What is the essence of white-box testing?

White-box testing is a software testing method that examines the internal structure, design, and logic of the code. The tester has access to the source code and designs test cases based on it.

Main principles:

• Code coverage: Test cases are developed to achieve maximum coverage of various code elements (statements, branches, paths).
• Knowledge of internal structure: Requires understanding of how the program works internally.
• Detection: Errors in logic, incorrect execution paths, issues with loops and conditional statements.
• Automation: Often used in unit tests, integration tests, and security tests at the code level.

Coverage types:

• Statement Coverage: Executing each statement in the program at least once.
• Branch Coverage: Executing each condition (if, else, switch) as true and false.
• Path Coverage: Executing all possible execution paths through the program. (Most comprehensive but often impractical).

Example of branch coverage in Python:

# Example function
def calculate_discount(price, is_member):
    if price > 100 and is_member: # Branch 1 (true)
        discount = price * 0.1  # Branch 2 (executed if Branch 1 is 'true')
    else: # Branch 3 (false)
        discount = 0            # Branch 4 (executed if Branch 1 is 'false')
    return price - discount

# Test cases for branch coverage:
# 1. price > 100 and is_member == True (covering Branch 1 true, Branch 2)
# 2. price <= 100 or is_member == False (covering Branch 1 false, Branch 3, Branch 4)

Advantages:

• Detects errors early in development.
• Ensures that important parts of the code are tested.
• Helps understand how the code works and improve its design.

Disadvantages:

• Requires deep knowledge of the code.
• Can be labor-intensive for large and complex systems.
• Does not test the program's operation from the end-user perspective (which is the task of black-box testing).

The essence is to check "how" the program works, not just "what" it does.