Robust Programming

Robust Programming

Robust programming means writing code that handles unexpected inputs and errors gracefully without crashing. This topic covers defensive design, testing, and maintainability.

---

Defensive Design

Defensive design anticipates problems before they happen. The key techniques are:

Input Validation

Validation checks that data meets specific criteria before processing it. Common validation checks include:

Check	Purpose	Example
Range check	Value within acceptable range	Age must be 0–150
Type check	Correct data type	Price must be a number
Length check	Correct number of characters	Password must be 8–20 chars
Presence check	Field is not empty	Username cannot be blank
Format check	Matches expected pattern	Email must contain @
Lookup check	Value exists in a list	Country must be in approved list

Example (Python)

while True:
    age = input("Enter your age: ")
    if age.isdigit() and 0 <= int(age) <= 150:
        age = int(age)
        break
    else:
        print("Invalid. Enter a number between 0 and 150.")

Authentication

• Username and password — most common method
• Two-factor authentication (2FA) — adds a second verification step (e.g. text code)
• Access levels restrict what different users can do

Input Sanitisation

Sanitisation cleans input to prevent malicious data (e.g. removing <script> tags to prevent SQL injection and cross-site scripting).

---

Testing

Testing ensures the program works correctly. There are several types:

Types of Testing

Type	Description	When Used
Iterative testing	Testing during development, after each module	Throughout development
Final/terminal testing	Testing the complete program	End of development
Unit testing	Testing individual components in isolation	During development
Integration testing	Testing components working together	After unit testing
Alpha testing	In-house testing by developers	Before release
Beta testing	Testing by a limited group of end users	Before full release

Test Data

You must use three types of test data:

Data Type	Purpose	Example (for age 0-150)
Normal	Typical valid input	25, 67, 10
Boundary	Edge of valid range	0, 1, 149, 150
Erroneous	Invalid input that should be rejected	-5, 200, "abc", ""

Trace Tables

A trace table tracks the value of each variable as a program executes, line by line. They help you dry run code to find logic errors.

x = 1
WHILE x <= 4
    x = x * 2
END WHILE

Iteration	x	x <= 4?
Start	1	True
1	2	True
2	4	True
3	8	False — exit

---

Types of Errors

Error	Description	Example
Syntax error	Breaks language rules; won't run	Missing colon after if
Logic error	Runs but gives wrong results	Using + instead of -
Runtime error	Crashes during execution	Division by zero

---

Maintainability

Writing code that others (or your future self) can understand and update:

• Comments — explain the purpose of code sections using # in Python
• Meaningful variable names — totalScore not x
• Indentation — consistent indentation shows code structure
• Sub-programs — break code into reusable functions and procedures
• Constants — use named constants like MAX_LIVES = 3 instead of magic numbers
• Modular design — separate program into logical modules

Example of Good vs Bad Practice

Bad:

x = 10
y = x * 1.2

Good:

VAT_RATE = 1.2  # Current UK VAT rate
price_before_vat = 10
price_after_vat = price_before_vat * VAT_RATE

---

Exam Tips

• Always give specific examples of validation — don't just say "check the input"
• When writing test plans, include all three types of test data (normal, boundary, erroneous) with expected outcomes
• Trace tables are frequently asked — practise following code line by line
• Remember the difference between validation (is it reasonable?) and verification (is it what the user intended?) — verification uses methods like double entry or screen confirmation