def __init__(self, intervals, alphabet, classes=None,
inverted=False, complete=False):
super(Character, self).__init__(consumes=True, size=1)
self.__simple = CharSet(intervals, alphabet)
self.alphabet = alphabet
self.classes = classes if classes else []
self.inverted = inverted
self.complete = complete
class Character(BaseNode, SimpleCompilableMixin):
'''
Match a single character. Currently the `__contains__` method should be
used for testing; that will call the `BaseAlphabet` as required.
How can this be improved?
- `intervals` define simple character ranges (eg. 0-9).
- `alphabet` is the alphabet used.
- `classes` is a list of `(class_, label, invert)` triplets, where:
- `class_` is a method on `alphabet` (eg. `.digit`)
- `label` is used for display
- `invert` is true if `class_` should fail
- `inverted` is a global boolean that inverts the entire result (if `True`
the test should fail).
- `complete` is True if the test (without `invert`) will always succeed.
'''
def __init__(self, intervals, alphabet, classes=None,
inverted=False, complete=False):
super(Character, self).__init__(consumes=True, size=1)
self.__simple = CharSet(intervals, alphabet)
self.alphabet = alphabet
self.classes = classes if classes else []
self.inverted = inverted
self.complete = complete
def _kargs(self):
kargs = super(Character, self)._kargs()
kargs['intervals'] = self.__simple.intervals
return kargs
def _compile_args(self):
return [self]
def append_interval(self, interval):
'''Add an additional interval.'''
self.__simple.append(interval, self.alphabet)
def append_class(self, class_, label, inverted=False):
'''Add a character class (see class docs).'''
for (class2, _, inverted2) in self.classes:
if class_ == class2:
self.complete = self.complete or inverted != inverted2
# if inverted matches, complete, else we already have it
return
self.classes.append((class_, label, inverted))
def invert(self):
'''Invert the selection.'''
self.inverted = not self.inverted
def __contains__(self, character):
result = self.complete
if not result:
for (class_, _, invert) in self.classes:
result = class_(character) != invert
if result:
break
if not result:
result = character in self.__simple
if self.inverted:
result = not result
return result
def __str__(self):
'''This returns (the illegal) [^] for all and [] for none.'''
if self.complete:
return '[]' if self.inverted else '[^]'
contents = ''.join('\\' + label for (_, label, _) in self.classes)
contents += self.__simple.to_str(self.alphabet)
return '[' + ('^' if self.inverted else '') + contents + ']'
def __hash__(self):
return hash(str(self))
def __bool__(self):
return bool(self.classes or self.__simple)
def __nonzero__(self):
return self.__bool__()
def simplify(self):
'''Reduce to a simpler opcode if possible.'''
if self.complete:
if self.inverted:
return NoMatch()
else:
return Dot(True)
else:
if self.classes or self.inverted:
return self
else:
return self.__simple.simplify(self.alphabet, self)
