Usage Guide

This guide explains how to use the modern React-based Mermaid to Dataverse application to transform Mermaid entity-relationship diagrams into Microsoft Dataverse solutions, tables, columns, and relationships with intelligent validation and Common Data Model (CDM) integration.

Getting Started

1. Initial Setup

The application is deployed to Azure App Service. See Deployment Guide for deployment instructions.

2. Prerequisites

• Mermaid ERD file (.mmd extension) with valid ER diagram syntax
• Dataverse environment with admin access
• Global Choices file (optional) - JSON file with choice definitions if needed

Overview

The application provides a modern React wizard interface with Fluent UI v9 components for uploading Mermaid ERD files and automatically creating corresponding Dataverse entities. Once deployed, access the React application: https://your-app-name.azurewebsites.net

Key Features:

• Smart CDM Detection: Automatically identifies Common Data Model entities
• Intelligent Validation: Real-time error detection and auto-corrections
• Type Detection: Smart data type mapping based on field names
• Modern UI: Intuitive React wizard
• Secure: Azure Managed Identity authentication

Use the Wizard

Step 1: ERD Upload & Intelligent Validation

• Browse: Upload your Mermaid ERD file (.mmd extension) using the modern file upload component

• Real-time Parsing: Immediate ERD structure analysis and entity detection

• Smart Validation: Comprehensive validation with automatic error detection:

  • CDM Detection: Identifies entities matching Common Data Model (e.g., Account, Contact)

• Naming Conflicts: Detects conflicts with Dataverse system columns
• Status Columns: Automatically filters out status columns (uses Dataverse built-in status)
• Primary Key Issues: Validates primary key structure
• Data Type Mapping: Smart type detection based on field names
• Auto-Corrections: View suggested fixes with side-by-side comparison

• Select Next to proceed

Step 2: Solution & Publisher Configuration

• Solution Name: Name for your Dataverse solution (e.g., "Customer Management") or choose from an existing solution
• Publisher Selection: Choose an existing publisher or create a new one with the React interface
• Publisher Prefix: 3-8 character prefix (e.g., "cmgt") with real-time validation
• Select Next to continue

Step 3: Global Choices Integration

• Upload JSON: Upload JSON file with global choice definitions

• Browse Existing: View and select from existing Dataverse global choices

• Preview Choices: Real-time preview of choice sets before deployment
• Combined Selection: Mix new and existing choices in a single deployment
• See Global Choices Guide for detailed documentation
• Select Next to continue

Step 4: Review & Deploy

• Configuration Summary: Complete review of all settings with modern UI

• CDM Integration Summary: Clear display of CDM vs. custom entity choices
• Real-time Validation: Final validation before deployment
• Deploy: Select Deploy to Dataverse to create your solution with live progress tracking

And finally

Intelligent Validation & Auto-Corrections

The React application includes sophisticated validation that provides real-time feedback and intelligent fixes:

Common Data Model (CDM) Detection

The system automatically analyzes your entities and detects potential CDM matches:

What happens:

CDM Detection Results:

• Contact detected: Matches CDM Contact entity
• Account detected: Matches CDM Account entity
• 💡 Recommendation: Use existing CDM entities for better Power Platform integration

User Choice:

• Use CDM entities: Leverage existing Contact and Account with built-in relationships
• Create custom entities: Proceed with your custom entity definitions
• Mixed approach: Use CDM for some, custom for others

Automatic Problem Detection & Fixes

Primary Column Conflicts

Problem detected:

Customer {
    string customer_id PK
    string name "Customer name"  // Conflicts with auto-generated primary column
}

Auto-suggested fix:

ERD
Customer {
    string customer_id PK
    string customer_name "Customer name"  // Renamed to avoid conflict
}

Status Column Filtering

Original ERD:

Order {
    string order_id PK
    string status "Order status"  // Will be automatically ignored
}

System handling:

• Status columns automatically filtered out during deployment
• Built-in Dataverse status fields (Active/Inactive) used instead
• Suggestion provided to create choice columns manually for custom status values

System Column Conflicts

Problem detected:

Task {
    string task_id PK
    string ownerid FK "Task owner"  // Conflicts with Dataverse system column
}

Auto-suggested fix:

Task {
    string task_id PK
    string task_ownerid FK "Task owner reference"  // Prefixed to avoid conflict
}

Smart Type Detection

The system includes intelligent field name analysis for automatic type improvement:

Email Detection:

Contact {
    string user_email "User email"      // → Automatically becomes 'email' type
    string contact_email "Contact email" // → Automatically becomes 'email' type
}

Phone Detection:

Contact {
    string mobile_phone "Mobile number"  // → Automatically becomes 'phone' type
    string office_tel "Office telephone" // → Automatically becomes 'phone' type
}

URL Detection:

Company {
    string website_url "Company website" // → Automatically becomes 'url' type
    string homepage "Company homepage"    // → Automatically becomes 'url' type
}

Complete Data Type Support

Mermaid Type	Dataverse Type	Description	Auto-Detection
string	Single Line Text	Text up to 4,000 chars	Email, Phone, URL detection
text / memo	Multiple Lines Text	Large text fields	N/A
email	Email	Email with validation	Auto-detected from field names
phone	Phone	Phone with formatting	Auto-detected from field names
url	URL	URL with validation	Auto-detected from field names
int / integer	Whole Number	Integer values	N/A
decimal	Decimal Number	Fixed precision decimal	N/A
money	Currency	Currency with formatting	N/A
boolean / bool	Two Options	Yes/No values	N/A
datetime	Date and Time	Full timestamp	N/A
date / dateonly	Date Only	Date without time	N/A
float / double	Floating Point	Floating point numbers	N/A
ticker	Ticker Symbol	Stock ticker	N/A
timezone	Time Zone	Time zone identifier	N/A
language	Language	Language LCID	N/A
duration	Duration	Duration in minutes	N/A
file	File	File storage	N/A
image	Image	Image storage	N/A

Supported Constraints & Markers

Constraint	Meaning	Implementation	Smart Validation
PK	Primary Key	Creates GUID primary key + auto-generated name field	Validates single PK per entity
FK	Foreign Key	Used for automatic relationship creation	Validates target entity exists
UK	Unique Key	Creates unique constraint	Validates uniqueness rules

Advanced Relationship Examples with CDM Integration

Smart Processing Results:

• CDM Detection: Account & Contact identified as CDM entities
• Type Detection: website_url → URL, business_email → Email, mobile_phone → Phone
• Validation: All relationships validated, junction table recognized
• Mixed Solution: CDM entities + custom entities in single deployment
• Optimized: Best of both worlds - standard CDM + business-specific entities

---

Rollback Deployments

The application provides modular rollback functionality, giving you granular control over which components to remove from Dataverse while preserving deployment history.

Accessing Rollback

1. Navigate to Deployment History (view past deployments)
2. Find the deployment you want to rollback
3. Select the "Rollback" button
4. Select components to remove

Rollback Options

You can choose exactly what to rollback:

Available Components

• Custom Entities - Delete custom tables created by deployment
• Global Choices - Delete custom global choice sets
• Solution - Remove the solution container
• Publisher - Delete the publisher

Dependency Rules

The system automatically enforces safe rollback order:

1. Custom Entities require Relationships to be deleted first - this happens automagically
2. Solution requires all Custom Entities + CDM Entities to be removed (auto-selected)
3. Publisher requires Solution to be deleted first (auto-selected)

Multiple Rollbacks

You can perform multiple sequential rollbacks on the same deployment.

Smart Tracking:

• System tracks what was deleted in each rollback
• Prevents attempting to delete already-removed components
• Each rollback shows only the components selected for that specific operation
• Deployment status updates automatically:
  • deployed - Active deployment with all components
  • modified - Partial rollback (some components remain)
  • rolled-back - Complete rollback (all components deleted)

Rollback Best Practices

⚠️ Important Considerations

1. CDM vs Custom: CDM entities are only removed from solution, not from Dataverse
2. Publisher Dependencies: Cannot delete publisher if solution exists
3. Irreversible: Rollback cannot be undone (re-deploy if needed)

Viewing Rollback History

Each deployment card in the history shows:

• Status badge (deployed/modified/rolled-back)
• Rollback badges listing what was rolled back in each operation
• Rollback numbering in chronological order
• Component details for each rollback

Example Display:

Deployment: university_v1.0
Status: modified
Rollbacks:
  #1 - Relationships (Jan 15, 2025)
  #2 - Custom Entities (Jan 16, 2025)