def test_overloading(file_regression: FileRegressionFixture) -> None: try: # Create and register a custom encoder @sdjson.encoders.register(Decimal) def encoder_1(obj): return "Result from first registration" # Test that we get the expected output from the first encoder assert sdjson.dumps(Decimal(1)) == '"Result from first registration"' # Create and register a new custom encoder that overloads the previous one @sdjson.encoders.register(Decimal) def encoder_2(obj): return "Result from second registration" # Test that we get the expected output from the second encoder assert sdjson.dumps(Decimal(2)) == '"Result from second registration"' print(sdjson.encoders.registry.items()) check_file_regression(remove_memaddr(str(sdjson.encoders.registry.items())), file_regression) finally: # Cleanup sdjson.encoders.unregister(Decimal)
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_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_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_highly_nested_objects_encoding():
# See #12051
l, d = [], {}
for x in range(100000):
l, d = [l], {'k': d}
with pytest.raises(RecursionError):
sdjson.dumps(l)
with pytest.raises(RecursionError):
sdjson.dumps(d)
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_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_non_string_keys_dict() -> None:
data = {'a': 1, (1, 2): 2}
# TODO:
if platform.python_implementation() == "PyPy":
match_string = r"key \(1, 2\) is not a string"
elif sys.version_info[:2] > (3, 6):
match_string = "keys must be str, int, float, bool or None, not tuple"
else:
match_string = "keys must be a string"
with pytest.raises(TypeError, match=match_string):
sdjson.dumps(data)
def test_multiple_files(): from .glossia import thorn from .glossia import talon # Test that we get the expected output when encoding a Decimal assert sdjson.dumps(Decimal(1)) == '"1"' # Test that we get the expected output when encoding a Fraction assert sdjson.dumps(Fraction(2, 3)) == '"2/3"' # Cleanup sdjson.encoders.unregister(Decimal) sdjson.encoders.unregister(Fraction)
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_dictrecursion():
x = {}
x["test"] = x
try:
sdjson.dumps(x)
except ValueError:
pass
else:
pytest.fail("didn't raise ValueError on dict recursion")
x = {}
y = {"a": x, "b": x}
# ensure that the marker is cleared
sdjson.dumps(x)
def test_highly_nested_objects_encoding() -> None:
# See #12051
l: List[List]
d: Dict[str, Dict]
l, d = [], {}
for x in range(100000):
l, d = [l], {'k': d}
with pytest.raises(RecursionError):
sdjson.dumps(l)
with pytest.raises(RecursionError):
sdjson.dumps(d)
def test_fraction_str():
# Create and register a custom encoder for Fraction that turns it into a str
@sdjson.encoders.register(Fraction)
def encode_fraction_str(obj):
return str(obj)
assert sdjson.dumps(Fraction(13, 10)) == '"13/10"'
assert sdjson.dumps(Fraction(3, 4)) == '"3/4"'
assert sdjson.dumps(Fraction(9, 11)) == '"9/11"'
assert sdjson.dumps(Fraction(140, 144)) == '"35/36"'
assert sdjson.dumps(Fraction(2, 7)) == '"2/7"'
# Cleanup
sdjson.encoders.unregister(Fraction)
def test_fraction_float():
# Create and register a custom encoder for Fraction that turns it into a float
@sdjson.encoders.register(Fraction)
def encode_fraction_float(obj):
return float(obj)
assert sdjson.dumps(Fraction(13, 10)) == "1.3"
assert sdjson.dumps(Fraction(3, 4)) == "0.75"
assert sdjson.dumps(Fraction(9, 11)) == "0.8181818181818182"
assert sdjson.dumps(Fraction(140, 144)) == "0.9722222222222222"
assert sdjson.dumps(Fraction(2, 7)) == "0.2857142857142857"
# Cleanup
sdjson.encoders.unregister(Fraction)
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_dictrecursion() -> None:
x: Dict[str, Dict] = {}
x["test"] = x
try:
sdjson.dumps(x)
except ValueError:
pass
else:
pytest.fail("didn't raise ValueError on dict recursion")
x = {}
y = {'a': x, 'b': x}
# ensure that the marker is cleared
sdjson.dumps(x)
def test_unregister() -> None: # Create and register a custom encoder for Decimal that turns it into a string @sdjson.encoders.register(Decimal) def encode_str(obj): return str(obj) # Test that we get the expected output from the first encoder assert sdjson.dumps(Decimal(1)) == '"1"' # Unregister that encoder sdjson.unregister_encoder(Decimal) # We should now get an error with pytest.raises(TypeError, match="Object of type [']*Decimal[']* is not JSON serializable") as e: sdjson.dumps(Decimal(2)) assert e.type is TypeError
def spectrum_search(self,
mass_spec: MassSpectrum,
n_hits: int = 5) -> List[SearchResult]:
"""
Perform a Quick Spectrum Search of the mass spectral library.
:param mass_spec: The mass spectrum to search against the library.
:param n_hits: The number of hits to return.
:return: List of possible identities for the mass spectrum.
"""
if not isinstance(mass_spec, MassSpectrum):
raise TypeError(
"`mass_spec` must be a pyms.Spectrum.MassSpectrum object.")
retry_count = 0
# Keep trying until it works
while retry_count < 240:
try:
res = requests.post(
f"http://localhost:5001/search/quick/?n_hits={n_hits}",
json=sdjson.dumps(mass_spec),
)
print(res.text)
return hit_list_from_json(res.text)
except requests.exceptions.ConnectionError:
time.sleep(0.5)
retry_count += 1
raise TimeoutError("Unable to communicate with the search server.")
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_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 full_search_with_ref_data(
self,
mass_spec: MassSpectrum,
n_hits: int = 5,
) -> List[Tuple[SearchResult, ReferenceData]]:
"""
Perform a Full Spectrum Search of the mass spectral library, including reference data.
:param mass_spec: The mass spectrum to search against the library.
:param n_hits: The number of hits to return.
:return: List of tuples consisting of the possible identities
for the mass spectrum and the reference data from the library.
"""
if not isinstance(mass_spec, MassSpectrum):
raise TypeError(
"`mass_spec` must be a pyms.Spectrum.MassSpectrum object.")
retry_count = 0
# Keep trying until it works
while retry_count < 240:
try:
res = requests.post(
f"http://localhost:5001/search/spectrum_with_ref_data/?n_hits={n_hits}",
json=sdjson.dumps(mass_spec))
return hit_list_with_ref_data_from_json(res.text)
except requests.exceptions.ConnectionError:
time.sleep(0.5)
retry_count += 1
raise TimeoutError("Unable to communicate with the search server.")
def check(indent, expected):
d1 = sdjson.dumps(h, indent=indent)
assert d1 == expected
sio = StringIO()
sdjson.dump(h, sio, indent=indent)
assert sio.getvalue() == expected
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_encode_mutated() -> None:
a = [object()] * 10
def crasher(obj):
del a[-1]
assert sdjson.dumps(a, default=crasher) == "[null, null, null, null, null]"
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 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_encode_symbols():
s = "\u273f\u2661\u273f" # Рю┐РЎАРю┐
encoded = sd_ujson.dumps(s)
encoded_json = sdjson.dumps(s)
assert len(encoded) == len(s) * 6 + 2 # 6 characters + quotes
assert encoded == encoded_json
decoded = sd_ujson.loads(encoded)
assert s == decoded
# sd_ujson outputs an UTF-8 encoded str object
encoded = sd_ujson.dumps(s, ensure_ascii=False)
# json outputs an unicode object
encoded_json = sdjson.dumps(s, ensure_ascii=False)
assert len(encoded) == len(s) + 2 # original length + quotes
assert encoded == encoded_json
decoded = sd_ujson.loads(encoded)
assert s == decoded
