Code Quality Analyzer with AI

A comprehensive AI-powered code quality analyzer with multi-language support. This tool analyzes code quality across multiple programming languages using:

• Multi-language parsing to extract code metrics from Python, JavaScript, TypeScript, Java, C++, Go, Rust, Ruby, and PHP
• Rule-based detectors for code smells (long functions, unused imports, deep nesting, complexity)
• Language-specific linters integration (flake8, pylint, ESLint, Checkstyle, cppcheck, golangci-lint, clippy, RuboCop, PHP_CodeSniffer)
• ML classifier using scikit-learn to classify code snippets as "good" or "bad"
• CLI to analyze files and Flask web UI to upload and analyze code with autofix options
• Sample dataset and comprehensive unit tests

Supported Languages

Language	Parser Type	Linter Integration	Status
Python	AST	flake8, pylint	✅ Full
JavaScript/TS	Regex	ESLint	✅ Full
Java	Regex	Checkstyle, PMD	✅ Full
C/C++	Regex	cppcheck, clang-tidy	✅ Full
Go	Regex	golangci-lint	✅ Full
Rust	Regex	clippy	✅ Full
Ruby	Regex	RuboCop	✅ Full
PHP	Regex	PHP_CodeSniffer	✅ Full

Quickstart

1. Create and activate a Python environment (>= 3.9 recommended).
2. Install requirements:

pip install -r requirements.txt

3. (Optional) Install language-specific linters for enhanced analysis:

# JavaScript/TypeScript
npm install -g eslint

# Java (Ubuntu/Debian)
apt-get install checkstyle

# C/C++ (Ubuntu/Debian)
apt-get install cppcheck

# Go
go install github.com/golangci/golangci-lint/cmd/golangci-lint@latest

# Rust
rustup component add clippy

# Ruby
gem install rubocop

# PHP
composer global require squizlabs/php_codesniffer

4. Train a small demo model (optional):

python -m code_quality_analyzer.cli train --dataset datasets/synthetic_dataset.csv --model-out models/code_quality_model.joblib

5. Analyze any supported file:

# Python
python -m code_quality_analyzer.cli analyze --file examples/bad_example.py --model models/code_quality_model.joblib

# JavaScript
python -m code_quality_analyzer.cli analyze --file app.js --model models/code_quality_model.joblib

# Java
python -m code_quality_analyzer.cli analyze --file Main.java --model models/code_quality_model.joblib

# Any supported language (auto-detected by extension)
python -m code_quality_analyzer.cli analyze --file <your-file> --model models/code_quality_model.joblib

6. Run the web app:

python -m code_quality_analyzer.webapp

Project Structure

• code_quality_analyzer/ - main package
  • parser.py - Multi-language parsing and feature extraction (Python AST, regex-based for others)
  • detectors.py - Rule-based detectors and linter integrations for all supported languages
  • ml_classifier.py - Model training and prediction helpers
  • suggestion_engine.py - Auto-fix suggestions
  • cli.py - CLI wrapper with multi-language support
  • webapp.py - Flask demo for uploading and analyzing code
• datasets/ - Small synthetic datasets
• examples/ - Sample files in various languages
• tests/ - Unit tests
• requirements.txt - Required Python packages

Multi-Language Analysis Examples

Python Analysis

from code_quality_analyzer.parser import extract_features_from_file
from code_quality_analyzer.detectors import RuleBasedDetector

features = extract_features_from_file('script.py')
detector = RuleBasedDetector()
issues = detector.detect_all_languages(source_code, 'python')

JavaScript/TypeScript Analysis

features = extract_features_from_file('app.js')
issues = detector.detect_all_languages(source_code, 'javascript')

Java Analysis

features = extract_features_from_file('Main.java')
issues = detector.detect_all_languages(source_code, 'java')

Language detection is automatic based on file extension.

Notes

This project supports comprehensive multi-language analysis. You can extend it with:

• Tree-sitter integration for more precise AST parsing across all languages
• Better ML dataset with language-specific embeddings
• Enhanced auto-fix engine using language-specific AST rewriters
• CI integration for scanning across multi-language repositories
• Custom rule definitions per language

Contributions & improvements are welcome.

Hosting / Deployment

This repo contains a Dockerfile and docker-compose.yml to run the Flask web app behind Gunicorn. The container exposes port 5000.

Build and run locally with Docker:

docker build -t code-quality-analyzer:latest .
docker run --rm -p 5000:5000 -v ${PWD}\models:/app/models -e MODEL_PATH=/app/models/code_quality_model.joblib code-quality-analyzer:latest

Using Docker Compose:

docker-compose up --build

Heroku / PaaS (example): create an app, set a BUILDPACK or use the Docker container; the included Procfile runs Gunicorn for you.

heroku create
git push heroku main
heroku ps:scale web=1

Note: Ensure model file models/code_quality_model.joblib is included or set the MODEL_PATH environment variable pointing to the model artifact.

Production Deployment with GitHub Container Registry (Recommended)

Best deployment method - No size limits, full multi-language support, ML features included.

Automatic Deployment with GitHub Actions

The project includes automated CI/CD that builds and publishes Docker images to GitHub Container Registry (GHCR) on every push to main:

1. GitHub Actions automatically:

   • Builds multi-architecture Docker image (linux/amd64, linux/arm64)
   • Pushes to ghcr.io/<YOUR_USERNAME>/code-quality-analyzer:latest
   • Optionally uploads ML models to S3 (if configured)
   • Uses build cache for faster subsequent builds

2. Pull and run the pre-built image:

# Pull latest image from GHCR
docker pull ghcr.io/shahinshac/code-quality-analyzer:latest

# Run the container
docker run -d -p 5000:5000 \
  -e MODEL_PATH=/app/models/code_quality_model.joblib \
  ghcr.io/shahinshac/code-quality-analyzer:latest

3. Deploy to any cloud platform:
   • Railway: Connect GitHub repo, auto-deploys from GHCR
   • Render: Deploy from Docker registry
   • Fly.io: fly launch from Dockerfile
   • Google Cloud Run: Deploy from container registry
   • AWS ECS/Fargate: Use GHCR image
   • Azure Container Instances: Pull from GHCR

Manual Build and Push (Optional)

# Build locally
docker build -t ghcr.io/<YOUR_USERNAME>/code-quality-analyzer:latest .

# Login to GHCR
echo <GITHUB_TOKEN> | docker login ghcr.io -u <YOUR_USERNAME> --password-stdin

# Push to GHCR
docker push ghcr.io/<YOUR_USERNAME>/code-quality-analyzer:latest

Optional: S3 Model Storage

Configure these GitHub repository secrets/variables to enable S3 model upload:

Secrets:

• AWS_ACCESS_KEY_ID - AWS access key
• AWS_SECRET_ACCESS_KEY - AWS secret key

Variables:

• AWS_REGION - AWS region (default: us-east-1)
• AWS_S3_BUCKET - S3 bucket name for model storage

The workflow will automatically upload models to S3 and generate presigned URLs for download.

Contributions & improvements are welcome.