Codemetrica

A Python library for calculating code metrics and detecting code smells in Python codebases using tree-sitter.

Features

• Comprehensive code metrics calculation
• Code smell detection
• Support for analyzing Python files, classes, and methods
• Easy-to-use API
• Powered by tree-sitter for robust parsing

Prerequisites

• Python 3.10 or higher
• GCC compiler (for building tree-sitter grammar)
• Git (for cloning tree-sitter repository)

Installation

# Clone the repository
git clone https://github.com/yourusername/codemetrica.git
cd codemetrica

# Install the package
pip install -e .

The installation process will:

1. Clone the tree-sitter-python repository
2. Build the tree-sitter grammar
3. Install the package in development mode

Quick Start

from codemetrica import File, GodClass

# Analyze a Python file
f = File("example.py")

# Get all classes in the file
classes = f.get_classes()

# Check for code smells
for class_obj in classes:
    if GodClass.detect(class_obj):
        print(f"Found God Class: {class_obj.name}")

Available Metrics

Size Metrics

• Lines of Code (LOC)
• Source Lines of Code (SLOC)
• Kilo Lines of Code (KLOC)
• Comment Lines of Code (CLOC)
• Number of Methods (NOM)
• Number of Classes (NOC)
• Number of Fields (NOF)
• Number of Public Methods (NOPM)

Complexity Metrics

• Cyclomatic Complexity (CC)
• Weighted Methods per Class (WMC)
• NPath Complexity
• Essential Complexity
• Halstead Metrics
• Nested Block Depth (NBD)
• Decision Density

Object-Oriented Metrics

• Depth of Inheritance Tree (DIT)
• Number of Children (NOC)
• Response for a Class (RFC)
• Coupling Between Objects (CBO)
• Lack of Cohesion of Methods (LCOM)
• And more...

Available Code Smells

General Smells

• Broken Modularization
• Complex Conditional
• Deficient Encapsulation
• Dispersed Coupling
• Empty catch clause
• And more...

Class Level Smells

• Brain Class
• Class Data Should Be Private
• Complex Class
• Data Class
• God Class
• Large Class
• Lazy Class

Method Level Smells

• Abstract Function Call From Constructor
• Brain Method
• Complex Method
• Long Method
• Long Parameter List

Development

To build the tree-sitter grammar manually:

# Clone tree-sitter-python if not already done
git clone https://github.com/tree-sitter/tree-sitter-python.git

# Build the grammar
gcc -shared -o build/my-languages.so -I./tree-sitter-python/src tree-sitter-python/src/parser.c -fPIC

License

MIT License