Lambda the Ultimate

inactiveTopic Static Type Inference (for Python) with Starkiller
started 3/25/2004; 2:39:07 AM - last post 4/1/2004; 12:24:22 AM
Ehud Lamm - Static Type Inference (for Python) with Starkiller  blueArrow
3/25/2004; 2:39:07 AM (reads: 7744, responses: 2)
Static Type Inference (for Python) with Starkiller
Michael Salib. PyCon DC 2004

Removing dynamic dispatch and binding leads to large performance benefits since their existence precludes many traditional optimization techniques,such as inlining. I have built a static type inferencer for Python called Starkiller. Given a Python source file, it can deduce the types of all expressions in the program without actually running it. Starkillerís primary goal is to take a Python source program as input and deduce the information needed to make native code compilation of that program easy. This paper is describes Starkillerís design and operation. It is partially adapted from a draft of my Masterís thesis.

Seems like a nice piece of work. Starkiller even goes as far as providing a mini-language that allows non-Python extension type authors to describe the type behavior of their extensions. These descriptions are used during inference to model the run time data flow.


Posted to Python by Ehud Lamm on 3/25/04; 2:44:33 AM

Ehud Lamm - Re: Static Type Inference (for Python) with Starkiller  blueArrow
3/29/2004; 5:59:35 AM (reads: 149, responses: 1)
Read what Matthias Felleisen has to say.

Isaac Gouy - Re: Static Type Inference (for Python) with Starkiller  blueArrow
4/1/2004; 12:24:22 AM (reads: 106, responses: 0)
Maybe a little type inference is better than no type inference at all?

Applications of Concrete-Type Inference Dec 2003

"We present a concrete-type inference tool for Smalltalk in a practical setting, demonstrating that it can reduce the runtime memory requirements significantly. Specifically, we have implemented our system in the Resilient Embedded System, a development platform for embedded systems."