https://github.com/statusfailed/hyperast

https://github.com/statusfailed/hyperast

Last synced: 12 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/statusfailed/hyperast
• Owner: statusfailed
• Created: 2024-04-26T14:57:27.000Z (almost 2 years ago)
• Default Branch: master
• Last Pushed: 2024-05-15T05:36:23.000Z (almost 2 years ago)
• Last Synced: 2025-01-31T08:18:55.217Z (about 1 year ago)
• Language: Python
• Size: 25.4 KB
• Stars: 1
• Watchers: 1
• Forks: 1
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# HyperAST

A (currently experimental) frontend to

[open-hypergraphs](https://github.com/statusfailed/open-hypergraphs/)

which lets you define open hypergraphs in a pointful

(i.e., not [point-free](https://en.wikipedia.org/wiki/Tacit_programming) ) style.

Here's a 1-bit adder circuit:

    @hypergraph

    def adder(x0, x1):

        y0 = x0 + x1 # sum

        y1 = x0 * x1 # carry

        return [y0, y1]

This builds an open hypergraph representing a string diagram like this:

![adder diagram](./propaganda/adder.png)

Unlike python code, hypergraphs are fully combinatorial: there is no implicit

assumption of 'causal' control flow.

For example, we can define a hypergraphs with *feedback*, such as a

[ring oscillator](https://en.wikipedia.org/wiki/Ring_oscillator):

    @hypergraph

    def ring_oscillator():

        x0 = node()

        x1 = ~x0 # we use x0 before declaring it - this is OK!

        unify(x0, ~x1) # connect x0 and x1

        return [x0]

This defines a string diagram like the one below:

![ring oscillator diagram](./propaganda/ring_oscillator.png)

# Alternative interface

There is also a WIP alternative interface which inspects python syntax to define

a hypergraph.

Variables serve as hypernodes while functions and python operators are (labeled)

hyperedges.

 The first example looks the same:

    @hypergraph

    def adder(x0, x1):

        y0 = x0 + x1 # sum

        y1 = x0 * x1 # carry

        return [y0, y1]

... but there is less overhead when writing hypergraphs with more unusual

structure:

    @hypergraph

    def ring_oscillator():

        x1 = ~x0 # we use x0 before declaring it - this is OK

        x0 = ~x1

        return [x0]

Unlike the normal interface, this style does not execute the python code; it

only inspects it.