qdiff

qdiff is a command-line tool and Python library for computing differences between two code quality analysis reports. It helps developers and researchers understand how code quality differs between two versions of a project.

qdiff is designed to work both as:

• ✅ an installable console application
• ✅ a reusable Python package

---

Features

• 🔍 Compare two code quality reports

• 📊 Generate a diff report in a specified output directory

• 🛠️ Supports multiple source tools:

  • dpy – DPy (for Python code analysis)
  • dc – Designite (for C# code analysis)
  • dj – DesigniteJava (for Java code analysis)

• 🔌 Clean Python API for integration into scripts, experiments, and dashboards

• 🤖 CI-friendly command-line interface

---

Installation

Prerequisites

• Python 3.9 or higher
• pip

---

Install locally (development mode)

From the project root directory:

pip install -e .

This installs qdiff in editable mode, allowing you to modify the source code without reinstalling.

Verify installation:

qdiff --help

---

Usage (Command Line)

Basic syntax

qdiff path1 path2 path3 tool

Arguments

Argument	Description
path1	Path to the first code quality report directory
path2	Path to the second code quality report directory
path3	Path to the output directory for the diff report
tool	Source tool that generated the reports (dpy, dc, or dj)

---

Example

qdiff reports/v1 reports/v2 reports/diff dpy

This compares:

• report from reports/v1
• report from reports/v2
• writes the diff to reports/diff
• assumes reports were generated by DPy

---

Help

qdiff --help

This displays full usage instructions and available options.

---

Usage as a Python Library

You can import and use qdiff directly in your Python code.

Example

from qdiff.smell_diff import SmellDiff
# The parameters define the columns in the dataframe to be treated as key and description columns; the description column can be passed an empty list
smell_diff = SmellDiff(['Project', 'Package', 'Class', 'Smell'],['Description'])
# read_csv is a simple function that read csv files using pandas library and return corresponding dataframes
df_old, df_new = read_csv('/path/to/ver1/DesignSmells.csv',
                           '/path/to/ver2/DesignSmells.csv')
diff_df = smell_diff.diff(df_old, df_new)

---

Building the Package (Optional)

If you want to create distributable artifacts (wheel + source archive):

Install build tool

python -m pip install build

Build

python -m build

Artifacts will be created in:

dist/