def get_grammar():
# Expression grammar with float numbers
E, T, F = [NonTerminal(name) for name in ['E', 'T', 'F']]
PLUS, MULT, OPEN, CLOSE = [
Terminal(value) for value in ['+', '*', '(', ')']
]
NUMBER = Terminal('number', RegExRecognizer(r'\d+(\.\d+)?'))
productions = [(E, (E, PLUS, T)), (E, (T, )), (T, (T, MULT, F)),
(T, (F, )), (F, (OPEN, E, CLOSE)), (F, (NUMBER, ))]
return Grammar.from_struct(productions, E)
def test_terminal_nonterminal():
# Production A is a terminal ("a") and non-terminal at the same time.
# Thus, it must be recognized as non-terminal.
grammar = """
S: A B;
A: "a" | B;
B: "b";
"""
g = Grammar.from_string(grammar)
assert NonTerminal("A") in g.nonterminals
assert Terminal("A") not in g.terminals
assert Terminal("B") in g.terminals
assert NonTerminal("B") not in g.nonterminals
# Here A should be non-terminal while B should be terminal.
grammar = """
S: A B;
A: B;
B: "b";
"""
g = Grammar.from_string(grammar)
assert NonTerminal("A") in g.nonterminals
assert Terminal("A") not in g.terminals
assert Terminal("B") in g.terminals
assert NonTerminal("B") not in g.nonterminals
grammar = """
S: A;
A: S;
A: 'x';
"""
g = Grammar.from_string(grammar)
assert NonTerminal("S") in g.nonterminals
assert NonTerminal("A") in g.nonterminals
assert Terminal("A") not in g.terminals
assert Terminal("x") in g.terminals
grammar = """
S: S S;
S: 'x';
S: EMPTY;
"""
g = Grammar.from_string(grammar)
assert NonTerminal("S") in g.nonterminals
assert Terminal("x") in g.terminals
assert NonTerminal("x") not in g.nonterminals
assert Terminal("S") not in g.terminals
def test_multiple_terminal_definition():
# A is defined multiple times as terminal thus it must be recognized
# as non-terminal with alternative expansions.
grammar = """
S: A A;
A: "a";
A: "b";
"""
g = Grammar.from_string(grammar)
assert NonTerminal("A") in g.nonterminals
assert Terminal("A") not in g.terminals
def test_first_empty_in_rhs():
"""
Test FIRST calculation when there are empty derivations in RHS of a
production.
"""
grammar = """
S: A C;
A: B | EMPTY;
B: "b";
C: "c";
"""
g = Grammar.from_string(grammar)
first_set = first(g)
assert EMPTY in first_set[NonTerminal('A')]
assert Terminal('B') in first_set[NonTerminal('A')]
assert Terminal('B') in first_set[NonTerminal('S')]
# 'A' can derive empty, thus 'C' must be in firsts of 'S'.
assert Terminal('C') in first_set[NonTerminal('S')]
from parglare import Grammar, NonTerminal, Terminal, RegExRecognizer
# Expression grammar with float numbers
E, T, F = [NonTerminal(name) for name in ['E', 'T', 'F']]
PLUS, MULT, OPEN, CLOSE = [Terminal(value) for value in ['+', '*', '(', ')']]
NUMBER = Terminal('number', RegExRecognizer(r'\d+(\.\d+)?'))
productions = [(E, (E, PLUS, T)), (E, (T, )), (T, (T, MULT, F)), (T, (F, )),
(F, (OPEN, E, CLOSE)), (F, (NUMBER, ))]
def get_grammar():
return Grammar.from_struct(productions, E)