def test_or(self):
ast = negation_normal_form(parse_ctlq("AF ('admin = bob' | AG ?)"))
self.assertTrue(check_ctlqx(ast))
fsm = self.init_model()
solution = {HashableDict({'admin': 'alice', 'state': 'waiting'}),
HashableDict({'admin': 'alice', 'state': 'processing'})}
self.assertCountEqual(bdd_to_set(fsm, solve_ctlqx(fsm, ast)), solution)
def test_bdd_to_set(self):
set_ = {HashableDict({'state': 'ready', 'request': 'TRUE'}),
HashableDict({'state': 'ready', 'request': 'FALSE'})}
init_nusmv()
fsm = self.init_model()
self.assertEqual(bdd_to_set(fsm, fsm.init), set_)
deinit_nusmv()
def test_au(self):
ast = negation_normal_form(parse_ctlq("A[? U 'state = processing']"))
self.assertTrue(check_ctlqx(ast))
fsm = self.init_model()
solution = {HashableDict({'admin': 'none', 'state': 'starting'}),
HashableDict({'admin': 'none', 'state': 'choosing'}),
HashableDict({'admin': 'alice', 'state': 'waiting'}),
HashableDict({'admin': 'bob', 'state': 'waiting'})}
self.assertCountEqual(bdd_to_set(fsm, solve_ctlqx(fsm, ast)), solution)
def test_or(self):
ast = negation_normal_form(parse_ctlq("AF ('admin = bob' | AG ?)"))
self.assertTrue(check_ctlqx(ast))
fsm = self.init_model()
solution = {
HashableDict({
'admin': 'alice',
'state': 'waiting'
}),
HashableDict({
'admin': 'alice',
'state': 'processing'
})
}
self.assertCountEqual(bdd_to_set(fsm, solve_ctlqx(fsm, ast)), solution)
def test_bdd_to_set(self):
set_ = {
HashableDict({
'state': 'ready',
'request': 'TRUE'
}),
HashableDict({
'state': 'ready',
'request': 'FALSE'
})
}
init_nusmv()
fsm = self.init_model()
self.assertEqual(bdd_to_set(fsm, fsm.init), set_)
deinit_nusmv()
def test_au(self):
ast = negation_normal_form(parse_ctlq("A[? U 'state = processing']"))
self.assertTrue(check_ctlqx(ast))
fsm = self.init_model()
solution = {
HashableDict({
'admin': 'none',
'state': 'starting'
}),
HashableDict({
'admin': 'none',
'state': 'choosing'
}),
HashableDict({
'admin': 'alice',
'state': 'waiting'
}),
HashableDict({
'admin': 'bob',
'state': 'waiting'
})
}
self.assertCountEqual(bdd_to_set(fsm, solve_ctlqx(fsm, ast)), solution)
def test_ax(self):
ast = negation_normal_form(parse_ctlq("AX ?"))
self.assertTrue(check_ctlqx(ast))
fsm = self.init_model()
solution = {HashableDict({'admin': 'none', 'state': 'choosing'})}
self.assertCountEqual(bdd_to_set(fsm, solve_ctlqx(fsm, ast)), solution)
def cli(model_path, query, order):
"""Solve QUERY that belongs to fragment CTLQx for model in MODEL_PATH."""
try:
# Parse `query` and transform it in NNF.
ast = negation_normal_form(parse_ctlq(query))
# Check that `query` belongs to fragment CTLQx.
if not check_ctlqx(ast):
click.echo("Error: {query} does not belong to CTLQx".format(query=query))
# Quit PyTLQ.
sys.exit()
# Initialize NuSMV.
with init_nusmv():
# Load model from `model_path`.
load(model_path)
# Enable dynamic reordering of the variables.
enable_dynamic_reordering()
# Check if an order file is given.
if order:
# Build model with pre-calculated variable ordering.
compute_model(variables_ordering=order)
else:
# Build model.
compute_model()
# Retrieve FSM of the model.
fsm = prop_database().master.bddFsm
# Solve `query` in `fsm`.
solution = solve_ctlqx(fsm, ast)
# Display solution.
click.echo("Solution states:")
if not solution:
click.echo("No solution")
# Quit PyTLQ.
sys.exit()
elif solution.is_false():
click.echo("False")
# Quit PyTLQ.
sys.exit()
else:
size = fsm.count_states(solution)
if size > 100:
if click.confirm(
"The number of states is too large"
" ({size}). Do you still want to print"
" them?".format(size=size)
):
pprint(bdd_to_set(fsm, solution))
else:
pprint(bdd_to_set(fsm, solution))
# Ask for further manipulations.
while True:
command = click.prompt(
"\nWhat do you want to do?"
"\n 1. Project the solution on a"
" subset of the variables"
"\n 2. Simplify the solution according"
" to Chan's approximate conjunctive"
" decomposition"
"\n 3. Quit PyTLQ"
"\nYour choice",
type=click.IntRange(1, 3),
default=3,
)
# Check if solution must be projected or simplified.
if command == 1 or command == 2:
# Gather more information.
click.echo("")
if command == 2:
maximum = click.prompt(
"Please enter the maximum"
" number of variables that must"
" appear in the conjuncts of"
" the simplification",
type=int,
default=1,
)
variables = click.prompt(
"Please enter the list of" " variables of interest," " separated by commas",
type=str,
default="all the variables",
)
# Format `variables`.
if variables == "all the variables":
variables = None
else:
variables = variables.replace(" ", "").split(",")
if command == 1:
# Project solution and display projection.
click.echo("\nProjection:")
click.echo(project(fsm, solution, variables))
else:
# Simplify solution and display simplification.
click.echo("\nApproximate conjunctive decomposition:")
click.echo(simplify(fsm, solution, maximum, variables))
# No further manipulations are needed.
else:
break
except Exception as error:
click.echo("Error: {msg}".format(msg=error))
def cli(model_path, query, order):
"""Solve QUERY that belongs to fragment CTLQx for model in MODEL_PATH."""
try:
# Parse `query` and transform it in NNF.
ast = negation_normal_form(parse_ctlq(query))
# Check that `query` belongs to fragment CTLQx.
if not check_ctlqx(ast):
click.echo('Error: {query} does not belong to CTLQx'
.format(query=query))
# Quit PyTLQ.
sys.exit()
# Initialize NuSMV.
with init_nusmv():
# Load model from `model_path`.
load(model_path)
# Enable dynamic reordering of the variables.
enable_dynamic_reordering()
# Check if an order file is given.
if order:
# Build model with pre-calculated variable ordering.
compute_model(variables_ordering=order)
else:
# Build model.
compute_model()
# Retrieve FSM of the model.
fsm = prop_database().master.bddFsm
# Solve `query` in `fsm`.
solution = solve_ctlqx(fsm, ast)
# Display solution.
click.echo('Solution states:')
if not solution:
click.echo('No solution')
# Quit PyTLQ.
sys.exit()
elif solution.is_false():
click.echo('False')
# Quit PyTLQ.
sys.exit()
else:
size = fsm.count_states(solution)
if size > 100:
if click.confirm('The number of states is too large'
' ({size}). Do you still want to print'
' them?'.format(size=size)):
pprint(bdd_to_set(fsm, solution))
else:
pprint(bdd_to_set(fsm, solution))
# Ask for further manipulations.
while True:
command = click.prompt('\nWhat do you want to do?'
'\n 1. Project the solution on a'
' subset of the variables'
'\n 2. Simplify the solution according'
' to Chan\'s approximate conjunctive'
' decomposition'
'\n 3. Quit PyTLQ'
'\nYour choice',
type=click.IntRange(1, 3), default=3)
# Check if solution must be projected or simplified.
if command == 1 or command == 2:
# Gather more information.
click.echo('')
if command == 2:
maximum = click.prompt('Please enter the maximum'
' number of variables that must'
' appear in the conjuncts of'
' the simplification', type=int,
default=1)
variables = click.prompt('Please enter the list of'
' variables of interest,'
' separated by commas', type=str,
default='all the variables')
# Format `variables`.
if variables == 'all the variables':
variables = None
else:
variables = variables.replace(" ", "").split(',')
if command == 1:
# Project solution and display projection.
click.echo('\nProjection:')
click.echo(project(fsm, solution, variables))
else:
# Simplify solution and display simplification.
click.echo('\nApproximate conjunctive decomposition:')
click.echo(simplify(fsm, solution, maximum, variables))
# No further manipulations are needed.
else:
break
except Exception as error:
click.echo('Error: {msg}'.format(msg=error))
def test_ax(self):
ast = negation_normal_form(parse_ctlq("AX ?"))
self.assertTrue(check_ctlqx(ast))
fsm = self.init_model()
solution = {HashableDict({'admin': 'none', 'state': 'choosing'})}
self.assertCountEqual(bdd_to_set(fsm, solve_ctlqx(fsm, ast)), solution)