Extensible Design for the Output Method of PrintShape

[YongJin-Cho]

Extensible Design for the Output Method of PrintShape (see PrintShape)

Problem Statement

The PrintShape function currently supports a limited set of output methods. To enhance its versatility, we need to extend its capabilities to include various output methods such as:

• String printing: Output the shape as a string.
• File printing: Save the shape to a text file.
• BMP file printing: Export the shape as a BMP image file.

By designing the output methods to be easily extendable, future modifications and maintenance will be more straightforward, ensuring that new output methods can be added with minimal changes to the existing codebase.

Design Candidates

Candidate 1: PrintShape provides setOutput and setCharacters

In this design, the PrintShape class is responsible for setting the output method and the characters used for printing the shape. The class will have the following methods:

• setOutput(output_stream): Sets the output method (e.g., console, string, file).
• setCharacters(characters): Sets the characters used to represent the shape.

classDiagram
    class PrintShape {
        +print(size) // Prints the shape to the specified output_stream with characters
        +setOutput(output_stream)
        +setCharacters(characters)
    }

Loading

Candidate 2: PrintShape creates a boolean image and prints it via the Output interface

In this design, the PrintShape class creates a boolean image of the shape and delegates the printing to an Output interface. The Output interface will have a method to print the shape, and StreamOut, which extends the Output interface, will provide methods to set the output method and characters:

• Output.print(boolean_image): Prints the boolean image of the shape.
• StreamOut.setOutput(output_stream): Sets the output method.
• StreamOut.setCharacters(characters): Sets the characters used to represent the shape.

classDiagram
    class PrintShape {
        -output
        +print(size) // creates boolean image and prints it via output
    }
    
    class Output {
        <<interface>>
        +print(boolean_image)
    }
    
    class StreamOut {
        +setOutput(output_stream)
        +setCharacters(characters)
        +print(boolean_image)
    }
    
    PrintShape ..> Output
    Output <|.. StreamOut

Loading

Pros and Cons

Candidate 1: PrintShape provides setOutput and setCharacters

Pros:

• Simple and straightforward implementation
• Easy to understand for new developers
• Minimal code complexity
• Direct control over output methods

Cons:

• Violates the Single Responsibility Principle (SRP) - PrintShape is responsible for both shape creation and output handling
• Less flexible for adding new output methods (requires modifying the PrintShape class)
• Higher coupling between shape creation and output formatting
• More difficult to test output methods independently
• Not suitable for extensions that output colors rather than characters, such as BMP.

Candidate 2: PrintShape creates a boolean image and prints it via the Output interface

Pros:

• Follows the Open/Closed Principle - open for extension, closed for modification
• Separation of concerns - shape creation and output handling are separate
• More extensible - new output methods can be added by implementing the Output interface
• Easier to test each component independently
• Promotes loose coupling between components

Cons:

• More complex initial implementation
• Requires understanding of interfaces and polymorphism
• Slightly higher learning curve for new developers
• May involve more classes and code organization

Comparison Table

Aspect	Candidate 1: PrintShape provides setOutput and setCharacters	Candidate 2: PrintShape creates a boolean image and prints via Output interface
Architecture	Monolithic design	Separation of concerns via interfaces
Extensibility	Limited - requires modifying PrintShape	High - new output methods can be added by implementing Output interface
Complexity	Low - straightforward implementation	Medium - requires understanding of interfaces and polymorphism
Testability	Harder to test output methods independently	Easier to test components independently
SOLID Principles	Violates Single Responsibility Principle	Follows Open/Closed Principle
Learning Curve	Easy for new developers	Steeper initial learning curve
Maintenance	Higher coupling makes changes more difficult	Lower coupling facilitates maintenance
Flexibility	Limited for non-character outputs (e.g., BMP)	Supports diverse output formats through different implementations
Code Organization	Fewer classes, centralized logic	More classes, distributed logic

Design Outcome

After evaluating the design candidates, Candidate 2 has been selected as the preferred approach. By having PrintShape create a boolean image and delegating the printing responsibility to the Output interface, we achieve:

1. Better separation of concerns
2. Higher extensibility for adding new output methods
3. Improved testability of individual components
4. Compliance with SOLID principles, particularly the Open/Closed Principle

This design allows us to easily implement all the required output methods (string printing, file printing, BMP file printing) by creating different implementations of the Output interface without modifying the core PrintShape logic.

The modular architecture will make it straightforward to add more output formats in the future, such as SVG, HTML, or console output with ANSI colors, by simply creating new implementations of the Output interface.