Program Debugging

Overview

This is our first post about the The Debugging Book! The goal of this post is to introduce different debugging methods and utilize these methods within our Chasten and Cellveyor tools.

Summary

The Introduction to Debugging chapter explores the topic of debugging through the example of fixing issues in a Python program that removes HTML tags. It explores common debugging challenges and introduces a systematic approach using the scientific method. A specific example involves testing a hypothesis about quote handling and using assertions for confirmation. The importance of a clear diagnosis and post-fix steps like testing, adding assertions, committing changes, and closing bug reports is emphasized. Overall, it provides insights into effective debugging strategies while highlighting the significance of thorough documentation and clear thinking during the debugging process.

The chapter also introduces a systematic debugging strategy based on the scientific method. It emphasizes the importance of understanding the code, fixing the problem instead of symptoms, and proceeding systematically. The scientific method is applied to debugging, involving formulating hypotheses, making predictions, testing hypotheses through experiments, and refining hypotheses based on observations.

The chapter provides a specific example where a hypothesis about the removal of double quotes from tagged input is tested and confirmed using the assert statement. The debugging process involves identifying defects, faults, and failures, and tracing the cause-effect chain to locate and fix the underlying issues. The passage concludes by highlighting the challenges of debugging, including the complexity of program states and the lack of specifications:

# Initial code segment with the hypothesis
elif c == '"' or c == "'" and tag:
    assert False  # <=== Just added
    quote = not quote

To test the hypothesis, the author introduces an assert statement to check whether the condition is faulty. They run the program with an input that triggers this condition:

result = remove_html_markup_with_quote_assert('"foo"')

The correct fix involves adding parentheses to ensure the corrected condition:

elif (c == '"' or c == "'") and tag:
    quote = not quote

The incorrect condition fails to handle single and double quotes correctly. The corrected condition ensures that the code inside the block is executed only when either a single or double quote is encountered and the tag variable is true.

After fixing the code, the author suggests post-fix activities, including checking for further occurrences of the defect, validating tests, adding assertions for future correctness checks, committing the fix to the code repository, and closing the bug report. These activities ensure the long-term health and correctness of a software application.

Reflection

After reading and interpreting this article, we as software engineers can learn more about debugging. We learn about how simple debugging is and how we can make it legible with Python. Debugging means to simultaneously find and fix problems within our code. The author wants developers to approach debugging by using the scientific method.

This chapter uses a Python function to concretely explore the debugging process. In the provided function there is a bug that causes it to not work properly. The article shows in a step-by-step fashion what is wrong and explains how to debug this function. The message being conveyed in this article is to simply use a systematic and methodical approach when going about debugging. Of course, this is often easier said than done!

Action Items

The cause-and-effect chain can be helpful to keep in mind when writing test cases for our code. Having a deeper understanding of the steps in the cause-and-effect chain can allow us to write better test cases to debug our program. Knowing that the chain starts with a human action that results in an error which then results in errors in the code may help us figure out how to debug a section of the program or an input that causes an error.

Using the idea of the scientific method in this chapter can also be beneficial to look at bugs that are within the program or function. Taking the approach of the scientific method and creating a hypothesis as to what might be causing the error that the program or function is producing is a good first step in diagnosing the bug. Predicting what may happen when you modify a section or piece of code and checking it to see if the output is what you predicted it to be will help us debug a program or function. Overall, this chapter gives key insights into the best methodology to use when debugging a function or program. We can use it to fix bugs in our own programs!

Return to Blog Post Listing