Testing

Focuses on implementation of code
The intention is to write code that qualifies the requirements and break it up by finding bugs.
The tester should be a third party resource
Special Independent testing / Quality Assurance Departments
Testing from the small to large

Testing Strategies

Gives a road map / plan
What to do when to do, who to do etc.

Unit Testing

Individual functions called as ‘Units’ of the systems are tested independently.

Integration Testing

We combine modules together one by one and test them.

1. Stubs in Top-Down Testing:
   1. In top-down testing, the higher-level modules are tested first before the lower-level modules are available. Stubs are used to simulate the behavior of the lower-level modules that are not yet implemented. Stubs receive inputs from the higher-level modules and return pre-determined outputs. They are primarily used to provide the necessary responses to allow testing of the higher-level modules.
2. Drivers in Bottom-Up Testing:
   1. In bottom-up testing, the lower-level modules are tested first before the higher-level modules are available. Drivers are used to simulate the behavior of the higher-level modules that are not yet implemented. Drivers provide inputs to the lower-level modules and handle their outputs. They are primarily used to test the lower-level modules in isolation and verify their functionality.

Validation Testing

• Verification

Correct Working
Product right?

Link to original

• Validation

Correct Requirement implemented
Right Product?

Link to original

System Testing

Entire system is tested as in one organism.

Testing Techniques

Focus on how to do it

White Box

AKA Glass Box / Structural Testing
We look at the code, loops, if else etc.
Tests the processing logic
Costs more, expensive

Basic Path Testing

Look more in Basic Path Testing.

Loop Testing

Start with the inner loop and then proceed to outer loops

Conditional Testing

Based on evaluation of conditions

Black Box

AKA Functional Testing
We look the software from outer boundaries
Ignores the processing logic
Large number of testing that can only be done using the blackbox testing

Equivalence Class Partitioning

Inputs
Partitioning into different equivalence classes (Valid or Invalid)
Partitions are actually the buckets
Pick and execute for only one equivalence class, the reason is that all of them should give the same answer

Boundary Value Analysis (BVA)

The focus is specially on boundary
Continuous variables have boundaries
We focus specially on boundaries, as the most of the problems lie there

Quote

Bugs lurk in corners and congregate at boundaries

Example

Black box test of ticketing system using Boundary Value Analysis

Ticket price depends on:

Inputs:

• Vistor Status - M / OT
• Dady of Visit - WD / WED
• Time of Day - 1st shift / 2nd shift
• Age Bracket - Y / M / E

Hence, 24 possible combinations of input

The number of invalid equivalent classes will be the number of our invalid test cases

Test cases of boundary and non boundaries will be kept separate

We exhaust all the boundaries of all buckets of valid equivalence classes

Seealso

Differences between white box and black box?

Grey Box

Combination of the above two techniques

---

Automated Testing

• Coverage Monitor
  • Code?
• Qualification Testing
  • Qualify codes if it follows certain conventions or not
• Blackbox Functional Testing Tools
  • Selenium, Usability testing
  • Regression Testing
    • Existing test cases which are checked whenever the code is changed
    • Result of changing the code
• Load Testing
• Test Case Management Tools
  • Tracks the entire lifecycle for bug

Insight

The biggest consumer of QA is Testing and the biggest consumer of testing is ??

Info

Testing never stops. It just tosses from one end to another.