How ltplus-ag/nhmzh-plugin-qto works
Project type and purpose
• This is a full-stack, event-driven Quantity Take-Off (QTO) plugin for the NHMzh ecosystem.
• Frontend: React + TypeScript + Material-UI for interactive IFC model upload, display, editing.
• Backend: Python + FastAPI, using IfcOpenShell to parse IFC files, MongoDB for persistence, Kafka for event publishing.
• Additional microservice/library (qto_ifc-msg): standalone Node.js/TypeScript code that can publish QTO messages to Kafka and interface with MinIO.
File structure
• Top-level directories:
• frontend app under src/, with React pages, components, API client, types, styles, Vite config.
• backend under backend/, with main.py, parsing scripts, Kafka producer (qto_producer.py), MongoDB init, requirements, Dockerfile.
• qto_ifc-msg/: separate NPM package for Kafka + MinIO messaging, includes source, tests, Dockerfile, workflows.
• Config files & automation: Dockerfiles (Dockerfile.frontend, backend/Dockerfile.backend, qto_ifc-msg/Dockerfile), GitHub Actions workflows, nginx.conf, ESLint, tsconfigs.
README for insights
• Architecture section: confirms FastAPI + IfcOpenShell + Kafka + MongoDB on backend, React + MUI DataGrid on frontend.
• Kafka topics: PROJECT_UPDATED events, batched element data with full IFC metadata.
• Integration: downstream NHMzh modules (cost, LCA) consume these events.
Elements
• Backend and microservices in another (e.g., green)
Final advice
• Dockerized deployment by optionally grouping containers under a “Docker Compose” boundary.
• Keep labels consistent with code (use exact endpoint names, topic names, container names).
• Provide a small caption or footnote describing tech stack versions (e.g., React 18, FastAPI, Python 3.9, Kafka 2.x).
• Readability: position UI at the top/left, data stores at bottom, event bus in center.