def test_extend_grammar():
grammar1 = Grammar()
grammar1.add_pattern(grammar1.add_token('A'), 'a+')
grammar1.add_pattern(grammar1.add_token('B'), 'b+')
grammar1.add_pattern(grammar1.add_token('C'), 'c+')
grammar1.add_parselet('expr', kind=ParseletKind.Pratt)
grammar2 = Grammar()
grammar2.add_pattern(grammar2.add_token('A'), '_a+')
grammar2.add_pattern(grammar2.add_token('B'), '_b+')
grammar2.add_parselet('expr', kind=ParseletKind.Pratt)
result = Grammar()
initial_count = len(result.symbols)
result.extend(grammar1)
result.extend(grammar2)
assert 'A' in result.tokens
assert 'B' in result.tokens
assert 'C' in result.tokens
assert len(result.symbols) == initial_count + 4
assert len(result.parselets) == 1
assert len(result.patterns) == 5
assert {pattern.pattern.pattern
for pattern in result.patterns
} == {'a+', '_a+', 'b+', '_b+', 'c+'}
def test_add_incorrect_pratt_parser():
grammar = Grammar()
stmt_id = grammar.add_parselet('stmt',
kind=ParseletKind.Pratt,
result_type=SyntaxToken)
expr_id = grammar.add_parselet('expr',
kind=ParseletKind.Pratt,
result_type=SyntaxToken)
integer_id = grammar.add_token('Integer')
with pytest.raises(GrammarError):
grammar.add_parser(expr_id, make_optional(integer_id))
with pytest.raises(GrammarError):
grammar.add_parser(expr_id, make_sequence(stmt_id))
with pytest.raises(GrammarError):
grammar.add_parser(expr_id, make_sequence(expr_id, stmt_id))
with pytest.raises(GrammarError):
grammar.add_parser(expr_id,
make_sequence(expr_id, make_optional(stmt_id)))
with pytest.raises(GrammarError):
grammar.add_parser(expr_id, make_sequence(expr_id, expr_id))
def test_extend_fail_grammar():
grammar1 = Grammar()
grammar1.add_parselet('expr', kind=ParseletKind.Pratt)
grammar2 = Grammar()
grammar2.add_parselet('expr', kind=ParseletKind.Packrat)
with pytest.raises(GrammarError):
Grammar.merge(grammar1, grammar2)
def test_add_incorrect_parselet():
grammar = Grammar()
symbol_count = len(grammar.symbols)
for name in {'+', 'Name'}:
with pytest.raises(GrammarError):
grammar.add_parselet(name)
assert len(
grammar.tokens
) == symbol_count, "Count of symbols in grammar is changed after failed call"
assert len(
grammar.symbols
) == symbol_count, "Count of symbols in grammar is changed after failed call"
def test_add_parselet():
grammar = Grammar()
symbol_count = len(grammar.symbols)
expr_id = grammar.add_parselet('expr')
assert expr_id.kind == ParseletKind.Packrat
assert len(grammar.parselets) == 1
assert len(grammar.symbols) == symbol_count + 1
def test_make_parselet():
grammar = Grammar()
name_id = grammar.add_parselet('name')
comb = make_parselet(name_id)
assert isinstance(comb, ParseletCombinator)
assert comb.parser_id == name_id
assert comb.result_type == SyntaxNode
assert comb.variables == {}
def test_extend_packrat_grammar():
grammar1 = Grammar()
grammar1.add_token('Number')
grammar1.add_token('String')
expr_id = grammar1.add_parselet('expr', result_type=object)
grammar1.add_parser(expr_id, 'Number')
grammar1.add_parser(expr_id, 'String')
grammar2 = Grammar()
grammar2.add_token('Number')
grammar2.add_token('String')
expr_id = grammar2.add_parselet('expr', result_type=object)
grammar2.add_parser(expr_id, 'Number')
grammar2.add_parser(expr_id, 'String')
result = Grammar.merge(grammar1, grammar2)
expr_id = result.parselets['expr']
assert expr_id in result.tables
assert len(cast(PackratTable, result.tables[expr_id]).parselets) == 4
def test_make_sequence():
grammar = Grammar()
name_id = grammar.add_token('Name')
expr_id = grammar.add_parselet('expr')
comb = make_sequence(name_id, expr_id)
assert isinstance(comb, SequenceCombinator)
assert len(comb) == 2
assert isinstance(comb[0], TokenCombinator)
assert isinstance(comb[1], ParseletCombinator)
assert comb.result_type == SyntaxNode
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 test_flat_sequence():
grammar = Grammar()
name_id = grammar.add_token('Name')
expr_id = grammar.add_parselet('expr')
combinators = tuple(
flat_sequence(TokenCombinator(name_id),
ParseletCombinator(expr_id),
SequenceCombinator((
TokenCombinator(name_id),
ParseletCombinator(expr_id),
)),
kind=SequenceCombinator))
assert len(combinators) == 4
assert isinstance(combinators[0], TokenCombinator)
assert isinstance(combinators[1], ParseletCombinator)
assert isinstance(combinators[2], TokenCombinator)
assert isinstance(combinators[3], ParseletCombinator)
def test_flat_combinator():
grammar = Grammar()
name_id = grammar.add_token('Name')
expr_id = grammar.add_parselet('expr')
# convert token id to token combinator
comb = flat_combinator(name_id)
assert isinstance(comb, TokenCombinator)
assert comb.token_id is name_id
# convert parselet id to parselet combinator
comb = flat_combinator(expr_id)
assert isinstance(comb, ParseletCombinator)
assert comb.parser_id is expr_id
assert comb.priority is None
# don't convert combinator
comb = TokenCombinator(name_id)
result = flat_combinator(comb)
assert comb is result
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 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('Name'), r'[a-zA-Z_][a-zA-Z0-9]*')
grammar.add_pattern(grammar.add_token('Number'), r'[0-9]+')
make_implicit = grammar.add_implicit
expr_id = grammar.add_parselet('expr',
kind=ParseletKind.Pratt,
result_type=object)
# expr := value:Number
grammar.add_parser(expr_id, "value:Number",
make_call(lambda value: value.value, object))
# expr := lhs:expr op:'+' rhs:expr
grammar.add_parser(expr_id,
'lhs:expr "**" rhs:expr <899>',
make_call(lambda lhs, rhs: (lhs, '**', rhs), object),
priority=900)
# expr := lhs:expr op:'+' rhs:expr
grammar.add_parser(expr_id,
'lhs:expr "+" rhs:expr <600>',
make_call(lambda lhs, rhs: (lhs, '+', rhs), object),
priority=600)
# expr := lhs:expr op:'-' rhs:expr
grammar.add_parser(expr_id,
'lhs:expr "-" rhs:expr <600>',
make_call(lambda lhs, rhs: (lhs, '-', rhs), object),
priority=600)
# expr := lhs:expr op:'*' rhs:expr
grammar.add_parser(expr_id,
'lhs:expr "*" rhs:expr <700>',
make_call(lambda lhs, rhs: (lhs, '*', rhs), object),
priority=700)
# expr := lhs:expr op:'/' rhs:expr
grammar.add_parser(expr_id,
'lhs:expr "/" rhs:expr <700>',
make_call(lambda lhs, rhs: (lhs, '/', rhs), object),
priority=700)
# expr := op:'-' value:expr
grammar.add_parser(expr_id, '"-" value:expr <800>',
make_call(lambda value: ('-', value), object))
# expr := op:'-' value:expr
grammar.add_parser(expr_id, '"+" value:expr <800>',
make_call(lambda value: ('+', value), object))
# expr := '(' value:expr ')'
grammar.add_parser(expr_id, '"(" value:expr ")"',
make_return_variable('value'))
return grammar
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
def test_add_idempotent_parselet():
grammar = Grammar()
p1 = grammar.add_parselet('name')
p2 = grammar.add_parselet('name')
assert p1 is p2 and p1 == p2
def test_add_parselet_different_kind():
grammar = Grammar()
grammar.add_parselet('expr', kind=ParseletKind.Packrat)
with pytest.raises(GrammarError):
grammar.add_parselet('expr', kind=ParseletKind.Pratt)