Python to CLike language transpiler.
Project description
py2many: Python to many CLike languages transpiler
Why
Python is popular, easy to program in, but has poor runtime performance. We can fix that by transpiling a subset of the language into a more performant, statically typed language.
A second benefit is security. Writing security sensitive code in a low level language like C is error prone and could lead to privilege escalation. Specialized languages such as wuffs exist to address this use case. py2many can be a more general purpose solution to the problem where you can verify the source via unit tests before you transpile.
A third potential use case is to accelerate python code by transpiling it into an extension
Swift and Kotlin dominate the mobile app development workflow. However, there is no one solution that works well for lower level libraries where there is desire to share code between platforms. Kotlin Mobile Multiplatform (KMM) is a player in this place, but it hasn't really caught on. py2many provides an alternative.
Lastly, it's a great educational tool to learn a new language by implementing a backend for your favorite language.
Status
Rust is the language where the focus of development has been.
C++14 is historically the first language to be supported. C++17 is now required for some features.
Preliminary support exists for Julia, Kotlin, Nim, Go, Dart, V, and D.
py2many can also emit Python 3 code that includes inferred type annotations, and revisions to the syntax intended to simplify parsing of the code.
History
Based on Julian Konchunas' pyrs.
Based on Lukas Martinelli Py14 and Py14/python-3 branch by Valentin Lorentz.
Example
Original Python version.
def fib(i: int) -> int:
if i == 0 or i == 1:
return 1
return fib(i - 1) + fib(i - 2)
Transpiled Rust code:
fn fib(i: i32) -> i32 {
if i == 0 || i == 1 {
return 1;
}
return (fib((i - 1)) + fib((i - 2)));
}
Transpiled code for other languages:
https://github.com/adsharma/py2many/tree/main/tests/expected (fib*)
Trying it out
Requirements:
- Python 3.8+
Local installation:
pip3 install --user # installs to $HOME/.local
OR
sudo pip3 install # installs systemwide
Add the py2many script to your $PATH and run:
Transpiling:
py2many --cpp=1 tests/cases/fib.py
py2many --rust=1 tests/cases/fib.py
py2many --julia=1 tests/cases/fib.py
py2many --kotlin=1 tests/cases/fib.py
py2many --nim=1 tests/cases/fib.py
py2many --dart=1 tests/cases/fib.py
py2many --go=1 tests/cases/fib.py
py2many --dlang=1 tests/cases/fib.py
Compiling:
clang tests/expected/fib.cpp
rustup run nightly cargo build -Zscript --manifest-path tests/expected/fib.rs
...
dmd -run tests/cases/fib.d
Many of the transpilers rely on a language specific formatter to parse the output and reformat it.
Typically this is the most prominent formatter for the language, such as rustfmt
for Rust.
Most of the transpilers also rely on external libraries to provide bridges from Python constructs to the target language.
The steps to install these external libraries can be found in .github/workflows/main.yml
.
Contributing
See CONTRIBUTING.md for how to test your changes and contribute to this project.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file py2many-0.5.1.tar.gz
.
File metadata
- Download URL: py2many-0.5.1.tar.gz
- Upload date:
- Size: 179.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.1.1 CPython/3.12.3
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 715183808172e7e08d6a49411167dfa0a2fdd79a4bfc1b88349f1f778be3cc5e |
|
MD5 | 1035dafc4a382fcff5674478f0723404 |
|
BLAKE2b-256 | 8bcc40c32a97078f16786573f1451338fa3e516fd27ff4497705d554416404db |
File details
Details for the file py2many-0.5.1-py3-none-any.whl
.
File metadata
- Download URL: py2many-0.5.1-py3-none-any.whl
- Upload date:
- Size: 135.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/5.1.1 CPython/3.12.3
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 4a1f4d7ce6eb8cb1eb2fde74799d457371113bc901e3a9795ccd3eb06352cf8d |
|
MD5 | 3213cbc079168d039fd9f9ba05224443 |
|
BLAKE2b-256 | 0f5c626cb2bc7b0ce610a4ad7a9bc9761b44d3014f0f918ee3370efe3bc868c9 |