README.md

Debugr

Debugr is a command-line tool that leverages Claude AI to assist with debugging and testing tasks. It is written using only Go standard libraries, without any additional dependencies.

Installation

To install Debugr, follow these steps:

1. Ensure you have Go installed on your system (version 1.16 or later).

2. Clone the repository:

   git clone https://github.com/abayomipopoola/debugr.git
   cd debugr
   

3. Install the binary:

   go install ./cmd/debugr
   

This will compile the program and install the binary in your $GOPATH/bin directory. Make sure this directory is in your system's PATH.

Usage

After installation, you can run Debugr from anywhere in your terminal:

debugr [flags] <prompt>

Flags:

• --context <file>: Specify a single file as context
• --context-dir <directory>: Specify a directory as context
• --debug: Enable debug mode
• --dry-run: Print suggested actions without executing

Examples:

1. Write test for this simple addNumber function:

Directory structure:

$ tree
.
└── adder
   └── adder.go

adder.go:

package adder

func addNumbers(x, y int) int {
	return x + y
}

Before running Debugr, set your ANTHROPIC_API_KEY environment variable:

export ANTHROPIC_API_KEY=your_api_key_here

Then run:

debugr --debug --context adder/adder.go "write test for the addNumber"

Output:

...
This will run the test and show you the results. If everything is working correctly, you should see a message indicating that the test passed.
To write a test for the `addNumbers` function, we need to create a new test file in the same package. Here's how we can do that:
Execute this action? (y/n): y
File create: adder/adder_test.go
Now you can run the test using the following command:
Execute this action? (y/n): y
ok      test_examples/adder     0.745s
This will run the test and show you the results. If everything is working correctly, you should see a message indicating that the test passed.

Updated directory structure:

$ tree
.
└── adder
   ├── adder.go
   └── adder_test.go

adder_test.go:

// ...
func TestAddNumbers(t *testing.T) {
	tests := []struct {
		name     string
		x        int
		y        int
		expected int
	}{
		{"positive numbers", 2, 3, 5},
		{"negative numbers", -2, -3, -5},
		{"mixed numbers", -2, 3, 1},
		{"zeros", 0, 0, 0},
	}
	for _, tt := range tests {
		t.Run(tt.name, func(t *testing.T) {
			result := addNumbers(tt.x, tt.y)
			if result != tt.expected {
				t.Errorf("addNumbers(%d, %d) = %d; want %d", tt.x, tt.y, result, tt.expected)
			}
		})
	}
}

2. Fixing a Python code bug and testing, with directory context:

Directory structure:

$ tree
.
├── binarygap
│   └── binary_gap.py
└── twosum
    └── twosum.py

twosum.py (incorrect solution):

"""
Given an array of integers nums and an integer target, return indices
of the two numbers such that they add up to target.
"""
def solution(nums, target):
    ls = len(nums)
    for i in range(ls):
        for j in range(i, ls):   # bug: should start from i+1
            if nums[i] + nums[j] == target:
                return [i, j]

nums = [3,2,4]
target = 6
print(solution(nums, target))

Run:

debugr --context-dir . "debug the twosum: maintain the inefficient solution for now"

Output:

...
To debug the twosum solution while maintaining its current (inefficient) approach, we can make a few adjustments to improve its correctness and add some debugging information. Here are the steps:
1. Modify the inner loop to start from i+1 to avoid using the same element twice:
Execute this action? (y/n):

Updated twosum.py:

def solution(nums, target):
    ls = len(nums)
    for i in range(ls):
        for j in range(i + 1, ls ):  # Start from i+1 to avoid using the same element twice
            print(f"Checking: nums[{i}] = {nums[i]}, nums[{j}] = {nums[j]}")  # Debugging print
            if nums[i] + nums[j] == target:
                return [i, j]
    return None  # Return None if no solution is found

# Test cases
nums1 = [3, 2, 4]
target1 = 6
print("Test case 1:", solution(nums1, target1))
...

3. Optimizing code and writing tests:

Directory structure:

$ tree
.
├── binarygap
│   └── binary_gap.py
└── twosum
    └── twosum.py

Run:

debugr --context-dir . "debug the twosum: maintain the inefficient solution for now"

output:

...
Certainly! I'll optimize the `twosum` function for better time complexity and provide a unit test for it. Here are the steps:
1. Optimize the `twosum` function:
Execute this action? (y/n): y
File modify: twosum/twosum.py
2. Create a new file for unit tests:
Execute this action? (y/n): y
File create: twosum/test_twosum.py
3. Run the unit tests:
The optimized `solution` function now has a time complexity of O(n) instead of O(n^2). It uses a dictionary to store the complement of each number as we iterate through the list, allowing us to find the solution in a single pass.

Updated directory structure:

$ tree
.
├── binarygap
│ └── binary_gap.py
└── twosum
├── test_twosum.py
└── twosum.py

Updated twosum.py:

def solution(nums, target):
    num_dict = {}
    for i, num in enumerate(nums):
        complement = target - num
        if complement in num_dict:
            return [num_dict[complement], i]
        num_dict[num] = i
    return None  # Return None if no solution is found

# Test cases
if __name__ == "__main__":
    nums1 = [3, 2, 4]
    target1 = 6
    print("Test case 1:", solution(nums1, target1))
...

test_twosum.py:

import unittest
from twosum import solution

class TestTwoSum(unittest.TestCase):
    def test_basic_case(self):
        self.assertEqual(solution([2, 7, 11, 15], 9), [0, 1])

    def test_same_numbers(self):
        self.assertEqual(solution([3, 3], 6), [0, 1])
    ...

if __name__ == '__main__':
    unittest.main()

TODO

• Implement code auto-formatting (currently leveraging code editor auto-format on save).
• Add support for more languages (currently supports Go, Python, and JavaScript).