def dot(category):
nfa = NFA()
nfa.classifier = CharClassifier([0])
nfa.add_state(0)
nfa.add_state(1)
nfa.add_transition(0, 1, 1)
nfa.mark_starting(0)
nfa.mark_accepting(1, category)
return nfa
def atom(char, category):
nfa = NFA()
nfa.classifier = CharClassifier([ord(char), ord(char) + 1])
nfa.add_state(0)
nfa.add_state(1)
nfa.add_transition(0, 1, 1)
nfa.mark_starting(0)
nfa.mark_accepting(1, category)
return nfa
def char_set(markers, accept, category):
nfa = NFA()
nfa.classifier = CharClassifier(markers)
nfa.add_state(0)
nfa.add_state(1)
for domain in accept:
nfa.add_transition(0, 1, domain)
nfa.mark_starting(0)
nfa.mark_accepting(1, category)
return nfa
def test_ambiguous_subcons(self):
nfa = NFA()
for i in range(1, 7):
nfa.add_state(i)
nfa.add_transition(1, 2, 'i')
nfa.add_transition(2, 3, 'f')
nfa.add_transition(1, 4, 'i')
nfa.add_transition(1, 5, 'a')
nfa.add_transition(1, 6, 'f')
for i in range(4, 7):
for char in 'iaf':
nfa.add_transition(i, i, char)
nfa.mark_starting(1)
nfa.mark_accepting(3, 'if') # if
for i in range(4, 7):
nfa.mark_accepting(i, 'id') # id
trans_matrix, starting_state, accepting_states = subset_cons(nfa, {'if': CategoryInfo('IF', 0),
'id': CategoryInfo('ID', -1)})
d1 = frozenset({1})
d2 = frozenset({2, 4})
d3 = frozenset({3, 4})
d4 = frozenset({4})
d5 = frozenset({5})
d6 = frozenset({6})
self.assertEqual(trans_matrix,
{d1: {'a': d5,
'i': d2,
'f': d6},
d2: {'a': d4,
'i': d4,
'f': d3},
d3: {'a': d4,
'i': d4,
'f': d4},
d4: {'a': d4,
'i': d4,
'f': d4},
d5: {'a': d5,
'i': d5,
'f': d5},
d6: {'a': d6,
'i': d6,
'f': d6},
})
self.assertEqual(starting_state, d1)
self.assertEqual(accepting_states, {d2: 'id',
d3: 'if',
d4: 'id',
d5: 'id',
d6: 'id'})
def cat(nfa_list):
nfa_list = list(nfa_list)
cat_nfa = NFA()
cat_classifier, class_maps = merge_classifiers(nfa.classifier
for nfa in nfa_list)
cat_nfa.classifier = cat_classifier
for i, z in enumerate(zip(nfa_list, class_maps)):
nfa, class_map = z # add everything to cat_nfa
for start, d in nfa.trans_matrix.items():
cat_nfa.add_state((i, start))
for char, end_set in d.items():
if char == epsilon:
for end in end_set:
cat_nfa.add_transition((i, start), (i, end), epsilon)
else:
for domain in class_map[char]:
for end in end_set:
cat_nfa.add_transition((i, start), (i, end),
domain)
# cat_nfa.alphabet |= nfa.alphabet
for i in range(len(nfa_list) - 1): # link these nfa
this = nfa_list[i]
next_ = nfa_list[i + 1]
for state in this.accepting_states:
cat_nfa.add_transition((i, state), (i + 1, next_.starting_state),
epsilon)
cat_nfa.mark_starting((0, nfa_list[0].starting_state))
for state, category in nfa_list[-1].accepting_states.items():
cat_nfa.mark_accepting((len(nfa_list) - 1, state), category)
return cat_nfa
def alt(nfa_list):
nfa_list = list(nfa_list)
alt_nfa = NFA()
alt_classifier, class_maps = merge_classifiers(nfa.classifier
for nfa in nfa_list)
alt_nfa.classifier = alt_classifier
for i, z in enumerate(zip(nfa_list, class_maps)):
nfa, class_map = z
for start, d in nfa.trans_matrix.items():
alt_nfa.add_state((i, start))
for char, end_set in d.items():
if char == epsilon:
for end in end_set:
alt_nfa.add_transition((i, start), (i, end), epsilon)
else:
for domain in class_map[char]:
for end in end_set:
alt_nfa.add_transition((i, start), (i, end),
domain)
# alt_nfa.alphabet |= nfa.alphabet
for state, category in nfa.accepting_states.items():
alt_nfa.mark_accepting((i, state), category)
alt_nfa.add_state(1) # starting state
for i, nfa in enumerate(nfa_list):
alt_nfa.add_transition(1, (i, nfa.starting_state), epsilon)
alt_nfa.mark_starting(1)
return alt_nfa
def build_nfa1(): # NFA on page 151 of the Dragon Book
nfa = NFA()
for i in range(4):
nfa.add_state(i)
nfa.add_transition(0, 0, 2)
nfa.add_transition(0, 1, 1)
nfa.add_transition(1, 1, 1)
nfa.add_transition(1, 2, 2)
nfa.add_transition(2, 1, 1)
nfa.add_transition(2, 3, 2)
nfa.add_transition(3, 0, 2)
nfa.add_transition(3, 1, 1)
nfa.mark_starting(0)
nfa.mark_accepting(3, 0)
nfa.classifier = CharClassifier([ord('a'), ord('b'), ord('b') + 1])
return nfa
def build_test_nfa(): # example NFA on EC pp.51
nfa = NFA()
for i in range(10):
nfa.add_state(i)
nfa.mark_starting(0)
nfa.mark_accepting(9, 1)
nfa.classifier = CharClassifier(
[ord('a'), ord('b'), ord('c'),
ord('c') + 1])
nfa.add_transition(0, 1, 1)
nfa.add_transition(1, 2, epsilon)
nfa.add_transition(2, 9, epsilon)
nfa.add_transition(2, 3, epsilon)
nfa.add_transition(3, 4, epsilon)
nfa.add_transition(3, 6, epsilon)
nfa.add_transition(4, 5, 2)
nfa.add_transition(5, 8, epsilon)
nfa.add_transition(6, 7, 3)
nfa.add_transition(7, 8, epsilon)
nfa.add_transition(8, 3, epsilon)
nfa.add_transition(8, 9, epsilon)
return nfa