Logical Unification

Logical unification in Python, extensible via dispatch.

Examples

unification has built-in support for most Python data types:

>>> from unification import *
>>> unify(1, 1)
{}
>>> unify(1, 2)
False
>>> x = var('x')
>>> unify((1, x), (1, 2))
{~x: 2}
>>> unify((x, x), (1, 2))
False

Custom classes can be made "unifiable" with the unifiable decorator:

@unifiable
class Account(object):
    def __init__(self, id, name, balance):
        self.id = id
        self.name = name
        self.balance = balance

>>> data = [Account(1, 'Alice', 100),
            Account(2, 'Bob', 0),
            Account(2, 'Charlie', 0),
            Account(2, 'Denis', 400),
            Account(2, 'Edith', 500)]
>>> id, name, balance = var('id'), var('name'), var('balance')
>>> [unify(Account(id, name, balance), acct) for acct in data]
[{~name: 'Alice', ~balance: 100, ~id: 1},
{~name: 'Bob', ~balance: 0, ~id: 2},
{~name: 'Charlie', ~balance: 0, ~id: 2},
{~name: 'Denis', ~balance: 400, ~id: 2},
{~name: 'Edith', ~balance: 500, ~id: 2}]
>>> [unify(Account(id, name, 0), acct) for acct in data]
[False,
{~name: 'Bob', ~id: 2},
{~name: 'Charlie', ~id: 2},
False,
False]

unification also supports function dispatch through pattern matching:

>> from unification.match import *
>>> n = var('n')

@match(0)
def fib(n):
    return 0


@match(1)
def fib(n):
    return 1


@match(n)
def fib(n):
    return fib(n - 1) + fib(n - 2)

>>> map(fib, [0, 1, 2, 3, 4, 5, 6, 7, 8, 0])
[0, 1, 1, 2, 3, 5, 8, 13, 21, 34]

The pattern matching can be fairly complex:

>> name, amount = var('name'), var('amount')

@match({'status': 200, 'data': {'name': name, 'credit': amount}})
def respond(name, amount):
    balance[name] +=  amount


@match({'status': 200, 'data': {'name': name, 'debit': amount}})
def respond(name, amount):
    balance[name] -= amount


@match({'status': 404})
def respond():
    print("Bad Request")

See the full example in the examples directory.

Performance and Reliability

Unification stresses extensibility over performance, preliminary benchmarks show that this is 2-5x slower than straight tuple-based unification.

unification's approach is reliable; although one caveat is set unification, which is challenging to do in general. It should work well in moderately complex cases, but it may break down under very complex ones.

Installation

Using pip:

pip install python-unification

To install from source:

git clone git@github.com:pythological/unification.git
cd unification
pip install -r requirements.txt

Tests can be run with the provided Makefile:

make check

About

This project is a fork of unification.