Finite State Machines in Python

python-fsm module can be used to build and describe deterministic finite-state automata.

Example: Operation of a microwave oven (full list of examples)

from fsm import State, Transducer, get_graph

microwave = Transducer('Microwave Oven')

closed_i = State(r'CLOSED\nidle', initial=True)
closed_p = State(r'CLOSED\nprocessing')
opened_i = State(r'OPEN\nidle')
paused = State('PAUSED')
settings = State(r'SETUP')

closed_i[r'PSB, TK /\nset program,\nset timer'] = closed_i
closed_i['SSB / start'] = closed_p
closed_i['ODH / open door'] = opened_i
closed_i['ELS / enter setup'] = settings
settings['SSB / save setup'] = settings
settings['ELS / leave setup'] = closed_i
opened_i['DL / shut door'] = closed_i
closed_p['PRB / pause'] = paused
closed_p['ODH / open door'] = opened_i
closed_p['TO / ready'] = closed_i
paused['SSB / stop'] = closed_i
paused['PRB / resume'] = closed_p
paused[r'PSB, TK /\nreset program,\nreset timer'] = paused
paused['ODH / open door'] = opened_i

get_graph(microwave).draw('microwave.png', prog='dot')

Overview of implemented types

At the center of the module is the State class representing a state which can be used in a finite state machine of any type. States can be used in the following automata:

FiniteStateMachine

Semiautomaton base for all types that follow. This class implements the process() method which takes an iterator as input and processes it. Semiautomata don't return any values from process() http://en.wikipedia.org/wiki/Semiautomaton

Acceptor

Acceptor's process() will return either True or False depending on whether its final state is classified as accepting or not. http://en.wikipedia.org/wiki/Finite_state_machine#Acceptors_and_recognizers

Transducer

Transducer class extends FiniteStateMachine by implementing an output() method which takes an input value which is meant to be passed to a the current state and returns current state's name. http://en.wikipedia.org/wiki/Finite-state_machine#Transducers

MooreMachine

A specialized transducer. Its output() method returns an output value stored in the current state. http://en.wikipedia.org/wiki/Moore_machine

MealyMachine

another specialized transducer. Its output() method returns a value assigned to the transition that the input value caused.