damj is a tool for creating comprehensive prompts for language models by combining project files and applying custom processing options.
damj
is designed to help developers create effective prompts for large language models (LLMs) such as ChatGPT. By combining different project files and applying customizable processing options, damj
simplifies the process of generating prompts tailored to specific project contexts.
- Combine multiple project files into a single prompt.
- Apply customizable processing options to include/exclude comments, imports, and docstrings for python scripts.
- Support for processing Jupyter notebooks and including cell outputs.
- Easy integration and usage within Python projects.
You can install the latest release from PyPI:
pip install damj
git clone https://github.com/baselhusam/damj.git
cd damj
pip install .
import os
from damj import Damj
cwd = os.getcwd()
damj = Damj(cwd)
damj.project_info(
project_overview="This is a sample project.",
add_project_structure=True,
)
prompt = damj.create_prompt(
question="What is the purpose of this project?",
)
print(prompt)
Output:
# Project Overview
This is a sample project.
# Project Structure
| ├── LICENSE
| ├── README.md
| ├── pyproject.toml
| ├── requirements.txt
├── assets/
| ├── background.png
| ├── logo.png
├── damj/
| ├── __init__.py
| ├── damj.py
| ├── utils.py
# Question
What is the purpose of this project?
import os
from damj import Damj
cwd = os.getcwd()
damj = Damj(
cwd=cwd,
whitelist_files=["*.py"],
blacklist_files=[".venv", "__pycache__"],
snippet_marker="```"
)
damj.project_info(
project_overview="This is a sample project.",
add_project_structure=True,
add_files_content=True,
py_options={
"add_imports": True,
"add_comments": True,
"add_docstrings": False,
"ipynb_output": False
}
)
prompt = damj.create_prompt(
question="What is the purpose of this project?",
copy_to_clipboard=True,
to_markdown=False
)
print(prompt)
damj also provides several utility functions that can be used independently. These utilities include functions to get the project structure, get file content, and more.
The get_project_structure
function generates a markdown representation of the directory structure, excluding blacklisted files and directories.
from damj.utils import get_project_structure
# Get the project structure excluding .venv and __pycache__ directories
cwd = os.getcwd()
blacklist = [".venv", "__pycache__"]
project_structure = get_project_structure(cwd, blacklist)
print(project_structure)
The get_file_content
function retrieves the content of a file, applying the specified py_options
.
from damj.utils import get_file_content
py_options = {
"add_comments": True,
"add_imports": True,
"add_docstrings": False,
"ipynb_output": False
}
# Get the content of a Python file with the specified options
file_content = get_file_content("example.py", py_options)
print(file_content)
This project is licensed under the Apache Software License. See the LICENSE file for details.
Basel Mather (baselmathar@gmail.com)
Contributions are welcome! Please fork the repository and open a pull request with your changes.