a basic bytecode to machine code compiler

Steven D'Aprano steve+comp.lang.python at pearwood.info
Fri Apr 1 08:57:11 EDT 2011


On Thu, 31 Mar 2011 18:33:36 -0400, Rouslan Korneychuk wrote:

> I'm aware that PyPy already has a working JIT compiler, but I figure it
> will be a long time before they have a version of Python that is ready
> for everybody to use, so this could be useful in the mean time.

PyPy is ready to use *now*, if you are happy writing code that targets 
Python 2.5 and don't need C extensions.


[...]
> What do people think? Would I be wasting my time going further with
> this?

Depends on what your ultimate aim is. If it is to learn things yourself, 
then it is never a waste of time to learn new things.

If your aim is to get a good working project that you can be proud to put 
on your CV, then go right ahead.

If your aim is to contribute to a Python compiler that will actually be 
used by people other than yourself, I'm not so sure... personally, I 
expect that PyPy is the future of Python optimizing compilers, but I 
could be wrong.

I suggest you check out the competitors:

Shedskin is a Python to C++ compiler;
Psyco is a JIT specialising compiler;
Nuitka claims to be a C++ implementation that compiles to machine code; 
Berp claims to be a Haskell implementation that does the same;
Compyler claims to be a native x86 assembly compiler;
UnPython claims to be an experimental Python to C compiler.


Of the six, as far as I know only Shedskin and Psyco are widely used. 



Good luck, and remember: 

Release early, release often, and let the community know when you do!


-- 
Steven



More information about the Python-list mailing list