def test_paren():
expr = "\\A:( ab ) | c"
cfg = CFG(expr)
assert cfg.__str__() == "A: r'ab|c'"
assert cfg.parse('')[1] == False
assert cfg.parse('ab') == ('', 'ab')
assert cfg.parse('c') == ('', 'c')
assert cfg.parse('ac') == ('ac', False)
# this works properly:
d = {'g': Symbol(None)}
_g = (Symbol('a') + d['g'] + Symbol('b')) | Symbol('')
d['g'].update(_g)
assert d['g'].parse('') == ('', True)
d['g'].parse('ab') == ('', 'b')
d['g'].parse('aabb') == ('', 'b')
expr = "\\A:a + \\A + b"
cfg = CFG(expr)
assert cfg.parse('') == ('', True)
assert cfg.parse('ab') == ('', 'b')
assert cfg.parse('aabb') == ('', 'b')
#assert cfg.parse('aabb') == ('', 'b')
#expr = "\\A:(a + \\A + b) | \\0"
expr = "\\A:( a + \\A + b ) | \\0"
cfg = CFG(expr)
print(cfg)
assert cfg.parse('') == ('', True)
assert cfg.parse('ab') == ('', 'b')
assert cfg.parse('aabb') == ('', 'b')
def test_add_or():
a = Symbol('a')
b = Symbol('b')
c = Symbol('c')
d = (a + b) | c
assert d.parse('')[1] == False
assert d.parse('ab')[0] == ''
assert d.parse('c')[0] == ''
d = a + (b | c)
assert d.parse('')[1] == False
assert d.parse('ab') == ('', 'b')
assert d.parse('ac') == ('', 'c')
def test_add():
a = Symbol('a')
b = Symbol('b')
c = a + b
assert c.parse('')[1] == False
assert c.parse('ab')[0] == ''
assert c.parse('abb')[0] != ''
a = Symbol('a*')
b = Symbol('b*')
c = a + b
assert c.parse('')[1] == True
assert c.parse('ab')[0] == ''
assert c.parse('aab')[0] == ''
assert c.parse('abb')[0] == ''
assert c.parse('aa')[0] == ''
assert c.parse('bb')[0] == ''
assert c.parse('aba')[0] != ''
def test_cfg_manual():
"""
test:
- create mapping from cfg expr
- create aggregate Symbol with arithmetic and parse
"""
in_cfg = ("\G :\WORD \SPACE \G | \\0\n" "\WORD :\w+\n" "\SPACE : ")
lines = [line for line in in_cfg.split('\n') if line.strip()]
mappings = Symbol.create_mappings(lines)
sw = Symbol(mappings['WORD'])
ss = Symbol(mappings['SPACE'])
g = sw + ss
g.rrec()
#g = g | Symbol('')
st = "first second third fourth fifth"
g.parse(st)
def test_or():
a = Symbol('a')
b = Symbol('b')
c = a | b
assert c.parse('')[1] == False
assert c.parse('a')[0] == ''
assert c.parse('b')[0] == ''
a = Symbol('a*')
b = Symbol('b*')
c = a | b
assert c.parse('')[1] == True
assert c.parse('aa')[0] == ''
assert c.parse('aa')[1] == 'aa'
assert c.parse('aab')[0] == 'b'
assert c.parse('aaba')[0] == 'ba'
assert c.parse('bb')[0] == ''
assert c.parse('bb')[1] == 'bb'
assert c.parse('bba')[0] == 'a'
assert c.parse('bbab')[0] == 'ab'
def test_star():
a = Symbol('a*')
assert a.parse('')[1] == True
assert a.parse('aa')[0] == ''
assert a.parse('aa')[1] == 'aa'
assert a.parse('aab')[0] == 'b'
assert a.parse('aaba')[0] == 'ba'
def test_inp_cfg():
"""
Simple example:
arg1: CFG that only has two important nonterminals: WORD and SPACE
arg2: "WORD", meaning "return all parts of input that are WORD tokens"
"""
# -> space is included if it's part of regex ?
# here, there is NO SPACE after the colon
# however, each symbol must be separated by a space
# a space in your regex means there must be two spaces ?
in_cfg = ("\G :\WORD \SPACE \G | \\0\n" "\WORD :\w+\n" "\SPACE : ")
lines = [line.strip() for line in in_cfg.split('\n') if line.strip()]
mappings = Symbol.create_mappings(lines)
def test_rec():
d = {'a': Symbol(None)}
_a = (Symbol('a') + d['a']) | Symbol('')
d['a'].update(_a)
assert d['a'].parse('') == ('', True)
assert d['a'].parse('a') == ('', True)
assert d['a'].parse('aa') == ('', True)
d = {'g': Symbol(None)}
_g = (Symbol('a') + d['g'] + Symbol('b')) | Symbol('')
d['g'].update(_g)
assert d['g'].parse('') == ('', True)
assert d['g'].parse('ab') == ('', 'b')
assert d['g'].parse('aabb') == ('', 'b')
assert d['g'].parse('aaabb') == ('', False)
assert d['g'].parse('aabbb') == ('b', 'b')
def test_create_symbol():
"""
old test:
- manual symbol creation
- manual parsing
"""
st = "first second third fourth fifth"
sw = Symbol('\w+')
ss = Symbol(' ')
while st:
st, curr = sw.parse(st)
print("parsed: {}".format(curr))
st, space = ss.parse(st)