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 clone_any(use, original, restrict=None):
'''
We can always convert Any() to a regular expression; the only question
is whether we have an open range or not.
'''
if restrict is None:
char = Character([(alphabet_.min, alphabet_.max)], alphabet_)
else:
char = Character(((char, char) for char in restrict), alphabet_)
log.debug(fmt('Any: cloned {0}', char))
regexp = Sequence(alphabet_, char)
return RegexpContainer.build(original, regexp, alphabet_, regexp_type,
use)
def clone_literal(use, original, text):
'''
Literal values are easy to transform.
'''
chars = [Character([(c, c)], alphabet_) for c in text]
regexp = Sequence(alphabet_, *chars)
log.debug(fmt('Literal: cloned {0}', regexp))
return RegexpContainer.build(original, regexp, alphabet_, regexp_type,
use)
def __add_groups(self, src, groups, stack):
'''
The target nfa nodes identified above are now used to create dfa
nodes.
'''
for nfa_nodes in groups:
(intervals, terminals) = groups[nfa_nodes]
char = Character(intervals, self.__alphabet)
(new, dest) = self.dfa.node(nfa_nodes)
self.dfa.connect(src, dest, char)
if new:
stack.append((dest, nfa_nodes, terminals))
def test_single(self):
m = TaggedFragments(UNICODE)
m.append(Character([('b', 'c')], UNICODE), 'bc')
l = list(m)
assert l == [(('b', 'c'), ['bc'])], l