pylibexplorer 0.1.1

LLM-powered Python library explorer

PyLib Explorer

PyLib Explorer is an LLM-powered tool for exploring Python libraries. This library generates comprehensive README.md files for randomly selected or specified Python packages, helping developers understand new libraries and implement them in their applications.

Features

• 🔍 Random Python library discovery
• 🤖 OpenAI and Claude LLM integrations
• 📚 Detailed README.md file generation
• 🔧 Code examples and usage scenarios
• 📊 Popularity-based filtering
• 🧩 Easy-to-use API and CLI
• 🌍 Multi-language support (create READMEs in any language)

Installation

pip install pylib-explorer

Quick Start

Python API Usage

from pylib_explorer.core import LibExplorer

# You can provide your API key directly or use environment variables
# English is the default language (language="en")
explorer = LibExplorer(api_key="your_api_key_here", provider="openai")

# For README in a different language:
explorer_tr = LibExplorer(api_key="your_api_key_here", provider="openai", language="tr")

# Generate README for a random package
readme_content = explorer.generate_readme()
print(f"README file created: {explorer.output_dir}")

# or for a specific package:
readme_content = explorer.generate_readme(package_name="requests")

Command Line Usage

For a random library exploration:

# Set your OpenAI API key as an environment variable
export OPENAI_API_KEY=your_api_key_here

# Explore a random library (default language: English)
pylib-explorer

# Explore in Turkish
pylib-explorer --language tr

# Explore in German
pylib-explorer --language de

For a specific library exploration:

# English (default)
pylib-explorer --package requests

# Turkish
pylib-explorer --package requests --language tr

Using Claude API:

export ANTHROPIC_API_KEY=your_anthropic_api_key

pylib-explorer --provider claude --package pandas --language fr

Command Line Options

usage: pylib-explorer [-h] [--package PACKAGE] [--provider {openai,claude}]
                      [--api-key API_KEY] [--output-dir OUTPUT_DIR]
                      [--min-downloads MIN_DOWNLOADS] [--language LANGUAGE]
                      [--verbose]

Options:
  -h, --help            Show this help message and exit
  --package PACKAGE, -p PACKAGE
                        Name of the Python package to explore. If not specified, a random package will be selected.
  --provider {openai,claude}, -m {openai,claude}
                        LLM provider to use ('openai' or 'claude').
  --api-key API_KEY, -k API_KEY
                        API key for the LLM provider. If not specified, it will be read from environment variables.
  --output-dir OUTPUT_DIR, -o OUTPUT_DIR
                        Directory where README files will be saved.
  --min-downloads MIN_DOWNLOADS, -d MIN_DOWNLOADS
                        Minimum download count filter for random selection.
  --language LANGUAGE, -l LANGUAGE
                        Language for the README content (e.g.: 'en', 'tr', 'fr', 'de', etc.)
  --verbose, -v         Enable verbose logging mode.

Supported Languages

PyLib Explorer can generate READMEs in any language supported by the LLMs. To specify the language:

• You can use ISO language codes: en, tr, fr, de, es, it, etc.
• Or you can use full language names: English, Turkish, French, German, etc.

English is the default language (en).

How Language Selection Works

PyLib Explorer communicates your language choice to the LLM with clear and strong instructions. Here's how it works:

1. Your language choice (short codes like tr, en) is converted to full language names
2. The prompt sent to the LLM includes:
   • At the beginning: "IMPORTANT: You must prepare this README in {language} language!"
   • At the end: "REMINDER: All content must be in {language} language! This is very important."
3. These strong instructions ensure the LLM produces content in the specified language

If you request a specific language but get content in a different language, please report it through GitHub issues.

Setting Up API Keys

PyLib Explorer requires API keys for either OpenAI or Claude APIs. You can provide these keys in these ways:

1. Environment Variables: Set OPENAI_API_KEY or ANTHROPIC_API_KEY environment variables.
2. Direct Parameter: Provide API keys directly via code or CLI.
3. .env File: Create a .env file in your project's root directory:

   OPENAI_API_KEY=your_openai_api_key
   ANTHROPIC_API_KEY=your_anthropic_api_key
   

Development

Requirements

• Python 3.7+
• OpenAI and/or Anthropic API keys

Installation

git clone https://github.com/username/pylib-explorer.git
cd pylib-explorer
pip install -e ".[dev]"

Running Tests

pytest

License

Licensed under the MIT License. See LICENSE file for more information.

Contributing

Contributions are welcome! Please check out CONTRIBUTING.md.

Contact

For questions or feedback, please use GitHub Issues.