/.env ]; then' >> /start-firecrawl.sh && \
-    echo '    echo "✓ Environment file exists at apps/api/.env"' >> /start-firecrawl.sh && \
-    echo 'else' >> /start-firecrawl.sh && \
-    echo '    echo "⚠ No .env file found, using defaults"' >> /start-firecrawl.sh && \
-    echo 'fi' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'echo ""' >> /start-firecrawl.sh && \
-    echo 'echo "Building Firecrawl services..."' >> /start-firecrawl.sh && \
-    echo 'echo "Running: docker compose build"' >> /start-firecrawl.sh && \
-    echo 'docker compose build' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo 'echo ""' >> /start-firecrawl.sh && \
-    echo 'echo "Starting Firecrawl services..."' >> /start-firecrawl.sh && \
-    echo 'echo "Running: docker compose up"' >> /start-firecrawl.sh && \
-    echo 'echo ""' >> /start-firecrawl.sh && \
-    echo '' >> /start-firecrawl.sh && \
-    echo '# Run docker compose' >> /start-firecrawl.sh && \
-    echo 'exec docker compose up "$@"' >> /start-firecrawl.sh
-
-RUN chmod +x /start-firecrawl.sh \
-    && chown firecrawl:firecrawl /start-firecrawl.sh
-
-# Switch back to non-root user
-USER firecrawl
-
-# Expose common ports used by Firecrawl
-EXPOSE 3002 8080 6379 5432
+RUN git clone https://github.com/mendableai/firecrawl.git .
+
+# Install Node.js dependencies for API
+WORKDIR /home/firecrawl/app/apps/api
+RUN npm ci --only=production
+
+# Install Node.js dependencies for Playwright service (lightweight version)
+WORKDIR /home/firecrawl/app/apps/playwright-service-ts
+RUN npm ci --only=production
+
+# Build the applications
+WORKDIR /home/firecrawl/app/apps/api
+RUN npm run build
+
+WORKDIR /home/firecrawl/app/apps/playwright-service-ts
+RUN npm run build
+
+# Create directories for runtime
+WORKDIR /home/firecrawl/app
+RUN mkdir -p logs tmp data redis-data postgres-data
+
+# Copy environment template and create .env
+RUN cp apps/api/.env.example .env
+
+# Create a simple startup script for HF Spaces
+RUN cat > start.sh << 'EOF' && chmod +x start.sh
+#!/bin/bash
+
+# Set environment variables for HF Spaces
+export PORT=${PORT:-7860}
+export HOST=0.0.0.0
+export REDIS_URL=redis://localhost:6379
+export PLAYWRIGHT_MICROSERVICE_URL=http://localhost:3000/scrape
+export USE_DB_AUTHENTICATION=false
+export ENV=production
+export LOGGING_LEVEL=info
+
+# Start Redis in background (if we can)
+redis-server --daemonize yes --port 6379 --bind 127.0.0.1 --dir /home/firecrawl/app/redis-data || echo "Redis start failed, continuing..."
+
+# Start Playwright service in background
+cd /home/firecrawl/app/apps/playwright-service-ts
+npm start &
+PLAYWRIGHT_PID=$!
+
+# Wait a bit for Playwright to start
+sleep 5
+
+# Start API service
+cd /home/firecrawl/app/apps/api
+exec node dist/src/harness.js --start-docker
+EOF
+
+# Create a simple health check script
+RUN cat > health.sh << 'EOF' && chmod +x health.sh
+#!/bin/bash
+curl -f http://localhost:${PORT:-7860}/health 2>/dev/null || exit 1
+EOF
+
+# Create minimal configuration files
+RUN cat > /home/firecrawl/app/.env << 'EOF'
+# HF Spaces Configuration
+PORT=7860
+HOST=0.0.0.0
+ENV=production
+LOGGING_LEVEL=info
+
+# Disable database authentication for simplicity
+USE_DB_AUTHENTICATION=false
+
+# Redis (will try to use local instance)
+REDIS_URL=redis://localhost:6379
+REDIS_RATE_LIMIT_URL=redis://localhost:6379
+
+# Playwright service
+PLAYWRIGHT_MICROSERVICE_URL=http://localhost:3000/scrape
+BLOCK_MEDIA=true
+
+# Disable optional services that might not work in HF Spaces
+BULL_AUTH_KEY=fc-demo
+TEST_API_KEY=fc-demo-key
+
+# Model configuration (can be overridden with HF Spaces secrets)
+MODEL_NAME=gpt-3.5-turbo
+MODEL_EMBEDDING_NAME=text-embedding-ada-002
+EOF
+
+# Create a simple API-only version for HF Spaces constraints
+RUN cat > start-api-only.sh << 'EOF' && chmod +x start-api-only.sh
+#!/bin/bash
+
+# Simplified startup for HF Spaces - API only
+export PORT=${PORT:-7860}
+export HOST=0.0.0.0
+export ENV=production
+export LOGGING_LEVEL=info
+export USE_DB_AUTHENTICATION=false
+
+# Disable services that might not work
+export PLAYWRIGHT_MICROSERVICE_URL=http://localhost:3000/scrape
+export REDIS_URL=redis://localhost:6379
+
+echo "🔥 Starting Firecrawl API on port $PORT"
+echo "⚠️  Note: This is a simplified version for Hugging Face Spaces"
+echo "📝 Some features may be limited due to platform constraints"
+
+cd /home/firecrawl/app/apps/api
+
+# Create a simple in-memory fallback if Redis isn't available
+cat > dist/src/lib/redis-fallback.js << 'JSEOF'
+// Simple in-memory fallback for Redis when not available
+class MemoryStore {
+  constructor() {
+    this.store = new Map();
+  }
+  
+  async get(key) {
+    return this.store.get(key);
+  }
+  
+  async set(key, value, options = {}) {
+    this.store.set(key, value);
+    if (options.EX) {
+      setTimeout(() => this.store.delete(key), options.EX * 1000);
+    }
+    return 'OK';
+  }
+  
+  async del(key) {
+    return this.store.delete(key);
+  }
+}
+
+module.exports = { MemoryStore };
+JSEOF
+
+# Start the API service
+exec node dist/src/harness.js --start-docker
+EOF
+
+# Expose the port that HF Spaces expects
+EXPOSE 7860
 
 # Health check
 HEALTHCHECK --interval=30s --timeout=10s --start-period=60s --retries=3 \
-    CMD docker compose ps -q | xargs docker inspect --format='{{.State.Health.Status}}' | grep -v healthy && exit 1 || exit 0
-
-# Set entrypoint
-ENTRYPOINT ["/start-firecrawl.sh"]
+    CMD ./health.sh
 
-# Default command (can be overridden)
-CMD ["./start-firecrawl.sh"]
+# Default command - use the API-only version for HF Spaces
+CMD ["./start-api-only.sh"]