How ruiromano/pbip-demo works
Project Type and Purpose
• This is a PBIP “folder‐based” Power BI project. It organizes two Semantic Models (Model01, Model02) and four Reports (Report01–Report04), and provides CI/CD pipelines and best-practice analysis.
• Purpose: manage Power BI artifacts (models, reports, themes) in source control, automate validation (BPA) and deploy to a Fabric workspace via REST APIs and fabric-cicd CLI.
File Structure
• Top-level:
– .bpa: best practice analysis scripts & data (bpa.ps1, JSON reports)
– .github/workflows: GitHub Actions (bpa.yml, deploy.yml)
– .resources: reusable themes and TMDL fragments
– Model01.SemanticModel & Model02.SemanticModel: folder-based PBIX definitions, TMDL files (tables, relationships, roles, perspectives)
– Report01.Report … Report04.Report: each report’s definition (.pbir), pages, visuals, static resources
– deploy.py: Python script for local deployment with fabric-cicd CLI
• Key patterns: Separation of semantic models vs reports; CI/CD via GitHub Actions; scripts for local invocation.
README Insights
• CI/CD instructions: service principal, secrets, fabric-cicd, REST API
• Two workflows: bpa (best practice analysis) & deploy (push content to Fabric)
Main Components
A. Source Control & CI/CD
1. GitHub Repository
2. GitHub Actions – bpa.yml (Best Practice Analysis)
3. GitHub Actions – deploy.yml (Deployment)
B. Scripts & CLIs
1. PowerShell script (.bpa/bpa.ps1)
2. Python script (deploy.py) & fabric-cicd CLI
C. Security / Credentials
1. Service Principal (client ID/secret/tenant) stored in GitHub Secrets
D. Power BI Artifacts
1. Semantic Models (Model01, Model02)
2. Reports (Report01–Report04)
3. Shared Resources (.resources folder)
E. Fabric Environment
1. Fabric CRUD REST API endpoints
2. Target Fabric Workspace (content endpoint)
Relationships & Data Flows
• Developer pushes code → GitHub Actions trigger
– bpa.yml runs bpa.ps1 → outputs rules report to .bpa/
– deploy.yml uses deploy.py + fabric-cicd → calls Fabric REST API using SP → uploads models & reports to workspace
• deploy.py (local) → fabric-cicd → Fabric REST API → workspace
Architectural Patterns & Principles
• Folder-based PBIP for modularity & source control
• Separation of concerns: models vs reports vs themes
• CI/CD pipelines for validation (BPA) and deployment
• Infrastructure as code style via fabric-cicd CLI and scripts