dantemazza / boolean-solver

Flask app that uses Quine-McCluskey/Petrick's algorithms to find least-cost boolean expressions

www.booleansolver.com

BooleanSolver

A python implementation of the Quine-McCluskey algorithm and Petrick's method, inspired by the content of ECE124 and my own independent research.

Project Setup

python3 -m venv venv
source venv/bin/activate
pip3 install -r requirements.txt

Running

Here is a sample test case expressed as a minterm expansion:

8
m(0,1,2,5,6,13,14,15,137,124,198,199)
d(3,11)

Here is a maxterm expansion:

6
M(0,1,2,5,6,35,36,39,41,45,49,62,63)
d(3,11,50,51,54,55,56,57)

The case should be formatted as such:

• The first line is the amount of boolean inputs. Rather than writing, for example, f(a,b,c,d), we would put a 4.
• The second line is the max/min expansion. It is important that you include brackets, no spaces, and seperate each value with a comma
• The third line is dontcares, formatted in the same way as the second line. This is optional.

To run a test case, use the invoke task

inv solve <test_case_path>

or use the flask app!

python3 boolean_solver/__init__.py

and connect to http://localhost:5000

About

The program uses a prime implicant chart (as specified in the tabulation method) to locate all essential prime implicants. The remaining ones are covered using a Petrick's method expansion. Algebraic expressions are stored in list data structures as follows:

<class 'list'>: [[[0], [2]], [1, 3], [2, 5], [3, 6], [5, 6, 7]]

Inner-most lists are products. Lists of the innermost lists are sums. Lists of these lists are products and so on. The program represents our variables as numbers here; assuming we map each number to its alphabetical letter (i.e 0->a, 1->b, etc.), the equivalent boolean expression would be: (a + c)(bd + cf + dg + fgh)

boolean-solver's Issues

Add documentation

you need to add a README.md file so that viewers of this repository know what they're looking at! Additionally, you should include details on how to set up your repo, testing, etc.

need to improve runtime

test_case2 cannot finish, presumably because the Petrick's implementation is quite brute force and slow. Need to look into this

need to format prime implicant chart properly for functions of 7+ variables