https://github.com/tusharsadhwani/packaged
The easiest way to ship python applications.
https://github.com/tusharsadhwani/packaged
packaging python
Last synced: 6 months ago
JSON representation
The easiest way to ship python applications.
- Host: GitHub
- URL: https://github.com/tusharsadhwani/packaged
- Owner: tusharsadhwani
- License: gpl-2.0
- Created: 2024-05-08T21:46:44.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-06-28T11:25:57.000Z (over 1 year ago)
- Last Synced: 2024-11-25T09:17:47.228Z (11 months ago)
- Topics: packaging, python
- Language: Shell
- Homepage: https://packaged.live
- Size: 237 KB
- Stars: 193
- Watchers: 3
- Forks: 8
- Open Issues: 5
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# packaged
The easiest way to ship python applications.
### Demo apps available on the website: [packaged.live](https://packaged.live)
`packaged` can take any Python project, and package it into a self contained
executable, that can run on other machines without needing Python installed.
## Installation
```bash
pip install packaged
```
## Usage
```bash
packaged [] [--python-version=3.12]
```
Such as:
```bash
packaged my_project.sh 'pip install .' 'python -m your_package' path/to/project
```
This will package Python 3.12 with your application by default.
To specify a different Python version, use the `--python-version` flag, like so:
```bash
packaged my_project.sh 'pip install .' 'python -m your_package' path/to/project --python-version=3.10
```
## Examples
All examples below create a self contained executable. You can send the produced
binary file to another machine with the same OS and architecture, and it will
run the same.
You can also find the pre-built binaries on the [Releases page](https://github.com/tusharsadhwani/packaged/releases/latest).
### Mandelbrot (`numpy`, `matplotlib`, GUI)
```bash
packaged ./mandelbrot.sh 'pip install -r requirements.txt' 'python mandelbrot.py' ./example/mandelbrot --python-version=3.10
```
This produces a `./mandelbrot.sh` binary with:
- Python 3.10
- `matplotlib`
- `numba`
- `llvmlite`
- `pillow`
That outputs an interactive mandelbrot set GUI.
### Minesweeper (using `packaged.toml` for configuration)
You can use a `packaged.toml` file and simply do `packaged path/to/project` to
create your package. For example, try the `minesweeper` project:
```bash
packaged ./example/minesweeper
```
[This configuration](https://github.com/tusharsadhwani/packaged/blob/main/example/minesweeper/packaged.toml)
is used for building the package. The equivalent command to build the project
without `pyproject.toml` would be:
```bash
packaged minesweeper.sh 'pip install .' 'python -m minesweeper' ./example/minesweeper
```
### Posting (TUI based API testing app)
Posting is a Postman alternative that runs entirely in the terminal. A perfect
candidate to build an executable out of!
Since the dependencies themselves contain all the source code needed, you can
skip the last argument. With this, no other files will be packaged other than
what is produced in the build step.
```bash
packaged ./posting.sh 'pip install posting' 'posting'
```
This will simply package the `posting` app into a single file.
### Aliens (pygame)
Pygame ships with various games as well, `pygame.examples.aliens` is one of them:
```bash
packaged ./aliens 'pip install pygame' 'python -m pygame.examples.aliens'
```
Another one that you can try out is `pygame.examples.chimp`.
### IPython (console scripts)
Packages that expose shell scripts (like `ipython`) should also just work when
creating a package, and these scripts can be used as the startup command:
```bash
packaged ./ipython 'pip install ipython' 'ipython'
```
Now running `./ipython` runs a portable version of IPython!
## Local Development / Testing
To test and modify the package locally:
- Create and activate a virtual environment
- Run `pip install -r requirements-dev.txt` to do an editable install
- Run `pytest` to run tests
- Make changes as needed
### Type Checking
Run `mypy .`
### Create and upload a package to PyPI
Make sure to bump the version in `setup.cfg`.
Then run the following commands:
```bash
rm -rf build dist
python setup.py sdist bdist_wheel
```
Then upload it to PyPI using [twine](https://twine.readthedocs.io/en/latest/#installation):
```bash
twine upload dist/*
```
## License
The package is Licensed under GNU General Public License v2 (GPLv2). However,
note that **the packages created with `packaged` are NOT licensed under GPL**.
This is because the archives created are just data for the package, and
`packaged` is not a part of the archives created.
That means that you can freely use `packaged` for commercial use.
Read the [License section for Makeself](https://github.com/megastep/makeself?tab=readme-ov-file#license) for more information.