mirror of
https://github.com/Memo-2023/mana-monorepo.git
synced 2026-05-15 21:59:40 +02:00
NestJS-based web crawler service for structured content extraction. Features: - Depth-controlled crawling with URL pattern filtering - robots.txt compliance - HTML/PDF/Markdown content extraction - BullMQ job queue for async processing - Redis caching layer - Prometheus metrics
55 lines
1.5 KiB
Docker
55 lines
1.5 KiB
Docker
# Build stage
|
|
FROM node:20-alpine AS builder
|
|
|
|
RUN corepack enable && corepack prepare pnpm@9.15.0 --activate
|
|
|
|
WORKDIR /app
|
|
|
|
# Copy workspace files
|
|
COPY pnpm-workspace.yaml ./
|
|
COPY pnpm-lock.yaml ./
|
|
COPY package.json ./
|
|
|
|
# Copy service files
|
|
COPY services/mana-crawler/package.json ./services/mana-crawler/
|
|
|
|
# Copy shared packages
|
|
COPY packages/shared-drizzle-config/package.json ./packages/shared-drizzle-config/
|
|
|
|
# Install dependencies
|
|
RUN pnpm install --frozen-lockfile
|
|
|
|
# Copy source code
|
|
COPY services/mana-crawler ./services/mana-crawler
|
|
COPY packages/shared-drizzle-config ./packages/shared-drizzle-config
|
|
|
|
# Build
|
|
WORKDIR /app/services/mana-crawler
|
|
RUN pnpm build
|
|
|
|
# Production stage
|
|
FROM node:20-alpine AS runner
|
|
|
|
RUN corepack enable && corepack prepare pnpm@9.15.0 --activate
|
|
|
|
WORKDIR /app
|
|
|
|
# Copy package files
|
|
COPY --from=builder /app/pnpm-workspace.yaml ./
|
|
COPY --from=builder /app/pnpm-lock.yaml ./
|
|
COPY --from=builder /app/package.json ./
|
|
COPY --from=builder /app/services/mana-crawler/package.json ./services/mana-crawler/
|
|
COPY --from=builder /app/packages/shared-drizzle-config/package.json ./packages/shared-drizzle-config/
|
|
|
|
# Install production dependencies only
|
|
RUN pnpm install --frozen-lockfile --prod
|
|
|
|
# Copy built files
|
|
COPY --from=builder /app/services/mana-crawler/dist ./services/mana-crawler/dist
|
|
COPY --from=builder /app/packages/shared-drizzle-config/dist ./packages/shared-drizzle-config/dist
|
|
|
|
WORKDIR /app/services/mana-crawler
|
|
|
|
EXPOSE 3023
|
|
|
|
CMD ["node", "dist/main"]
|