Update Dockerfile

Dockerfile

@@ -1,22 +1,31 @@
 # ===== Build stage =====
 FROM node:20-bullseye AS builder
+
+# Cần git để clone
 RUN apt-get update && apt-get install -y git && rm -rf /var/lib/apt/lists/*
+
+# pnpm
 RUN corepack enable && corepack prepare pnpm@9.0.0 --activate
 
 WORKDIR /app
-RUN git clone --depth=1 https://github.com/mendableai/firecrawl.git .
+
+# Clone mã nguồn Firecrawl (pin theo default branch; có thể đổi sang tag/commit bằng ARG)
+ARG FIRECRAWL_REPO=https://github.com/mendableai/firecrawl.git
+RUN git clone --depth=1 ${FIRECRAWL_REPO} .
+
+# Cài deps workspace và build 2 service
 RUN pnpm install --frozen-lockfile
 
-# Build services
 WORKDIR /app/apps/playwright-service
 RUN pnpm build
+
 WORKDIR /app/apps/api
 RUN pnpm build
 
-# ===== Runtime stage (HF Space) =====
+# ===== Runtime stage =====
 FROM node:20-bullseye
 
-# Chromium deps cho Playwright
+# Chromium/Playwright deps (tối thiểu an toàn)
 RUN apt-get update && apt-get install -y \
     wget gnupg ca-certificates fonts-liberation libasound2 \
     libatk1.0-0 libatk-bridge2.0-0 libatspi2.0-0 libcups2 libdbus-1-3 \
@@ -27,24 +36,21 @@ RUN apt-get update && apt-get install -y \
 
 ENV NODE_ENV=production
 WORKDIR /app
+
+# Copy toàn bộ repo đã build
 COPY --from=builder /app /app
 
-# Cài chromium & deps cho Playwright
+# Cài browser cho Playwright
 RUN npx playwright install chromium && npx playwright install-deps
 
-# Script chạy 2 tiến trình, port lấy từ $PORT do HF cung cấp
-# (HF gán $PORT=app_port, mặc định 7860)
+# Script chạy 2 tiến trình: playwright-service (3000) + api ($PORT)
 RUN printf '#!/usr/bin/env bash\nset -e\n\
 export PORT=${PORT:-7860}\n\
 export HOST=${HOST:-0.0.0.0}\n\
-# Bạn set các ENV khác trong Space Settings: REDIS_URL, USE_DB_AUTHENTICATION, v.v.\n\
-# Quan trọng: API đọc PLAYWRIGHT_MICROSERVICE_URL. Playwright service dưới đây mặc định nghe :3000\n\
+# Bạn tự set các ENV khác trong Space Settings: USE_DB_AUTHENTICATION, REDIS_URL, v.v.\n\
 export PLAYWRIGHT_MICROSERVICE_URL=${PLAYWRIGHT_MICROSERVICE_URL:-http://127.0.0.1:3000/html}\n\
-# Start playwright microservice (port 3000, internal)\n\
 node /app/apps/playwright-service/dist/index.js &\n\
-# Start API (nghe $PORT để HF router map)\n\
 node /app/apps/api/dist/main.js\n' > /app/run.sh && chmod +x /app/run.sh
 
-# Expose đúng với app_port (HF sẽ map tự động, nhưng để rõ ràng)
 EXPOSE 7860
 CMD ["/app/run.sh"]