Dockerfile

# Dockerfile (place this in the root of your firecrawl project)

# 1. Base Image: Use a Node.js LTS version that includes build tools
FROM node:18

# 2. Environment Variables
ENV PNPM_HOME="/pnpm"
ENV PATH="$PNPM_HOME:$PATH"
# Prevent apt-get from asking questions
ENV DEBIAN_FRONTEND=noninteractive
# Set Node environment (can be overridden by supervisor conf or HF secrets)
ENV NODE_ENV=production

# 3. Install System Dependencies: Redis, Supervisor, Git, and utilities
RUN apt-get update && apt-get install -y --no-install-recommends \
    redis-server \
    supervisor \
    git \
    curl \
    wget \
    gnupg \
    # Clean up apt cache
    && rm -rf /var/lib/apt/lists/*

# 4. Install pnpm v9+ globally
RUN npm install -g pnpm@9

# 5. Set Application Directory
WORKDIR /usr/src/app

# 6. Copy Package Definitions & Install Dependencies (Leverages Docker cache)
# Copy root files first
COPY package.json pnpm-lock.yaml ./
# Copy workspace config if it exists (use wildcard *)
COPY pnpm-workspace.yaml* ./
# Copy the specific package.json for the api app
COPY apps/api/package.json ./apps/api/

# Install ALL monorepo dependencies using the lockfile (includes devDeps needed for playwright)
RUN pnpm install --frozen-lockfile

# 7. Install Playwright Browsers & Dependencies
# This command downloads browsers (e.g., Chromium) AND tries to install needed OS libraries.
# Run this using the 'api' package context, assuming playwright is its dependency.
# Specify the browser(s) you need (e.g., chromium). Check Firecrawl needs.
RUN pnpm --filter api exec playwright install --with-deps chromium

# 8. Copy Application Code
# Copy the rest of your Firecrawl project code into the image
COPY . .

# 9. Copy Supervisor Configuration
COPY supervisord.conf /etc/supervisor/conf.d/supervisord.conf

# 10. Configure Environment for Internal Communication (inside the container)
# --- !! UPDATED PORT !! ---
ENV PORT=7860 \
    HOST=0.0.0.0 \
    REDIS_URL=redis://localhost:6379 \
    REDIS_RATE_LIMIT_URL=redis://localhost:6379 \
    USE_DB_AUTHENTICATION=false \
    # Add any other required non-secret ENVs here
    LOGGING_LEVEL=INFO

# --- Configure Hugging Face Space specific settings ---
# --- !! UPDATED PORT !! ---
# Expose the port the internal application will listen on
EXPOSE 7860
# Health check endpoint (if Firecrawl has one, e.g., /test or /health)
# --- !! UPDATED PORT in HEALTHCHECK !! ---
# HEALTHCHECK --interval=30s --timeout=5s --start-period=30s --retries=3 \
#   CMD curl -f http://localhost:7860/test || exit 1
# (Uncomment and adjust HEALTHCHECK if you know the correct endpoint /test)

# 11. Start Supervisor
# This command starts supervisord, which in turn starts redis, the api, and the worker(s) based on supervisord.conf
CMD ["/usr/bin/supervisord", "-c", "/etc/supervisor/supervisord.conf"]