paperprobe 0.2.0

Tool to automatically generate scripts that show usage of scientific code from scientific papers

PaperProbe

Automatically analyze and generate usage examples for scientific code from research papers

Beautiful terminal user interface for analyzing scientific repositories

What is PaperProbe?

PaperProbe bridges the gap between scientific papers and their associated code repositories. It automatically:

• Scans research papers (arXiv URLs or local PDFs) for GitHub repository links
• Analyzes repositories to understand structure, dependencies, and usage patterns
• Generates usage examples using AI-powered code analysis
• Provides insights on repository statistics, code quality, and documentation

Perfect for researchers, developers, and students who want to quickly understand and utilize code from scientific publications.

Features

Interactive TUI

• Beautiful terminal interface built with Textual
• Real-time progress updates and visual feedback
• Keyboard shortcuts for efficient navigation

Multi-Source Support

• arXiv papers: Direct URL scanning
• Local PDFs: Parse papers from your filesystem
• GitHub URLs: Direct repository analysis

AI-Powered Analysis

• Intelligent code structure analysis
• Automated usage example generation
• Context-aware documentation extraction

Repository Insights

• GitHub statistics (stars, forks, contributors)
• Dependency analysis
• Code quality metrics
• File structure visualization

Installation

Using pipx (Recommended)

pipx install paperprobe

Using pip

pip install paperprobe

From Source

git clone https://github.com/Brook-B-Nigatu/PaperProbe.git
cd PaperProbe
pip install -e .

[!TIP] Package version may need to be specified. pipx install paperprobe==0.2.0

Prerequisites

• Python: 3.12 or higher (some dependencies don't work on Python v3.13)
• API Keys: AI API key for AI-powered analysis
• GitHub Token (optional): For enhanced GitHub API access

Configuration

Environment Variables

PaperProbe requires API keys to function.

Environment Variables (Recommended)

Add to your shell configuration file (~/.zshrc, ~/.bashrc, etc.):

export CONSTRUCTOR_KM_ID=""
export CONSTRUCTOR_API_KEY=""
export CONSTRUCTOR_API_URL=""
export GITHUB_TOKEN="" #optional

Then reload your shell:

source ~/.zshrc

Usage

Launch the TUI

paperprobe

Basic Workflow

1. Enter a source:

   • arXiv URL: https://arxiv.org/abs/2301.12345
   • Local PDF: /path/to/paper.pdf
   • GitHub URL: https://github.com/username/repo

2. Scan for repositories: PaperProbe extracts GitHub links from papers

3. Select a repository: Choose from the discovered repos

4. Choose analysis mode:

   • Basic: Quick repository overview
   • Deep: Comprehensive analysis with usage examples

5. View results: Interactive markdown display with insights

Keyboard Shortcuts

Shortcut	Action
Ctrl+S	Use sample URL
Ctrl+Q	Quit application
Enter	Submit input / Select item
↑/↓	Navigate lists

Sample Usage

Try the built-in sample by pressing Ctrl+S on the intro screen, or paste this URL:

https://github.com/Brook-B-Nigatu/PaperProbe

Project Structure

PaperProbe/
├── src/
│   ├── core/               # Core analysis logic
│   │   ├── llm_service.py  # AI service integration
│   │   └── task_manager.py # Async task orchestration
│   ├── github_repo/        # GitHub repository handling
│   ├── preprocessing_utilities/
│   │   └── pdf_parser.py   # PDF text extraction
│   ├── tool_providers/     # Analysis tool providers
│   └── ui/                 # Terminal UI components
│       ├── app.py          # Main TUI application
│       ├── controller.py   # Business logic
│       └── style.tcss      # TUI styling
├── pyproject.toml          # Project configuration
└── README.md

Development

Setup Development Environment

# Clone the repository
git clone https://github.com/Brook-B-Nigatu/PaperProbe.git
cd PaperProbe

# Install in editable mode
pip install -e .

# Or use uv for faster installs
uv pip install -e .

Running Tests

# Run linter
ruff check .

# Format code
ruff format .

License

This project is licensed under the MIT License - see the LICENSE.txt file for details.

Acknowledgments

• Built with Textual for the beautiful TUI
• Powered by LangChain and OpenAI for intelligent analysis
• PDF parsing with pdfminer.six

---

_{Made with ❤️ for the research community}