def test_follow(self):
"""ll.py: Test FOLLOW set creation"""
# Grammar 4.28 in Aho, Ullman et al (except NUM instead of id)
lexer = Lexer()
lexer.addtoken(name='NUM', rule=r'[0-9]+')
p = ll.LLParser(lexer)
p.ap('E', 'T EP', lambda x: x[0] + x[1], start_production=True)
p.ap('EP', "'+' T EP", lambda x: x[1] + x[2])
p.ap('EP', "", lambda x: 0)
p.ap('T', "F TP", lambda x: x[0] * x[1])
p.ap('TP', "'*' F TP", lambda x: x[1] * x[2])
p.ap('TP', "", lambda x: 1)
p.ap('F', "'(' E ')'", lambda x: x[1])
p.ap('F', "NUM", lambda x: int(x[0]))
follow_e = p.follow(Symbol('E'))
follow_ep = p.follow(Symbol('EP'))
self.assertEqual(len(follow_e), 2)
self.assertEqual(follow_e, follow_ep)
names = {t.name: t.rule.pattern for t in follow_e}
correct_names = {'LITERAL': '\\)', '_EOF': "$"}
self.assertEqual(names, correct_names)
follow_t = p.follow(Symbol('T'))
follow_tp = p.follow(Symbol('TP'))
self.assertEqual(len(follow_t), 3)
self.assertEqual(follow_t, follow_tp)
names = {t.rule.pattern: t.name for t in follow_t}
correct_names = {'\\)': 'LITERAL', '\\+': 'LITERAL', "$": '_EOF'}
self.assertEqual(names, correct_names)
follow_f = p.follow(Symbol('F'))
self.assertEqual(len(follow_f), 4)
names = {t.rule.pattern: t.name for t in follow_f}
correct_names = {
'\\*': 'LITERAL',
'\\)': 'LITERAL',
'\\+': 'LITERAL',
"$": '_EOF'
}
self.assertEqual(names, correct_names)
def test_grammar(self):
"""ll.py: Test a basic LL(1) grammar"""
# Grammar 4.28 in Aho, Ullman et al (except NUM instead of id)
lexer = Lexer()
lexer.addtoken(name='NUM', rule=r'[0-9]+')
p = ll.LLParser(lexer)
p.ap('E', 'T EP', lambda x: x[0] + x[1], start_production=True)
p.ap('EP', "'+' T EP", lambda x: x[1] + x[2])
p.ap('EP', "", lambda x: 0)
p.ap('T', "F TP", lambda x: x[0] * x[1])
p.ap('TP', "'*' F TP", lambda x: x[1] * x[2])
p.ap('TP', "", lambda x: 1)
p.ap('F', "'(' E ')'", lambda x: x[1])
p.ap('F', "NUM", lambda x: int(x[0]))
self.assertEqual(p.parse("2+3*4"), 14)
self.assertEqual(p.parse("5"), 5)
self.assertEqual(p.parse("1+1+1+1+1+1+1"), 7)
self.assertEqual(p.parse("(5+2)"), 7)
self.assertEqual(p.parse("1+((((((((((3))))))))))"), 4)
def test_first(self):
"""ll.py: Test FIRST set creation"""
# Grammar 4.28 in Aho, Ullman et al (except NUM instead of id)
lexer = Lexer()
lexer.addtoken(name='NUM', rule=r'[0-9]+')
p = ll.LLParser(lexer)
p.ap('E', 'T EP', lambda x: x[0] + x[1], start_production=True)
p.ap('EP', "'+' T EP", lambda x: x[1] + x[2])
p.ap('EP', "", lambda x: 0)
p.ap('T', "F TP", lambda x: x[0] * x[1])
p.ap('TP', "'*' F TP", lambda x: x[1] * x[2])
p.ap('TP', "", lambda x: 1)
p.ap('F', "'(' E ')'", lambda x: x[1])
p.ap('F', "NUM", lambda x: int(x[0]))
first_f = p.first(SymbolString((Symbol('F'), )))
first_t = p.first(SymbolString((Symbol('T'), )))
first_e = p.first(SymbolString((Symbol('E'), )))
self.assertTrue(first_f == first_t == first_e)
self.assertEqual(len(first_f), 2)
names = {t.name: t.rule.pattern for t in first_f}
correct_names = {'LITERAL': '\\(', 'NUM': '[0-9]+'}
self.assertEqual(names, correct_names)
first_ep = p.first(SymbolString((Symbol('EP'), )))
self.assertEqual(len(first_ep), 2)
names = {t.name: t.rule.pattern for t in first_ep}
correct_names = {'LITERAL': '\\+', '_EPSILON': ''}
self.assertEqual(names, correct_names)
first_tp = p.first(SymbolString((Symbol('TP'), )))
self.assertEqual(len(first_tp), 2)
names = {t.name: t.rule.pattern for t in first_tp}
correct_names = {'LITERAL': '\\*', '_EPSILON': ''}
self.assertEqual(names, correct_names)
