def random_get_span(inputs, delimiter_dict, type_dict):
"""Samples random GetSpan expression."""
while True:
is_delimiters = [random.randint(1, 2), random.randint(1, 2)]
if delimiter_dict and is_delimiters[0] == 1:
r1 = random_delimiter(delimiter_dict)
indices = [i for i in dsl.INDEX if abs(i) <= delimiter_dict[r1]]
i1 = random.choice(indices)
else:
r1 = random_type(type_dict)
indices = [i for i in dsl.INDEX if abs(i) <= type_dict[r1]]
i1 = random.choice(indices)
if delimiter_dict and is_delimiters[1] == 1:
r2 = random_delimiter(delimiter_dict)
indices = [i for i in dsl.INDEX if abs(i) <= delimiter_dict[r2]]
i2 = random.choice(indices)
else:
r2 = random_type(type_dict)
indices = [i for i in dsl.INDEX if abs(i) <= type_dict[r2]]
i2 = random.choice(indices)
expr = dsl.GetSpan(
r1, i1, random_boundary(), r2, i2, random_boundary())
if not _is_output_empty(expr, inputs):
return expr
def random_get_span(inputs, delimiter_dict, type_dict):
"""Samples random GetSpan expression."""
while True:
is_delimiters = [random.randint(1, 2), random.randint(1, 2)]
if delimiter_dict and is_delimiters[0] == 1:
r1 = random_delimiter(delimiter_dict)
indices = [i for i in dsl.INDEX if abs(i) <= delimiter_dict[r1]]
i1 = random.choice(indices)
else:
r1 = random_type(type_dict)
indices = [i for i in dsl.INDEX if abs(i) <= type_dict[r1]]
i1 = random.choice(indices)
if delimiter_dict and is_delimiters[1] == 1:
r2 = random_delimiter(delimiter_dict)
indices = [i for i in dsl.INDEX if abs(i) <= delimiter_dict[r2]]
i2 = random.choice(indices)
else:
r2 = random_type(type_dict)
indices = [i for i in dsl.INDEX if abs(i) <= type_dict[r2]]
i2 = random.choice(indices)
expr = dsl.GetSpan(r1, i1, random_boundary(), r2, i2,
random_boundary())
# Make sure outputs are non-empty.
try:
if min(len(expr(input_value)) for input_value in inputs) == 0:
continue
except: # pylint: disable=[bare-except]
continue
return expr
def test_programs(self):
program1 = dsl.Concat(dsl.GetToken(dsl.Type.ALPHANUM, 3),
dsl.GetFrom(':'), dsl.GetFirst(dsl.Type.CHAR, 4))
self.assertEqual(program1('Ud 9:25,JV3 Obb'), '2525,JV3 ObbUd 9')
self.assertEqual(program1('zLny xmHg 8:43 A44q'), '843 A44qzLny')
program2 = dsl.Concat(
dsl.Compose(
dsl.Replace(' ', ','),
dsl.GetSpan(dsl.Type.PROP_CASE, 1, dsl.Boundary.START,
dsl.Type.PROP_CASE, 4, dsl.Boundary.END)),
dsl.ConstStr('.'), dsl.GetToken(dsl.Type.PROP_CASE, -1))
self.assertEqual(program2('Jacob Ethan James Alexander Michael'),
'Jacob,Ethan,James,Alexander.Michael')
self.assertEqual(program2('Earth Fire Wind Water Pluto Sun'),
'Earth,Fire,Wind,Water.Sun')
def test_decode(self):
id_token_table, token_id_table = tokens.build_token_tables()
self.assertEqual(len(token_id_table), len(id_token_table))
program = dsl.Concat(
dsl.Compose(
dsl.Replace(' ', ','),
dsl.GetSpan(dsl.Type.PROP_CASE, 1, dsl.Boundary.START,
dsl.Type.PROP_CASE, 4, dsl.Boundary.END)),
dsl.ConstStr('.'), dsl.GetToken(dsl.Type.PROP_CASE, -1))
encoding = program.encode(token_id_table)
self.assertEqual(encoding[-1], token_id_table[dsl.EOS])
decoded_program = dsl.decode_program(encoding, id_token_table)
self.assertEqual(
decoded_program('Jacob Ethan James Alexander Michael'),
'Jacob,Ethan,James,Alexander.Michael')
self.assertEqual(decoded_program('Earth Fire Wind Water Pluto Sun'),
'Earth,Fire,Wind,Water.Sun')