def test_extend_implicit_grammar():
grammar1 = Grammar()
grammar1.add_implicit('(')
result = Grammar()
result.extend(grammar1)
assert result.patterns[0].token_id == result.tokens['(']
assert result.patterns[0].priority == -len('(')
assert result.patterns[0].is_implicit
def test_add_brackets():
grammar = Grammar()
open_id = grammar.add_implicit('(')
close_id = grammar.add_implicit(')')
assert grammar.brackets == set()
assert grammar.open_brackets == set()
assert grammar.close_brackets == set()
grammar.add_brackets(open_id, close_id)
assert grammar.brackets == {(open_id, close_id)}
assert grammar.open_brackets == {open_id}
assert grammar.close_brackets == {close_id}
assert grammar.bracket_pairs[open_id] == close_id
def test_extend_brackets_grammar():
grammar1 = Grammar()
grammar1.add_brackets(grammar1.add_implicit('('),
grammar1.add_implicit(')'))
grammar2 = Grammar()
grammar2.add_brackets(grammar2.add_implicit('('),
grammar2.add_implicit(')'))
grammar2.add_brackets(grammar2.add_implicit('['),
grammar2.add_implicit(']'))
result = Grammar.merge(grammar1, grammar2)
assert result.brackets == {(result.tokens['['], result.tokens[']']),
(result.tokens['('], result.tokens[')'])}
def create_core_grammar() -> Grammar:
""" This function is used for initialize default grammar """
grammar = Grammar()
grammar.add_pattern(grammar.add_token('Comment'), RE_COMMENT)
grammar.add_pattern(grammar.add_token('Whitespace'), RE_WHITESPACE)
grammar.add_pattern(grammar.add_token('Name'), RE_NAME)
grammar.add_pattern(grammar.add_token('NewLine'), RE_NEWLINE)
grammar.add_pattern(grammar.add_token('String'), RE_STRING_SINGLE)
grammar.add_pattern(grammar.add_token('String'), RE_STRING_DOUBLE)
grammar.add_pattern(grammar.add_token('Integer'), RE_NUMBER_BINARY)
grammar.add_pattern(grammar.add_token('Integer'), RE_NUMBER_OCTAL)
grammar.add_pattern(grammar.add_token('Integer'), RE_NUMBER_DECIMAL)
grammar.add_pattern(grammar.add_token('Integer'), RE_NUMBER_HEXADECIMAL)
grammar.add_pattern(grammar.add_token('Float'), RE_FLOAT_POINT)
grammar.add_pattern(grammar.add_token('Float'), RE_FLOAT_EXPONENT)
grammar.add_implicit('(')
grammar.add_implicit(')')
grammar.add_implicit('[')
grammar.add_implicit(']')
grammar.add_implicit('{')
grammar.add_implicit('}')
grammar.add_implicit('<')
grammar.add_implicit('>')
grammar.add_trivia(grammar.tokens['Comment'])
grammar.add_trivia(grammar.tokens['Whitespace'])
grammar.add_brackets(grammar.tokens['('], grammar.tokens[')'])
grammar.add_brackets(grammar.tokens['['], grammar.tokens[']'])
grammar.add_brackets(grammar.tokens['{'], grammar.tokens['}'])
return grammar
def test_add_packrat_parser():
grammar = Grammar()
stmt_id = grammar.add_parselet('stmt',
kind=ParseletKind.Packrat,
result_type=SyntaxToken)
star_id = grammar.add_implicit('*')
assert grammar.add_parser(
stmt_id,
make_sequence(grammar.add_implicit('('), stmt_id,
grammar.add_implicit(')')))
assert grammar.add_parser(stmt_id,
make_sequence(grammar.add_implicit('(')))
assert grammar.add_parser(stmt_id, star_id)
assert grammar.add_parser(stmt_id, stmt_id)
def convert_node(grammar: Grammar, node: CombinatorNode, location: Location) -> Combinator:
if isinstance(node, SequenceNode):
return make_sequence(*(convert_node(grammar, child, location) for child in node.combinators))
if isinstance(node, RepeatNode):
return make_repeat(convert_node(grammar, node.combinator, location))
if isinstance(node, OptionalNode):
return make_optional(convert_node(grammar, node.combinator, location))
if isinstance(node, NamedNode):
return make_named(node.name.value, convert_node(grammar, node.combinator, location))
if isinstance(node, ImplicitNode):
token_id = grammar.add_implicit(ast.literal_eval(node.value.value), location=location)
return make_token(token_id)
if isinstance(node, ReferenceNode):
name = node.name.value
if name in grammar.tokens:
if node.priority:
raise DiagnosticError(location, f'Token combinator can not have priority')
return make_token(grammar.tokens[name])
elif name in grammar.parselets:
priority = node.priority and ast.literal_eval(node.priority.value)
return make_parselet(grammar.parselets[name], priority)
else:
raise DiagnosticError(location, f"Not found symbol {name} in grammar")
raise NotImplementedError(f'Not implemented conversion from node to combinator: {type(node).__name__}')
def test_add_implicit_token():
grammar = Grammar()
token_id = grammar.add_implicit('+')
assert token_id.name == '+'
assert token_id.description == '+'
assert token_id.is_implicit
assert '+' in grammar.tokens
assert len(grammar.patterns) == 1
pattern = grammar.patterns[0]
assert pattern.token_id == token_id
assert pattern.pattern == re.compile(re.escape('+'))
assert pattern.priority < 0
assert pattern.is_implicit
def grammar() -> Grammar:
grammar = Grammar()
whitespace_id = grammar.add_pattern(grammar.add_token('Whitespace'),
r'\s+')
grammar.add_trivia(whitespace_id)
grammar.add_pattern(grammar.add_token('Number'), r'[0-9]+')
grammar.add_pattern(grammar.add_token('Name'), r'[a-zA-Z_][a-zA-Z0-9]+')
grammar.add_implicit("for")
grammar.add_implicit("while")
grammar.add_implicit("+")
grammar.add_implicit("-")
return grammar
def test_add_pratt_parser():
grammar = Grammar()
expr_id = grammar.add_parselet('expr',
kind=ParseletKind.Pratt,
result_type=SyntaxToken)
integer_id = grammar.add_token('Integer')
string_id = grammar.add_token('String')
plus_id = grammar.add_implicit('+')
star_id = grammar.add_implicit('*')
table = cast(PrattTable, grammar.tables[expr_id])
assert table.prefix_tokens == set()
assert grammar.add_parser(expr_id, integer_id)
assert integer_id in table.prefix_tokens, "Cleanup of pratt table prefix tokens is not worked"
assert grammar.add_parser(expr_id, make_named('value', string_id))
assert string_id in table.prefix_tokens, "Cleanup of pratt table prefix tokens is not worked"
assert grammar.add_parser(expr_id, make_sequence(expr_id, plus_id,
expr_id))
assert grammar.add_parser(
expr_id,
make_sequence(make_named('lhs', expr_id), make_named('op', star_id),
expr_id))
def create_combinator_grammar() -> Grammar:
"""
Create grammar for parse combinator definition
P.S. This grammar is used for bootstrap process of initial grammar, e.g. definition of combinators in grammar
"""
grammar = Grammar()
grammar.extend(create_core_grammar())
# tokens
name_id = grammar.tokens['Name']
string_id = grammar.tokens['String']
number_id = grammar.tokens['Integer']
colon_id = grammar.add_implicit(':')
parent_open_id = grammar.tokens['(']
parent_close_id = grammar.tokens[')']
square_open_id = grammar.tokens['[']
square_close_id = grammar.tokens[']']
curly_open_id = grammar.tokens['{']
curly_close_id = grammar.tokens['}']
less_id = grammar.tokens['<']
great_id = grammar.tokens['>']
# parse combinator definition
comb_id = grammar.add_parselet('combinator', result_type=CombinatorNode)
seq_id = grammar.add_parselet('combinator_sequence', result_type=SequenceNode)
# combinator := name: Name ":" combinator=combinator ; named variable
grammar.add_parser(
comb_id,
make_sequence(make_named('name', name_id), colon_id, make_named('combinator', comb_id)),
make_ctor(NamedNode)
)
# combinator := name: Name [ '<' priority: Number '>' ] ; reference to parselet or token
grammar.add_parser(
comb_id,
make_sequence(make_named('name', name_id), make_optional(less_id, make_named('priority', number_id), great_id)),
make_ctor(ReferenceNode)
)
# combinator := value: String ; reference to implicit token
grammar.add_parser(comb_id, make_named('value', string_id), make_ctor(ImplicitNode))
# combinator := '[' combinator: combinator_sequence ']' ; optional combinator
grammar.add_parser(
comb_id,
make_sequence(square_open_id, make_named('combinator', seq_id), square_close_id),
make_ctor(OptionalNode)
)
# combinator := '{' combinator: combinator_sequence '}' ; repeat combinator
grammar.add_parser(
comb_id,
make_sequence(curly_open_id, make_named('combinator', seq_id), curly_close_id),
make_ctor(RepeatNode)
)
# combinator := '(' combinator: combinator_sequence ')' ; parenthesis combinator
grammar.add_parser(
comb_id,
make_sequence(parent_open_id, make_named('combinator', seq_id), parent_close_id),
make_return_variable('combinator')
)
# combinator_sequence := combinators:combinator combinators:{ combinator } ; sequence combinator
grammar.add_parser(
seq_id,
make_sequence(make_named('combinators', comb_id), make_named('combinators', make_repeat(comb_id))),
make_ctor(SequenceNode)
)
return grammar
