Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/mattgreen/hython
Haskell-powered Python 3 interpreter
https://github.com/mattgreen/hython
haskell interpreter language python
Last synced: 3 days ago
JSON representation
Haskell-powered Python 3 interpreter
- Host: GitHub
- URL: https://github.com/mattgreen/hython
- Owner: mattgreen
- License: gpl-3.0
- Created: 2014-07-16T19:11:46.000Z (over 10 years ago)
- Default Branch: master
- Last Pushed: 2017-08-01T00:18:19.000Z (over 7 years ago)
- Last Synced: 2024-10-06T04:07:00.697Z (3 months ago)
- Topics: haskell, interpreter, language, python
- Language: Haskell
- Homepage:
- Size: 549 KB
- Stars: 580
- Watchers: 23
- Forks: 29
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.md
Awesome Lists containing this project
README
# hython
A toy Python 3 interpreter implemented in Haskell.
## Introduction
I wanted to learn Haskell, and I wanted a big project, so I decided to write a Python 3 interpreter. The result was extremely educational and easily the coolest project I've ever worked on. Because it's implemented in a naive fashion, it won't ever be a replacement for real Python implementations.
**Note:** Hython only implements most of the Python3 _language_. It doesn't contain much of a standard library, which is a big part of what makes Python pleasant to use. Adding all of the necessary machinery needed for the existing Python 3 standard library to function is an enormous undertaking that I'm not interested in.
## Status
It's finally done! Or at least, I'm declaring it that way.
## Features
* [x] Lexer
* [x] Parser
* [x] Most built-in data types, including `int`, `bool`, `string`, `list`, `dict` and `range`
* [x] Common unary and binary operators on common data types
* [x] A few built-in functions, including `print`
* [x] Variable assignment and lookup, with support for `nonlocal` and `global` keywords
* [x] Conditional expressions with `if` and `else`
* [x] All loop constructs: `for` and `while` with support for `break` and `continue` within them
* [x] Support for the `with` statement
* [x] Destructuring (`(a,b) = [1,2]`)
* [x] Functions, including nested functions, default parameters, and keyword parameters
* [x] Splat (`*` and `**`) operators in caller argument lists
* [x] Lambda expressions, with proper environment capture
* [x] Classes, including inheritance and proper method resolution order
* [x] Objects
* [x] Exception handling via `try`, with support for handlers, frame unwinding, `finally` handlers, and `else`, along with some built-in exception classes
* [x] Basic support for loading modules with the `import` statement
* [x] Simple REPL
* [x] Support for the `is` operator
* [ ] Support for generators and `yield`
* [ ] List/generator/dict/set comprehensions
* [ ] Index slicing
* [ ] Support for decorators / metaclasses
* [ ] Multi-line input for the REPL
## Code Metrics
`sloccount` output as of 10/1/16:
Totals grouped by language (dominant language first):
haskell: 2159 (70.83%)
yacc: 580 (19.03%) # parser
python: 309 (10.14%) # lib
## Examples
See the [test](https://github.com/mattgreen/hython/tree/master/test) directory for example code that works
## Building and running
1. Install [Stack](https://github.com/commercialhaskell/stack)
2. Clone the repository:
$ git clone https://github.com/mattgreen/hython.git
$ cd hython
3. Build:
$ make
4. Run a file:
$ ./hython test/fib.py
## REPL
Hython includes a simple REPL, which you can play around with:
$ ./hython
## Test Suite
Hython's test suite is rather simple: it ensures the output of Hython matches that of the system's `python3` for each test file.
To run the automated test suite:
$ make test
## Reference Information
* [Python 3.4 Language Reference](https://docs.python.org/3.4/reference/)
* [Alex + Happy Whitespace Handling](https://github.com/jmoy/alexhappy)
* [berp](https://github.com/bjpop/berp)
* [Language.Python](https://github.com/bjpop/language-python)