def token_ref(self, ast):
name = camel2py(ast).upper()
value = self.tokens.get(name)
if value and isinstance(value, model.Model):
return value
if name in self.token_rules:
exp = self.token_rules[name]
else:
exp = model.Decorator(model.RuleRef(name))
self.token_rules[name] = exp
return exp
def rule(self, ast):
name = camel2py(ast.name)
exp = ast.exp
if name[0].isupper():
name = name.upper()
if isinstance(exp, model.Token):
if name in self.token_rules:
self.token_rules[name].exp = exp # it is a model._Decorator
else:
self.token_rules[name] = exp
return None
elif not ast.fragment and not isinstance(exp, model.Sequence):
ref = model.RuleRef(name.lower())
if name in self.token_rules:
self.token_rules[name].exp = ref
else:
self.token_rules[name] = ref
name = name.lower()
return model.Rule(ast, name, exp, ast.params, ast.kwparams)
class TOKEN_TREE:
tokens: list
RULE = grammars.Rule(
ast=None,
name="TOKEN_TREE",
exp=grammars.Choice([
grammars.RuleRef("LITERAL"),
grammars.RuleRef("IDENT"),
grammars.RuleRef("LIFETIME"),
grammars.RuleRef("PUNCT"),
grammars.Sequence(
AST(sequence=[
grammars.Token("("),
grammars.Closure(grammars.RuleRef("TOKEN_TREE")),
grammars.Token(")"),
])),
grammars.Sequence(
AST(sequence=[
grammars.Token("{"),
grammars.Closure(grammars.RuleRef("TOKEN_TREE")),
grammars.Token("}"),
])),
grammars.Sequence(
AST(sequence=[
grammars.Token("["),
grammars.Closure(grammars.RuleRef("TOKEN_TREE")),
grammars.Token("]"),
])),
]),
params=None,
kwparams=None,
)
@classmethod
def from_ast(cls, ast) -> TOKEN_TREE:
return cls(ast)
def rule_ref(self, ast):
assert ast[0].islower()
return model.RuleRef(camel2py(ast))
def rule_ref(self, ast):
return model.RuleRef(camel2py(ast))
def node_to_tatsu(node: gll_grammar.RuleNode):
match node:
case gll_grammar.Empty():
return tatsu_grammars.EmptyClosure()
case gll_grammar.LabeledNode(name, item):
return tatsu_grammars.Named(AST(name=name, exp=node_to_tatsu(item)))
case gll_grammar.StringLiteral(s):
return tatsu_grammars.Token(ast=s)
case gll_grammar.CharacterRange(_, _):
raise NotImplementedError("character ranges")
case gll_grammar.SymbolName(name):
return tatsu_grammars.RuleRef(ast=name)
case gll_grammar.Concatenation(items):
return tatsu_grammars.Sequence(
ast=AST(sequence=list(map(node_to_tatsu, items)))
)
case gll_grammar.Alternation(items):
return tatsu_grammars.Choice(ast=list(map(node_to_tatsu, items)))
case gll_grammar.Option(item):
return tatsu_grammars.Optional(exp=node_to_tatsu(item))
case gll_grammar.Repeated(False, item, separator=None, allow_trailing=False):
return tatsu_grammars.Closure(exp=node_to_tatsu(item))
case gll_grammar.Repeated(True, item, separator=None, allow_trailing=False):
return tatsu_grammars.PositiveClosure(exp=node_to_tatsu(item))
case gll_grammar.Repeated(False, item, separator, allow_trailing=False):
return tatsu_grammars.Join(
ast=AST(exp=node_to_tatsu(item), sep=tatsu_grammars.Token(ast=separator))
)
case gll_grammar.Repeated(True, item, separator, allow_trailing=False):
return tatsu_grammars.PositiveJoin(
ast=AST(exp=node_to_tatsu(item), sep=tatsu_grammars.Token(ast=separator))
)
case gll_grammar.Repeated(False, item, separator, allow_trailing=True):
item = node_to_tatsu(item)
separator = tatsu_grammars.Token(ast=separator)
return tatsu_grammars.Sequence(
ast=AST(
sequence=[
tatsu_grammars.Closure(
exp=tatsu_grammars.Sequence(
ast=AST(
sequence=[
item,
separator,
]
)
),
),
tatsu_grammars.Optional(exp=item),
]
)
)
case gll_grammar.Repeated(True, item, separator, allow_trailing=True):
item = node_to_tatsu(item)
separator = tatsu_grammars.Token(ast=separator)
return tatsu_grammars.Sequence(
ast=AST(
sequence=[
item,
tatsu_grammars.Closure(
exp=tatsu_grammars.Sequence(
ast=AST(
sequence=[
separator,
item,
]
)
),
),
tatsu_grammars.Optional(exp=separator),
]
)
)
case _:
# should be unreachable
assert False, repr(node)