Data Views for Database Schema Visualisation
Create Data Views in persistence mode to visualise entity-relationship schemas derived from the central data model. Select entities with the entity picker, assign a database profile for type resolution, and inspect entities via the Entity Overlay.
Data Views in persistence mode replace traditional static ERD diagrams with an interactive graph-based approach. Entities appear as nodes on a force-directed canvas. Click any node to open the Entity Overlay, which shows the full property table with ERD-style columns - key roles, field names, concrete database types, nullable, unique, defaults, FK references, constraints, and comments. All entity data comes from the central model (model.neoarc); the view stores only layout and profile configuration.
Creating a Persistence Mode Data View
Database Profile Selection
Each persistence view references a databaseProfileId that controls type and name resolution. Profiles are registered in project.neoarc and stored as .neoarc-db-profile.json files. Five vendors are supported out of the box.
| Vendor | Default Naming | Example Type Mappings |
|---|---|---|
| SQL Server | PascalCase | string to NVARCHAR, integer to INT, boolean to BIT, uuid to UNIQUEIDENTIFIER |
| PostgreSQL | snake_case | string to VARCHAR, integer to INTEGER, boolean to BOOLEAN, uuid to UUID |
| MySQL | snake_case | string to VARCHAR, integer to INT, boolean to TINYINT(1), uuid to CHAR(36) |
| Oracle | SCREAMING_SNAKE_CASE | string to VARCHAR2, integer to NUMBER(10), boolean to NUMBER(1), uuid to RAW(16) |
| MongoDB | camelCase | string to String, integer to Int32, boolean to Boolean, uuid to UUID |
Relationship Edges and FK Inference
Edges between nodes on the graph canvas are inferred automatically from foreign key relationships defined in the model. When both the source and target entities of an FK relationship are included in the view, an edge is drawn between their nodes with cardinality labels. No manual connection drawing is needed.
Side-by-Side Entity Comparison
Pin an Entity Overlay to dock it below the canvas, then click another node to open a second overlay. This lets you compare two entities' schemas side by side. When docked overlays share a foreign key relationship, connecting lines are drawn between the relevant FK and PK rows.
Key Indicators
| Indicator | Key Role | Description |
|---|---|---|
| PK | primary | Primary key, the canonical identifier for the entity |
| FK | foreign | Foreign key, references another entity's key |
| NK | natural | Natural key, a business-meaningful identifier (e.g. email, SSN) |
| SK | surrogate | Surrogate key, a system-generated identifier (e.g. UUID, auto-increment) |
| CP | composite-part | Participates in a composite key |
Standalone ERD Table Shapes
For freeform diagramming without a central model, the standard diagram editor includes standalone ERD table shapes. These shapes store their own columns, data types, and relationships directly in the diagram file. They support crow's foot notation, row-level ports, FK auto-fill, SQL Server import, and CSV export.
Standalone ERD tables are suitable for quick sketches and one-off diagrams. For structured data modelling, persistence mode Data Views are the recommended approach because they keep all entity definitions in a single source of truth.