def test_overlap(self):
m = TaggedFragments(UNICODE)
m.append(Character([('b', 'd')], UNICODE), 'bd')
m.append(Character([('c', 'e')], UNICODE), 'ce')
l = list(m)
assert l == [(('b', 'b'), ['bd']), (('c', 'd'), ['bd', 'ce']),
(('e', 'e'), ['ce'])], l
def test_overlap(self):
m = TaggedFragments(UNICODE)
m.append(Character([('b', 'd')], UNICODE), 'bd')
m.append(Character([('c', 'e')], UNICODE), 'ce')
l = list(m)
assert l == [(('b', 'b'), ['bd']), (('c', 'd'), ['bd', 'ce']),
(('e', 'e'), ['ce'])], l
def __fragment_transitions(self, nfa_nodes):
'''
From the given nodes we can accumulate the destination nodes and
terminals associated with each transition (edge/character).
At the same time we separate the character matches into non-overlapping
fragments.
'''
fragments = TaggedFragments(self.__alphabet)
for nfa_node in nfa_nodes:
for (dest, edge) in self.__nfa.transitions(nfa_node):
(nodes, terminals) = self.__nfa.connected([dest])
fragments.append(edge, (nodes, list(terminals)))
return fragments
def __fragment_transitions(self, nfa_nodes):
'''
From the given nodes we can accumulate the destination nodes and
terminals associated with each transition (edge/character).
At the same time we separate the character matches into non-overlapping
fragments.
'''
fragments = TaggedFragments(self.__alphabet)
for nfa_node in nfa_nodes:
for (dest, edge) in self.__nfa.transitions(nfa_node):
(nodes, terminals) = self.__nfa.connected([dest])
fragments.append(edge, (nodes, list(terminals)))
return fragments
def __build_table(self):
'''
Rewrite the graph as a transition table, with appropriate ordering.
'''
for src in self.__graph:
# construct an interval map of possible destinations and terminals
# given a character
fragments = TaggedFragments(self.__alphabet)
for (dest, char) in self.__graph.transitions(src):
fragments.append(char, (dest, self.__graph.terminal(dest)))
map_ = IntervalMap()
for (interval, dts) in fragments:
map_[interval] = dts
# collect empty transitions
empties = [(dest, self.__graph.terminal(dest))
# ordering here is reverse of what is required, which
# is ok because we use empties[-1] below
for dest in sorted(self.__graph.empty_transitions(src))]
self.__table[src] = (map_, empties)
def __build_table(self):
'''
Rewrite the graph as a transition table, with appropriate ordering.
'''
for src in self.__graph:
# construct an interval map of possible destinations and terminals
# given a character
fragments = TaggedFragments(self.__alphabet)
for (dest, char) in self.__graph.transitions(src):
fragments.append(char, (dest, self.__graph.terminal(dest)))
map_ = IntervalMap()
for (interval, dts) in fragments:
map_[interval] = dts
# collect empty transitions
empties = [
(dest, self.__graph.terminal(dest))
# ordering here is reverse of what is required, which
# is ok because we use empties[-1] below
for dest in sorted(self.__graph.empty_transitions(src))
]
self.__table[src] = (map_, empties)
def test_single(self):
m = TaggedFragments(UNICODE)
m.append(Character([('b', 'c')], UNICODE), 'bc')
l = list(m)
assert l == [(('b', 'c'), ['bc'])], l
def test_single(self):
m = TaggedFragments(UNICODE)
m.append(Character([('b', 'c')], UNICODE), 'bc')
l = list(m)
assert l == [(('b', 'c'), ['bc'])], l
def test_overlap(self):
m = TaggedFragments(UNICODE)
m.append(Character([("b", "d")], UNICODE), "bd")
m.append(Character([("c", "e")], UNICODE), "ce")
l = list(m)
assert l == [(("b", "b"), ["bd"]), (("c", "d"), ["bd", "ce"]), (("e", "e"), ["ce"])], l
def test_single(self):
m = TaggedFragments(UNICODE)
m.append(Character([("b", "c")], UNICODE), "bc")
l = list(m)
assert l == [(("b", "c"), ["bc"])], l
