Knowledge Representation

A propositional logic engine for representing and querying a knowledge base. Sentences are built from symbols and logical connectives; the model_check function determines whether a knowledge base entails a query by exhaustive truth-table enumeration.

Functions

logic.py

Classes

• EvaluationException – Raised when a symbol is evaluated against a model that does not contain it.
• Sentence (abstract) – Base class for all logical sentences. Defines evaluate(model), formula(), and symbols().
• Symbol(name) – A propositional variable. evaluate returns bool(model[name]) or raises EvaluationException.
• Not(operand) – Logical negation. formula renders as ¬.
• And(*conjuncts) – Logical conjunction. add(conjunct) appends a new conjunct. formula renders operands joined by ∧.
• Or(*disjuncts) – Logical disjunction. add(disjunct) appends a new disjunct. formula renders operands joined by ∨.
• Implication(antecedent, consequent) – Material implication. formula renders as antecedent => consequent.
• Biconditional(left, right) – Logical biconditional. formula renders as left <=> right.

Functions

• model_check(knowledge, query) – Returns True if knowledge entails query under all possible truth assignments to the symbols involved.

Installation

Install the required dependencies:

pip install -r requirements.txt

Usage

Quick run (Harry Potter puzzle)

Windows:

python harry.py

Linux/macOS:

python3 harry.py

Build a knowledge base in Python

from logic import And, Not, Or, Implication, Biconditional, Symbol, model_check

rain = Symbol("rain")
sun  = Symbol("sun")

knowledge = And(
    Or(rain, sun),
    Implication(rain, Not(sun)),
)

print(model_check(knowledge, sun))   # True
print(model_check(knowledge, rain))  # False

Testing

Windows

python -m pytest

Run a specific file:

python -m pytest tests/test_logic.py

Run a specific test:

python -m pytest tests/test_logic.py::test_function_name

Run tests matching a pattern:

python -m pytest -k "symbol"

Run without coverage (faster):

python -m pytest --no-cov

Linux/macOS

pytest

Contributing

Short contributions are welcome and appreciated.

1. Fork the repository.
2. Create a branch for your change.
3. Make your update and keep it focused.
4. Run tests before submitting:
   • Windows: python -m pytest
   • Linux/macOS: pytest
5. Open a Pull Request with a clear description of what and why.

Please keep PRs small, readable, and related to a single improvement.