repo2string 0.2.0

Copy any repository into the clipboard. And get the token count.

repo2string

repo2string is a Python package and CLI tool that gathers all files in a repository (or any folder), excluding ignored files as specified by a .gitignore (if present), and concatenates them into a single string. This is useful for copying the entire codebase as a context to large language models (LLMs) like ChatGPT.

Features:

• Recursively traverse directories.
• Skip files listed in .gitignore (if present) or skip only .git if no .gitignore exists.
• Generate a file tree (with absolute paths).
• Include the contents of all non-ignored files.
• Copy all text to your clipboard automatically.
• Token counting: Displays the token count of the entire prompt.
• Verbose mode (-v or --verbose): Also prints the token counts per file, sorted from highest to lowest.

Installation

You can install repo2string directly from PyPI:

pip install repo2string

Or install from source:

git clone https://github.com/szulcmaciej/repo2string.git
cd repo2string
pip install .

Development Setup

To set up the development environment:

1. Clone the repository:

   git clone https://github.com/szulcmaciej/repo2string.git
   cd repo2string
   

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   

3. Install the package in editable mode with development dependencies:

   pip install -e ".[test]"
   

4. Install pre-commit hooks:

   pre-commit install
   

The pre-commit hooks will:

• Run Ruff for linting and auto-formatting
• Run pytest to ensure all tests pass
• Block commits if any checks fail

To run tests manually:

pytest

Usage

repo2string [PATH] [--verbose]

• PATH is optional; defaults to . (current directory).
• --verbose or -v prints a token-count summary per file (descending).

Example:

repo2string /path/to/myproject --verbose

You will see console output summarizing the total token count, plus a per-file token breakdown if in verbose mode. The entire text is copied to your clipboard.

Now you can paste the combined repo data into ChatGPT or another LLM interface to work on your code with maximum context.

License

MIT