To help people better understand what Model Xtractor can do for them, I created a very cheap gig on Fiverr. There is no mention of Model Xtractor there, as Fiverr may not allow advertising own products this way. But people can send me a Microsoft Access, SQLite, Firebird or SQL Server CE database file, and I’ll send them back some static images from dynamic views generated with Model Xtractor.

1. Introduction

1.1. About This Gig

Click here for my Fiverr gig

I reverse engineer your MS Access, SQLite, Firebird or SQL Server CE database file, and generate six BETTER alternative views of your data model, without changing anything in your database.

Generated views could use:

  • Crow’s Foot, IDEF1X or a simplified notation
  • alternative colors and styles, from a spectacular theme
  • collapsed and minified tables (as shapes with no items
  • many-to-many relationships (bypassing the intersection tables)
  • nullable column data types and default values
  • table columns, keys, generated relationship items, and indexes
  • hidden table columns and column categories
  • description notes and … much more!

The basic package delivers six standard views, automatically generated by our internal tool, as static PNG images. For a little extra, you may ask for customization of these views, with different spectacular themes.

Look at the last presentation slide for all details.

The file you will send me must be for one of the following databases:

  • Microsoft Access – usually a .mdb or .accdb file;
  • SQLite.- usually a .db or .sqlite file;
  • Firebird – usually a .db file;
  • SQL Server CE – usually a .sdf file.

1.2. Microsoft Access – Relationships View

We’ll use just a typical Microsoft Access Northwind database as example. The process is similar to all other supported databases, such as SQLite, Firebird or SQL Server CE.

All you can get from within Microsoft Access’ IDE (Integrated Development Environment) in terms of ER data modeling is a Relationships view, with obvious limitations:

  • There is only one single view allowed per database;
  • Tables are all displayed expanded, with their column names;
  • You cannot show column data types and other column details;
  • There is one single specific notation of the relationships;
  • You have to manually arrange the tables, and there is no automatic layout;
  • The area gets quickly filled up when many tables;
  • Always same styles and colors;

2. Basic Package – No Revisions

I will generate six STANDARD image views of your data model.

For our previous Northwind database sample, the following types of extended views will be generated and delivered as PNG images.

These images should be accepted as they are, as the Basic package includes no revisions.

Position of shapes may vary.

2.1. Structural Model

This view is an alternative representation of the Microsoft Access Relationships view. It focuses on both internal table structure and the interconnectivity between tables. Tables are all expanded.

However, a stereotype icon can show different types of tables, as independent or dependent tables, intersection or lookup tables, and so on.

Relationship connectors follow a simplified Crow’s Foot notation: to-many relationships appear as a fork, and required foreign key values are displayed with plain lines.

The view replaces intermediate intersection tables – which are implementation details – by conceptual many-to-many relationships (something which is unique to this kind of diagrams). For instance, the Order Details intersection table has been replaced by a line with forks at both ends between Products and Orders.

Order Details appears now only as internal relationship items in the connected tables. Relationship items are internally generated and expand into actual connectors. When “collapsed”, they appear in gray within the table structure.

The table structure may also show indexes, at the end. And we can later include triggers, unique and check constraints.

2.2. Detailed Model

This view shows a mosaic representation, very rich in textual information, with the internal structure of all tables. There are no links between tables, as we do not focus here on interconnectivity.

Each shape may show three item categories: Keys, Columns and Indexes. Keys are table columns used as primary and/or foreign keys.

Table columns appear with their physical data type, the default value, and a star if nullable. Textual table and column descriptions appear as well.

2.3. Simple Model

This view is a clean alternative of the Detailed Model that eliminates clutter.

We show just table and column names, including internally generated relationship items (in gray) and indexes (at the end).

The view appears as a simple textual report about the tables and their internal structure.

2.4. Relationship Model

This view focuses on the interconnectivity between tables. Internal table structure is hidden, as the table shapes appear collapsed.

A stereotype icon can show different types of tables, as independent or dependent tables, intersection or lookup tables, and so on.

Relationship connectors follow the simplified Crow’s Foot notation described before: to-many relationships appear as a fork, and required foreign key values are displayed with plain lines.

Relationship connectors will also display Cascade Delete and Cascade Update icons – something no other type of diagram can show. Small empty rectangles mean NO ACTION, while a filled-in triangle represents a CASCADE condition. For instance, when a Customer is deleted, nothing happens to Orders linked to this customer (the empty white rectangle below Customers). However, when the ID of a Customer changes, related foreign key value will be automatically updated in the related Orders entries (the black triangle).

The view can be very useful for plenty of tables, when you want to focus on the foreign key constraints and do not care about the internal table structure.

2.5. Graph Model

This very clean view focuses on the interconnectivity between tables. Internal table structure is hidden.

The view simply represents the tables by their names, with straight lines for the relationship connectors.

Intersection tables – like Order Details – are omitted and replaced by conceptual many-to-many relationships – like the one between Orders and Products.

2.6. Topological Model

This very clean view focuses on the abstract level of interconnectivity between ALL tables. Both internal table structure and the actual table names are hidden.

This view represents ALL tables minified, by small bullets with no names. Relationships between appear as simple straight lines.

The view shows the complexity and the degree of connectivity between tables. This could be fairly important for large databases with plenty of tables.

3. Standard Package – Three Revisions

I will generate six CUSTOM image views of your data model.

You can ask for some types of changes on the six views described before:

  • Alternative standard ER Crow’s Foot or IDEF1X notation;
  • Tables you don’t want to appear at all in some models;
  • Tables you want to see either collapsed (with no columns) or expanded (with columns);
  • Tables you want to see minified (i.e. as a bullet, with no name);
  • Relationship connectors you want to see or not in a view;
  • Shapes manually moved to other positions;
  • Custom notes with your own description text;
  • Shape with same size, or shape positions automatically aligned;
  • Tables with or without all their items grouped in categories;
  • Columns with or without data types.

Remark category items and columns with data types are diagram switches. This means you can not show some shapes with categories, and others with no categories, in the same diagram. Or some tables with column data types, and others without.

3.1. Custom Structural Model

For instance, in this variation from the Structural Model, we made the following changes:

  • We displayed Categories, Customers and Shippers tables as collapsed;
  • We displayed Products table as minified (as a bullet);
  • We displayed category items;
  • We made some items and categories invisible (Employees shows only First and Last Name);
  • We displayed a note.

4. Premium Package – Custom Theme

I will generate six CUSTOM EXPERT image views of your data model, using a custom theme.

In addition to the previous types of changes, you can select a different theme for each view:

  1. Plain White
  2. All Silver
  3. All Teal
  4. Cassandra Navy
  5. Cassandra Navy Headless
  6. SQL Workbench

All following diagrams display the previous Custom Structural Model with a different theme.

4.1. Plain White Theme

4.2. All Silver Theme

4.3. All Teal Theme

4.4. Cassandra Navy Theme

4.5. Cassandra Navy Headless Theme

4.6. SQL Workbench Theme

Categories: Data Modeling

Cristian Scutaru

I designed and implemented the Data Xtractor suite, with Model Xtractor, Query Xtractor, and Visual Xtractor as separate modules. I am a software architect and developer with over 30 years professional experience. I’ve been working with relational databases for almost three decades and I was constantly unhappy with the relative limitation of those tools used to connect directly to a platform, and instantly extract and display data in flexible ways.