def test_linecol(data, line, col, idx):
with pytest.raises(sdjson.JSONDecodeError) as err:
sdjson.loads(data)
if platform.python_implementation() == "PyPy":
match = "Unexpected '!'"
else:
match = 'Expecting value'
if pypy:
assert err.value.msg.startswith(
match) # Fix for varying messages between PyPy versions
else:
assert err.value.msg == match
assert err.value.pos == idx
assert err.value.lineno == line
assert err.value.colno == col
if pypy:
assert re.match(
rf'{match}.*: line {line} column {col:d} \(char {idx:d}\)',
str(err.value))
else:
assert re.match(
rf'{match}: line {line} column {col:d} \(char {idx:d}\)',
str(err.value))
def test_list():
assert sdjson.dumps(list(BigNum)) == str([SMALL, BIG, HUGE, REALLY_HUGE])
assert sdjson.loads(sdjson.dumps(list(BigNum))) == list(BigNum)
assert sdjson.dumps(list(FloatNum)) == str([E, PI, TAU])
assert sdjson.loads(sdjson.dumps(list(FloatNum))) == list(FloatNum)
assert sdjson.dumps(list(WierdNum)) == '[Infinity, -Infinity, NaN]'
assert sdjson.loads(sdjson.dumps(list(WierdNum)))[:2] == list(WierdNum)[:2]
assert isnan(sdjson.loads(sdjson.dumps(list(WierdNum)))[2])
def test_weird_floats() -> None:
for enum, expected in zip(WierdNum, ("Infinity", "-Infinity", "NaN")):
assert sdjson.dumps(enum) == expected
if not isnan(enum):
assert float(sdjson.dumps(enum)) == enum
assert sdjson.loads(sdjson.dumps(enum)) == enum
else:
assert isnan(float(sdjson.dumps(enum)))
assert isnan(sdjson.loads(sdjson.dumps(enum)))
def test_weird_floats():
for enum, expected in zip(WierdNum, ('Infinity', '-Infinity', 'NaN')):
assert sdjson.dumps(enum) == expected
if not isnan(enum):
assert float(sdjson.dumps(enum)) == enum
assert sdjson.loads(sdjson.dumps(enum)) == enum
else:
assert isnan(float(sdjson.dumps(enum)))
assert isnan(sdjson.loads(sdjson.dumps(enum)))
def test_object_pairs_hook_with_unicode() -> None:
s = '{"xkd":1, "kcw":2, "art":3, "hxm":4, "qrt":5, "pad":6, "hoy":7}'
p = [("xkd", 1), ("kcw", 2), ("art", 3), ("hxm", 4), ("qrt", 5), ("pad", 6), ("hoy", 7)]
assert sdjson.loads(s, object_pairs_hook=lambda x: x) == p
od = sdjson.loads(s, object_pairs_hook=OrderedDict)
assert od == OrderedDict(p)
assert type(od) == OrderedDict
# the object_pairs_hook takes priority over the object_hook
assert sdjson.loads(s, object_pairs_hook=OrderedDict, object_hook=lambda x: None) == OrderedDict(p)
def test_encode_long_neg_conversion():
test_input = -9223372036854775808
output = sd_ujson.encode(test_input)
sdjson.loads(output)
sd_ujson.decode(output)
assert test_input == sdjson.loads(output)
assert output == sdjson.dumps(test_input)
assert test_input == sd_ujson.decode(output)
def test_parse_pass_2() -> None:
# from http://json.org/JSON_checker/test/pass2.json
JSON = r"""
[[[[[[[[[[[[[[[[[[["Not too deep"]]]]]]]]]]]]]]]]]]]
"""
# test in/out equivalence and parsing
res = sdjson.loads(JSON)
out = sdjson.dumps(res)
assert res == sdjson.loads(out)
def test_indent():
h = [
['blorpie'],
['whoops'],
[],
'd-shtaeou',
'd-nthiouh',
'i-vhbjkhnth',
{
'nifty': 87
},
{
'field': 'yes',
'morefield': False
},
]
expect = textwrap.dedent("""\
[
\t[
\t\t"blorpie"
\t],
\t[
\t\t"whoops"
\t],
\t[],
\t"d-shtaeou",
\t"d-nthiouh",
\t"i-vhbjkhnth",
\t{
\t\t"nifty": 87
\t},
\t{
\t\t"field": "yes",
\t\t"morefield": false
\t}
]""")
d1 = sdjson.dumps(h)
d2 = sdjson.dumps(h, indent=2, sort_keys=True, separators=(',', ': '))
d3 = sdjson.dumps(h, indent='\t', sort_keys=True, separators=(',', ': '))
d4 = sdjson.dumps(h, indent=2, sort_keys=True)
d5 = sdjson.dumps(h, indent='\t', sort_keys=True)
h1 = sdjson.loads(d1)
h2 = sdjson.loads(d2)
h3 = sdjson.loads(d3)
assert h1 == h
assert h2 == h
assert h3 == h
assert d2 == expect.expandtabs(2)
assert d3 == expect
assert d4 == d2
assert d5 == d3
def test_allow_nan():
for val in (float('inf'), float('-inf'), float('nan')):
out = sdjson.dumps([val])
if val == val: # inf
assert sdjson.loads(out) == [val]
else: # nan
res = sdjson.loads(out)
assert len(res) == 1
assert res[0] != res[0]
with pytest.raises(ValueError):
sdjson.dumps([val], allow_nan=False)
def test_failures() -> None:
for idx, doc in enumerate(JSONDOCS):
idx = idx + 1
if idx in SKIPS:
sdjson.loads(doc)
continue
try:
sdjson.loads(doc)
except sdjson.JSONDecodeError:
pass
else:
pytest.fail(f"Expected failure for fail{idx}.json: {doc!r}")
def test_failures():
for idx, doc in enumerate(JSONDOCS):
idx = idx + 1
if idx in SKIPS:
sdjson.loads(doc)
continue
try:
sdjson.loads(doc)
except sdjson.JSONDecodeError:
pass
else:
pytest.fail("Expected failure for fail{0}.json: {1!r}".format(
idx, doc))
def test_allow_nan() -> None:
for val in (float("inf"), float("-inf"), float("nan")):
out = sdjson.dumps([val])
if val == val: # inf
assert sdjson.loads(out) == [val]
else: # nan
res = sdjson.loads(out)
assert len(res) == 1
assert res[0] != res[0]
with pytest.raises(
ValueError,
match="Out of range float values are not JSON compliant"):
sdjson.dumps([val], allow_nan=False)
def test_encode_null_character():
test_input = "31337 \x00 1337"
output = sd_ujson.encode(test_input)
assert test_input == sdjson.loads(output)
assert output == sdjson.dumps(test_input)
assert test_input == sd_ujson.decode(output)
test_input = "\x00"
output = sd_ujson.encode(test_input)
assert test_input == sdjson.loads(output)
assert output == sdjson.dumps(test_input)
assert test_input == sd_ujson.decode(output)
assert '" \\u0000\\r\\n "' == sd_ujson.dumps(" \u0000\r\n ")
def test_bytes_decode():
for encoding, bom in [
('utf-8', codecs.BOM_UTF8),
('utf-16be', codecs.BOM_UTF16_BE),
('utf-16le', codecs.BOM_UTF16_LE),
('utf-32be', codecs.BOM_UTF32_BE),
('utf-32le', codecs.BOM_UTF32_LE),
]:
data = ["a\xb5\u20ac\U0001d120"]
encoded = sdjson.dumps(data).encode(encoding)
assert sdjson.loads(bom + encoded) == data
assert sdjson.loads(encoded) == data
with pytest.raises(UnicodeDecodeError):
sdjson.loads(b'["\x80"]')
# RFC-7159 and ECMA-404 extend JSON to allow documents that
# consist of only a string, which can present a special case
# not covered by the encoding detection patterns specified in
# RFC-4627 for utf-16-le (XX 00 XX 00).
assert sdjson.loads('"\u2600"'.encode('utf-16-le')) == '\u2600'
# Encoding detection for small (<4) bytes objects
# is implemented as a special case. RFC-7159 and ECMA-404
# allow single codepoint JSON documents which are only two
# bytes in utf-16 encodings w/o BOM.
assert sdjson.loads(b'5\x00') == 5
assert sdjson.loads(b'\x007') == 7
assert sdjson.loads(b'57') == 57
def test_parse_pass_3() -> None:
# from http://json.org/JSON_checker/test/pass3.json
JSON = r"""
{
"JSON Test Pattern pass3": {
"The outermost value": "must be an object or array.",
"In this test": "It is an object."
}
}
"""
# test in/out equivalence and parsing
res = sdjson.loads(JSON)
out = sdjson.dumps(res)
assert res == sdjson.loads(out)
def test_encode_to_utf8():
test_input = b"\xe6\x97\xa5\xd1\x88"
test_input = test_input.decode("utf-8")
enc = sd_ujson.encode(test_input, ensure_ascii=False)
dec = sd_ujson.decode(enc)
assert enc == sdjson.dumps(test_input, ensure_ascii=False)
assert dec == sdjson.loads(enc)
def test_sdjson():
d = Device(
1000,
"Television",
DeviceType.RC,
MagicMapping(
make="Samsung",
smart=True,
ports=Counter([Port.HDMI, Port.HDMI, Port.HDMI, Port.VGA]),
),
)
expected = """{
"device_id": 1000,
"display_name": "Television",
"device_type": 1,
"configuration": {
"make": "Samsung",
"smart": true,
"ports": {
"1": 3,
"2": 1
}
}
}"""
assert sdjson.dumps(d, indent=2) == expected
loaded_device = Device.from_dict(sdjson.loads(sdjson.dumps(d)))
# the Counter won't be equal because the enum's have become disassociated
assert loaded_device.device_id == d.device_id
assert loaded_device.display_name == d.display_name
assert loaded_device.configuration["make"] == d.configuration["make"]
assert loaded_device.configuration["smart"] == d.configuration["smart"]
def test_dict_keys():
s, b, h, r = BigNum
e, p, t = FloatNum
i, j, n = WierdNum
d = {
s: 'tiny',
b: 'large',
h: 'larger',
r: 'largest',
e: "Euler's number",
p: 'pi',
t: 'tau',
i: 'Infinity',
j: '-Infinity',
n: 'NaN',
}
nd = sdjson.loads(sdjson.dumps(d))
assert nd[str(SMALL)] == 'tiny'
assert nd[str(BIG)] == 'large'
assert nd[str(HUGE)] == 'larger'
assert nd[str(REALLY_HUGE)] == 'largest'
assert nd[repr(E)] == "Euler's number"
assert nd[repr(PI)] == 'pi'
assert nd[repr(TAU)] == 'tau'
assert nd['Infinity'] == 'Infinity'
assert nd['-Infinity'] == '-Infinity'
assert nd['NaN'] == 'NaN'
def test_separators() -> None:
h = [
["blorpie"],
["whoops"],
[],
"d-shtaeou",
"d-nthiouh",
"i-vhbjkhnth",
{
"nifty": 87
},
{
"field": "yes",
"morefield": False
},
]
expect = textwrap.dedent("""\
[
[
"blorpie"
] ,
[
"whoops"
] ,
[] ,
"d-shtaeou" ,
"d-nthiouh" ,
"i-vhbjkhnth" ,
{
"nifty" : 87
} ,
{
"field" : "yes" ,
"morefield" : false
}
]""")
d1 = sdjson.dumps(h)
d2 = sdjson.dumps(h, indent=2, sort_keys=True, separators=(" ,", " : "))
h1 = sdjson.loads(d1)
h2 = sdjson.loads(d2)
assert h1 == h
assert h2 == h
assert d2 == expect
def helper(expected_output, **encode_kwargs):
if "escape_forward_slashes" not in encode_kwargs:
encode_kwargs["escape_forward_slashes"] = True,
output = sd_ujson.encode(test_input, **encode_kwargs)
assert output == expected_output
if encode_kwargs.get("escape_forward_slashes", True):
assert test_input == sdjson.loads(output)
assert test_input == sd_ujson.decode(output)
def test_double_long_issue():
sut = {"a": -4342969734183514}
encoded = sdjson.dumps(sut)
decoded = sdjson.loads(encoded)
assert sut == decoded
encoded = sd_ujson.encode(sut)
decoded = sd_ujson.decode(encoded)
assert sut == decoded
def test_double_long_decimal_issue():
sut = {"a": -12345678901234.56789012}
encoded = sdjson.dumps(sut)
decoded = sdjson.loads(encoded)
assert sut == decoded
encoded = sd_ujson.encode(sut)
decoded = sd_ujson.decode(encoded)
assert sut == decoded
def __test_invalid_input(data, msg, idx):
with pytest.raises(sdjson.JSONDecodeError) as err:
sdjson.loads(data)
if pypy:
assert err.value.msg.startswith(
msg) # Fix for varying messages between PyPy versions
else:
assert err.value.msg == msg
assert err.value.pos == idx
assert err.value.lineno == 1
assert err.value.colno == idx + 1
if pypy:
assert re.match(
rf'{msg}.*: line 1 column {idx + 1:d} \(char {idx:d}\)',
str(err.value))
else:
assert re.match(rf'{msg}: line 1 column {idx + 1:d} \(char {idx:d}\)',
str(err.value))
def test_dump_skipkeys() -> None:
v = {b"invalid_key": False, "valid_key": True}
with pytest.raises(TypeError):
sdjson.dumps(v)
s = sdjson.dumps(v, skipkeys=True)
o = sdjson.loads(s)
assert "valid_key" in o
assert b"invalid_key" not in o
def test_dump_skipkeys():
v = {b'invalid_key': False, 'valid_key': True}
with pytest.raises(TypeError):
sdjson.dumps(v)
s = sdjson.dumps(v, skipkeys=True)
o = sdjson.loads(s)
assert 'valid_key' in o
assert b'invalid_key' not in o
def test_separators():
h = [
['blorpie'],
['whoops'],
[],
'd-shtaeou',
'd-nthiouh',
'i-vhbjkhnth',
{'nifty': 87},
{'field': 'yes', 'morefield': False},
]
expect = textwrap.dedent(
"""\
[
[
"blorpie"
] ,
[
"whoops"
] ,
[] ,
"d-shtaeou" ,
"d-nthiouh" ,
"i-vhbjkhnth" ,
{
"nifty" : 87
} ,
{
"field" : "yes" ,
"morefield" : false
}
]"""
)
d1 = sdjson.dumps(h)
d2 = sdjson.dumps(h, indent=2, sort_keys=True, separators=(' ,', ' : '))
h1 = sdjson.loads(d1)
h2 = sdjson.loads(d2)
assert h1 == h
assert h2 == h
assert d2 == expect
def test_object_pairs_hook() -> None:
s = '{"xkd":1, "kcw":2, "art":3, "hxm":4, "qrt":5, "pad":6, "hoy":7}'
p = [("xkd", 1), ("kcw", 2), ("art", 3), ("hxm", 4), ("qrt", 5),
("pad", 6), ("hoy", 7)]
assert sdjson.loads(s, object_pairs_hook=lambda x: x) == p
assert sdjson.load(StringIO(s), object_pairs_hook=lambda x: x) == p
od = sdjson.loads(s, object_pairs_hook=OrderedDict)
assert od == OrderedDict(p)
assert type(od) == OrderedDict
# the object_pairs_hook takes priority over the object_hook
assert sdjson.loads(s,
object_pairs_hook=OrderedDict,
object_hook=lambda x: None) == OrderedDict(p)
# check that empty object literals work (see #17368)
assert sdjson.loads("{}", object_pairs_hook=OrderedDict) == OrderedDict()
assert sdjson.loads('{"empty": {}}',
object_pairs_hook=OrderedDict) == OrderedDict([
("empty", OrderedDict())
])
def test_string_with_utf8_bom():
import sys
# see #18958
bom_json = "[1,2,3]".encode('utf-8-sig').decode('utf-8')
with pytest.raises(sdjson.JSONDecodeError) as e:
sdjson.loads(bom_json)
# TODO:
if sys.version_info.major >= 3 and sys.version_info.minor == 7:
assert 'BOM' in str(e)
with pytest.raises(sdjson.JSONDecodeError) as e:
sdjson.load(StringIO(bom_json))
# TODO:
if sys.version_info.major >= 3 and sys.version_info.minor == 7:
assert 'BOM' in str(e)
# make sure that the BOM is not detected in the middle of a string
bom_in_str = '"{}"'.format(''.encode('utf-8-sig').decode('utf-8'))
assert sdjson.loads(bom_in_str) == '\ufeff'
assert sdjson.load(StringIO(bom_in_str)) == '\ufeff'
def test_highly_nested_objects_decoding() -> None:
# test that loading highly-nested objects doesn't segfault when C
# accelerations are used. See #12017
with pytest.raises(RecursionError):
sdjson.loads('{"a":' * 100000 + '1' + '}' * 100000)
with pytest.raises(RecursionError):
sdjson.loads('{"a":' * 100000 + "[1]" + '}' * 100000)
with pytest.raises(RecursionError):
sdjson.loads('[' * 100000 + '1' + ']' * 100000)
def test_dict():
assert dict(ref_data) == ref_data.__dict__ == ref_data_dict_non_recursive
assert sdjson.loads(sdjson.dumps(ref_data)) == ref_data_dict
assert ReferenceData.from_dict(dict(ref_data)) == ref_data
for obj in [
test_string,
test_int,
test_float,
test_list_ints,
test_list_strs,
test_dictionary,
test_numbers,
test_tuple,
test_lists,
test_sequences,
test_lists,
]:
with pytest.raises(TypeError):
ReferenceData.from_dict(obj) # type: ignore