Lecture Notes: Object-Oriented Database Management System (ODBMS) with Objectivity DB

Introduction

• Discussion on the current topic: Object-oriented DBMS, specifically Objectivity DB.
• Focus: Building an application using C++ and defining schemas using DDL (Data Definition Language) in C++.
• DDL syntax is kept similar to C++ to minimize learning curve for developers.

Defining Schema in C++

• Class Definition: Process of defining classes and attributes using C++ syntax.
• Associations: Represent real-life relationships between different objects.
• Storage Mechanisms:
  • Inline: Specific memory area allocated for the application. More efficient for direct object references.
  • Non-Inline: Uses a system-wide specific area for storing references between objects. Less efficient due to potential searching overhead.

Association Types & Storage

• Cardinality: One-to-one, one-to-many, many-to-many relationships.
• Inline Behavior: Each reference has its specific area, preventing ambiguity and enhancing access efficiency.
• Non-Inline Behavior: When adopting non-inline, all references stored in a general system-specific area.
  • Efficiency: Non-inline may be memory efficient but has access inefficiency.
  • Memory Overhead: Storage behavior involves certain memory overheads; tables indicate space requirements for different storage models.

Defining Associations in C++

• Syntax Formats:
  • Unidirectional Association: One object can access another, but not vice versa.
  • Bidirectional Association: Both objects can access each other.
  • Example: Syntax for defining different types of associations (one-to-one, one-to-many, many-to-many).
• Classes and Links: Defined links between different classes, specifying behaviors like copy and delete.

Efficiency Comparison

• Memory Efficiency: Non-inline is memory efficient as it uses the same space for all objects but has limitations on object reference count.
• Access Efficiency: Inline is access efficient as it has dedicated memory for specific associations but is memory inefficient.

Changing Storage Behavior

• Modify DDL: Change from non-inline to inline or vice versa by modifying the DDL and reprocessing it.
• Adopt Changes: Ensure the interface of the application is updated to reflect changes in storage behavior.

Behavior Specifications

• Propagation: Operations performed on an object propagate to its references (e.g., delete propagation).
• Copy Behavior: Different options like copy-copy, copy-move, and copy-delete.
• Versioning Behavior: Similar to copying, determines how references are managed during version creation.

Practical Examples

• Examples of defining behaviors for different classes and associations.
• Emphasis on automatic handling by DBMS, ensuring bidirectional links are maintained as relationships change.

SQL++ in Objectivity DB

• SQL++: Used for direct access by specific users, similar to SQL for relational databases, but for objects.
• Modes:
  • Interactive Mode: Direct commands and immediate responses.
  • ODBC Mode: Embedded within applications using ODBC modules.
• General Architecture: Conversion of SQL++ commands to object operations handled by Objectivity DB kernel.
• Comparison with Relational Concepts: Tables vs. object collections, columns vs. attributes, etc.

Object Access and Limitations

• Use by Specific Users: Only certain users (e.g., DBA) have the rights for direct access.
• Constraints: SQL++ supports basic object types only, due to complexity.
• Inheritance and Views: Inheritance aligns with views within SQL++ environments.

Summary

• Objectivity DB: Combines ODBMS principles with C++ and DDL for schema definition and management.
• EFF and IEF: Efficient in different contexts based on usage and requirements.
• Behavior Definitions: Crucial for managing relationships and object handling within the database.

Next Steps

• Transition from DDL to Objectivity SQL++ for object-oriented programming and direct data access, covering methods and operations in an OOP context.

End of lecture.