How civicdatalab/bma-dashboard-superset works
Project type and purpose
• From the name “dashboard-superset” and minimal README, this appears to be an extension or customization layer on top of Apache Superset—a full-stack, data-viz/dashboarding platform. The goal is likely to provide a prescriptive set of dashboards or UI enhancements for Superset.
(current and expected) file structure
• At the moment there are only LICENSE and README.md, which suggests the code hasn’t been laid out yet in your repo. In a typical Superset-extension project you would see:
• a “superset_ext” or “plugin” directory containing Python files (Flask views, hooks)
• a “superset_ext/static” folder for JavaScript/TypeScript or CSS assets
• a “superset_ext/templates” folder for Jinja templates
• a configuration file (e.g. setup.py or pyproject.toml)
• possibly a “docker” or “ci” folder for deployment/build scripts
README for additional insights
• As it stands the README only names the project. When expanded, look for:
• an “Architecture” or “How it works” section
• a list of dependencies (e.g. Superset version, Python, Node)
• any code snippets showing plugin registration or configuration
Main components
a. User’s Web Browser (React/JS front‐end)
b. Superset Core Server (Flask + Superset’s internal API + plugin hooks)
c. Custom Dashboard Plugin (your Python & JS code, registered via Superset’s plugin framework)
d. Metadata Database (PostgreSQL or MySQL storing Superset metadata: dashboards, slices, users)
e. External Data Sources (e.g. PostgreSQL, Presto, Druid—where the dashboards read their data)
f. Build & Deployment Pipeline (Docker, CI/CD scripts, helm charts, etc.)
Relationships and interactions
• Superset Server interacting with the Metadata DB for dashboard definitions.
• Superset Server querying External Data Sources for chart data.
• Build/deploy pipeline feeding container images or config into the Superset runtime environment.
• If an auth service is used, show the flow of tokens or redirects between browser, Superset, and the auth provider.
Architectural patterns and design principles
• Plugin Architecture: highlight how your extension code hooks into Superset’s plugin registry.
• MVC within Superset: model (metadata DB), view (React components), controller (Flask routes).
• Layered separation: build pipeline → infrastructure (containers/helm) → runtime (Superset + plugin).
Relevant technologies, frameworks, libraries
• Backend: Python 3.x, Flask, Superset’s plugin framework
• Frontend: React (JS/TS), Ant Design (if used by Superset), custom CSS/JS in “static”
• Data stores: PostgreSQL/MySQL for metadata, plus any external catalogs
• CI/CD: Docker, GitHub Actions/Jenkins, helm/Terraform
Next steps
• Populate your repo with the expected directory structure and configuration files.
• Flesh out the README’s “Architecture” section with brief component descriptions.