Awesome
ormsgpack
ormsgpack is a fast msgpack serialization library for Python derived from orjson, with native support for various Python types.
ormsgpack supports CPython 3.9, 3.10, 3.11, 3.12 and 3.13. Releases follow semantic versioning and serializing a new object type without an opt-in flag is considered a breaking change.
ormsgpack is licensed under both the Apache 2.0 and MIT licenses. The repository and issue tracker is github.com/aviramha/ormsgpack, and patches may be submitted there. There is a CHANGELOG available in the repository.
Usage
Install
To install a wheel from PyPI:
pip install --upgrade "pip>=20.3" # manylinux_x_y, universal2 wheel support
pip install --upgrade ormsgpack
To build a wheel, see packaging.
Quickstart
This is an example of serializing, with options specified, and deserializing:
>>> import ormsgpack, datetime, numpy
>>> data = {
... "type": "job",
... "created_at": datetime.datetime(1970, 1, 1),
... "status": "🆗",
... "payload": numpy.array([[1, 2], [3, 4]]),
... }
>>> ormsgpack.packb(data, option=ormsgpack.OPT_NAIVE_UTC | ormsgpack.OPT_SERIALIZE_NUMPY)
b'\x84\xa4type\xa3job\xaacreated_at\xb91970-01-01T00:00:00+00:00\xa6status\xa4\xf0\x9f\x86\x97\xa7payload\x92\x92\x01\x02\x92\x03\x04'
>>> ormsgpack.unpackb(_)
{'type': 'job', 'created_at': '1970-01-01T00:00:00+00:00', 'status': '🆗', 'payload': [[1, 2], [3, 4]]}
Serialize
def packb(
__obj: Any,
default: Optional[Callable[[Any], Any]] = ...,
option: Optional[int] = ...,
) -> bytes: ...
packb() serializes Python objects to msgpack.
It natively serializes
bytes, str, dict, list, tuple, int, float, bool,
dataclasses.dataclass, typing.TypedDict, datetime.datetime,
datetime.date, datetime.time, uuid.UUID, numpy.ndarray, and
None instances. It supports arbitrary types through default. It
serializes subclasses of str, int, dict, list,
dataclasses.dataclass, and enum.Enum. It does not serialize subclasses
of tuple to avoid serializing namedtuple objects as arrays. To avoid
serializing subclasses, specify the option ormsgpack.OPT_PASSTHROUGH_SUBCLASS.
The output is a bytes object.
The global interpreter lock (GIL) is held for the duration of the call.
It raises MsgpackEncodeError on an unsupported type. This exception message
describes the invalid object with the error message
Type is not msgpack serializable: .... To fix this, specify
default.
It raises MsgpackEncodeError on a str that contains invalid UTF-8.
It raises MsgpackEncodeError if a dict has a key of a type other than str or bytes,
unless OPT_NON_STR_KEYS is specified.
It raises MsgpackEncodeError if the output of default recurses to handling by
default more than 254 levels deep.
It raises MsgpackEncodeError on circular references.
It raises MsgpackEncodeError if a tzinfo on a datetime object is
unsupported.
MsgpackEncodeError is a subclass of TypeError.
default
To serialize a subclass or arbitrary types, specify default as a
callable that returns a supported type. default may be a function,
lambda, or callable class instance. To specify that a type was not
handled by default, raise an exception such as TypeError.
>>> import ormsgpack, decimal
>>> def default(obj):
... if isinstance(obj, decimal.Decimal):
... return str(obj)
... raise TypeError
...
>>> ormsgpack.packb(decimal.Decimal("0.0842389659712649442845"))
TypeError: Type is not msgpack serializable: decimal.Decimal
>>> ormsgpack.packb(decimal.Decimal("0.0842389659712649442845"), default=default)
b'\xb80.0842389659712649442845'
>>> ormsgpack.packb({1, 2}, default=default)
TypeError: Type is not msgpack serializable: set
The default callable may return an object that itself
must be handled by default up to 254 times before an exception
is raised.
It is important that default raise an exception if a type cannot be handled.
Python otherwise implicitly returns None, which appears to the caller
like a legitimate value and is serialized:
>>> import ormsgpack, decimal
>>> def default(obj):
... if isinstance(obj, decimal.Decimal):
... return str(obj)
...
>>> ormsgpack.packb({"set":{1, 2}}, default=default)
b'\x81\xa3set\xc0'
>>> ormsgpack.unpackb(_)
{'set': None}
To serialize a type as a MessagePack extension type, return an
ormsgpack.Ext object. The instantiation arguments are an integer in
the range [0, 127] and a bytes object, defining the type and
value, respectively.
>>> import ormsgpack, decimal
>>> def default(obj):
... if isinstance(obj, decimal.Decimal):
... return ormsgpack.Ext(0, str(obj).encode())
... raise TypeError
...
>>> ormsgpack.packb(decimal.Decimal("0.0842389659712649442845"), default=default)
b'\xc7\x18\x000.0842389659712649442845'
option
To modify how data is serialized, specify option. Each option is an integer
constant in ormsgpack. To specify multiple options, mask them together, e.g.,
option=ormsgpack.OPT_NON_STR_KEYS | ormsgpack.OPT_NAIVE_UTC.
OPT_NAIVE_UTC
Serialize datetime.datetime objects without a tzinfo and numpy.datetime64
objects as UTC. This has no effect on datetime.datetime objects that have
tzinfo set.
>>> import ormsgpack, datetime
>>> ormsgpack.packb(
... datetime.datetime(1970, 1, 1, 0, 0, 0),
... )
b'\xb31970-01-01T00:00:00'
>>> ormsgpack.unpackb(_)
'1970-01-01T00:00:00'
>>> ormsgpack.packb(
... datetime.datetime(1970, 1, 1, 0, 0, 0),
... option=ormsgpack.OPT_NAIVE_UTC,
... )
b'\xb91970-01-01T00:00:00+00:00'
>>> ormsgpack.unpackb(_)
'1970-01-01T00:00:00+00:00'
OPT_NON_STR_KEYS
Serialize dict keys of type other than str. This allows dict keys
to be one of str, int, float, bool, None, datetime.datetime,
datetime.date, datetime.time, enum.Enum, and uuid.UUID.
>>> import ormsgpack, datetime, uuid
>>> ormsgpack.packb(
... {uuid.UUID("7202d115-7ff3-4c81-a7c1-2a1f067b1ece"): [1, 2, 3]},
... option=ormsgpack.OPT_NON_STR_KEYS,
... )
b'\x81\xd9$7202d115-7ff3-4c81-a7c1-2a1f067b1ece\x93\x01\x02\x03'
>>> ormsgpack.unpackb(_)
{'7202d115-7ff3-4c81-a7c1-2a1f067b1ece': [1, 2, 3]}
>>> ormsgpack.packb(
... {datetime.datetime(1970, 1, 1, 0, 0, 0): [1, 2, 3]},
... option=ormsgpack.OPT_NON_STR_KEYS | ormsgpack.OPT_NAIVE_UTC,
... )
b'\x81\xb91970-01-01T00:00:00+00:00\x93\x01\x02\x03'
>>> ormsgpack.unpackb(_)
{'1970-01-01T00:00:00+00:00': [1, 2, 3]}
These types are generally serialized how they would be as
values, e.g., datetime.datetime is still an RFC 3339 string and respects
options affecting it.
This option has the risk of creating duplicate keys. This is because non-str
objects may serialize to the same str as an existing key, e.g.,
{"1970-01-01T00:00:00+00:00": true, datetime.datetime(1970, 1, 1, 0, 0, 0): false}.
The last key to be inserted to the dict will be serialized last and a msgpack deserializer will presumably take the last
occurrence of a key (in the above, false). The first value will be lost.
This option is not compatible with ormsgpack.OPT_SORT_KEYS.
OPT_OMIT_MICROSECONDS
Do not serialize the microsecond component of datetime.datetime,
datetime.time and numpy.datetime64 instances.
>>> import ormsgpack, datetime
>>> ormsgpack.packb(
... datetime.datetime(1970, 1, 1, 0, 0, 0, 1),
... )
b'\xba1970-01-01T00:00:00.000001'
>>> ormsgpack.unpackb(_)
'1970-01-01T00:00:00.000001'
>>> ormsgpack.packb(
... datetime.datetime(1970, 1, 1, 0, 0, 0, 1),
... option=ormsgpack.OPT_OMIT_MICROSECONDS,
... )
b'\xb31970-01-01T00:00:00'
>>> ormsgpack.unpackb(_)
'1970-01-01T00:00:00'
OPT_PASSTHROUGH_BIG_INT
Enables passthrough of big (Python) ints. By setting this option, one can set a default function for ints larger than 63 bits, smaller ints are still serialized efficiently.
>>> import ormsgpack
>>> ormsgpack.packb(
... 2**65,
... )
TypeError: Integer exceeds 64-bit range
>>> ormsgpack.packb(
... 2**65,
... option=ormsgpack.OPT_PASSTHROUGH_BIG_INT,
... default=lambda _: {"type": "bigint", "value": str(_) }
... )
b'\x82\xa4type\xa6bigint\xa5value\xb436893488147419103232'
>>> ormsgpack.unpackb(_)
{'type': 'bigint', 'value': '36893488147419103232'}
OPT_PASSTHROUGH_DATACLASS
Passthrough dataclasses.dataclass instances to default. This allows
customizing their output but is much slower.
>>> import ormsgpack, dataclasses
>>> @dataclasses.dataclass
... class User:
... id: str
... name: str
... password: str
...
>>> def default(obj):
... if isinstance(obj, User):
... return {"id": obj.id, "name": obj.name}
... raise TypeError
...
>>> ormsgpack.packb(User("3b1", "asd", "zxc"))
b'\x83\xa2id\xa33b1\xa4name\xa3asd\xa8password\xa3zxc'
>>> ormsgpack.packb(User("3b1", "asd", "zxc"), option=ormsgpack.OPT_PASSTHROUGH_DATACLASS)
TypeError: Type is not msgpack serializable: User
>>> ormsgpack.packb(
... User("3b1", "asd", "zxc"),
... option=ormsgpack.OPT_PASSTHROUGH_DATACLASS,
... default=default,
... )
b'\x82\xa2id\xa33b1\xa4name\xa3asd'
OPT_PASSTHROUGH_DATETIME
Passthrough datetime.datetime, datetime.date, and datetime.time instances
to default. This allows serializing datetimes to a custom format, e.g.,
HTTP dates:
>>> import ormsgpack, datetime
>>> def default(obj):
... if isinstance(obj, datetime.datetime):
... return obj.strftime("%a, %d %b %Y %H:%M:%S GMT")
... raise TypeError
...
>>> ormsgpack.packb({"created_at": datetime.datetime(1970, 1, 1)})
b'\x81\xaacreated_at\xb31970-01-01T00:00:00'
>>> ormsgpack.packb({"created_at": datetime.datetime(1970, 1, 1)}, option=ormsgpack.OPT_PASSTHROUGH_DATETIME)
TypeError: Type is not msgpack serializable: datetime.datetime
>>> ormsgpack.packb(
... {"created_at": datetime.datetime(1970, 1, 1)},
... option=ormsgpack.OPT_PASSTHROUGH_DATETIME,
... default=default,
... )
b'\x81\xaacreated_at\xbdThu, 01 Jan 1970 00:00:00 GMT'
This does not affect datetimes in dict keys if using OPT_NON_STR_KEYS.
OPT_PASSTHROUGH_SUBCLASS
Passthrough subclasses of builtin types to default.
>>> import ormsgpack
>>> class Secret(str):
... pass
...
>>> def default(obj):
... if isinstance(obj, Secret):
... return "******"
... raise TypeError
...
>>> ormsgpack.packb(Secret("zxc"))
b'\xa3zxc'
>>> ormsgpack.packb(Secret("zxc"), option=ormsgpack.OPT_PASSTHROUGH_SUBCLASS)
TypeError: Type is not msgpack serializable: Secret
>>> ormsgpack.packb(Secret("zxc"), option=ormsgpack.OPT_PASSTHROUGH_SUBCLASS, default=default)
b'\xa6******'
This does not affect serializing subclasses as dict keys if using
OPT_NON_STR_KEYS.
OPT_PASSTHROUGH_TUPLE
Passthrough tuples to default.
>>> import ormsgpack
>>> ormsgpack.packb(
... (9193, "test", 42),
... )
b'\x93\xcd#\xe9\xa4test*'
>>> ormsgpack.unpackb(_)
[9193, 'test', 42]
>>> ormsgpack.packb(
... (9193, "test", 42),
... option=ormsgpack.OPT_PASSTHROUGH_TUPLE,
... default=lambda _: {"type": "tuple", "value": list(_)}
... )
b'\x82\xa4type\xa5tuple\xa5value\x93\xcd#\xe9\xa4test*'
>>> ormsgpack.unpackb(_)
{'type': 'tuple', 'value': [9193, 'test', 42]}
OPT_SERIALIZE_NUMPY
Serialize numpy.ndarray instances. For more, see
numpy.
OPT_SERIALIZE_PYDANTIC
Serialize pydantic.BaseModel instances.
OPT_SORT_KEYS
Serialize dict keys and pydantic model fields in sorted order. The default
is to serialize in an unspecified order.
This can be used to ensure the order is deterministic for hashing or tests. It has a substantial performance penalty and is not recommended in general.
>>> import ormsgpack
>>> ormsgpack.packb({"b": 1, "c": 2, "a": 3})
b'\x83\xa1b\x01\xa1c\x02\xa1a\x03'
>>> ormsgpack.packb({"b": 1, "c": 2, "a": 3}, option=ormsgpack.OPT_SORT_KEYS)
b'\x83\xa1a\x03\xa1b\x01\xa1c\x02'
The sorting is not collation/locale-aware:
>>> import ormsgpack
>>> ormsgpack.packb({"a": 1, "ä": 2, "A": 3}, option=ormsgpack.OPT_SORT_KEYS)
b'\x83\xa1A\x03\xa1a\x01\xa2\xc3\xa4\x02'
dataclass also serialize as maps but this has no effect on them.
OPT_UTC_Z
Serialize a UTC timezone on datetime.datetime and numpy.datetime64 instances
as Z instead of +00:00.
>>> import ormsgpack, datetime
>>> ormsgpack.packb(
... datetime.datetime(1970, 1, 1, 0, 0, 0, tzinfo=datetime.timezone.utc),
... )
b'\xb91970-01-01T00:00:00+00:00'
>>> ormsgpack.packb(
... datetime.datetime(1970, 1, 1, 0, 0, 0, tzinfo=datetime.timezone.utc),
... option=ormsgpack.OPT_UTC_Z
... )
b'\xb41970-01-01T00:00:00Z'
Deserialize
def unpackb(
__obj: Union[bytes, bytearray, memoryview],
/,
ext_hook: Optional[Callable[[int, bytes], Any]] = ...,
option: Optional[int] = ...,
) -> Any: ...
unpackb() deserializes msgpack to Python objects. It deserializes to dict,
list, int, float, str, bool, bytes and None objects.
bytes, bytearray, memoryview input are accepted.
ormsgpack maintains a cache of map keys for the duration of the process. This causes a net reduction in memory usage by avoiding duplicate strings. The keys must be at most 64 bytes to be cached and 512 entries are stored.
The global interpreter lock (GIL) is held for the duration of the call.
It raises MsgpackDecodeError if given an invalid type or invalid
msgpack.
MsgpackDecodeError is a subclass of ValueError.
ext_hook
To deserialize extension types, specify the optional ext_hook
argument. The value should be a callable and is invoked with the
extension type and value as arguments.
>>> import ormsgpack, decimal
>>> def ext_hook(tag, data):
... if tag == 0:
... return decimal.Decimal(data.decode())
... raise TypeError
...
>>> ormsgpack.packb(
... ormsgpack.Ext(0, str(decimal.Decimal("0.0842389659712649442845")).encode())
... )
b'\xc7\x18\x000.0842389659712649442845'
>>> ormsgpack.unpackb(_, ext_hook=ext_hook)
Decimal('0.0842389659712649442845'
option
unpackb() supports the OPT_NON_STR_KEYS option, that is similar to original msgpack's strict_map_key=False.
Be aware that this option is considered unsafe and disabled by default in msgpack due to possibility of HashDoS.
Types
dataclass
ormsgpack serializes instances of dataclasses.dataclass natively. It serializes
instances 40-50x as fast as other libraries and avoids a severe slowdown seen
in other libraries compared to serializing dict.
It is supported to pass all variants of dataclasses, including dataclasses
using __slots__, frozen dataclasses, those with optional or default
attributes, and subclasses. There is a performance benefit to not
using __slots__.
Dataclasses are serialized as maps, with every attribute serialized and in the order given on class definition:
>>> import dataclasses, ormsgpack, typing
>>> @dataclasses.dataclass
... class Member:
... id: int
... active: bool = dataclasses.field(default=False)
...
>>> @dataclasses.dataclass
... class Object:
... id: int
... name: str
... members: typing.List[Member]
...
>>> ormsgpack.packb(Object(1, "a", [Member(1, True), Member(2)]))
b'\x83\xa2id\x01\xa4name\xa1a\xa7members\x92\x82\xa2id\x01\xa6active\xc3\x82\xa2id\x02\xa6active\xc2'
Performance
--------------------------------------------------------------------------------- benchmark 'dataclass': 2 tests --------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_dataclass_ormsgpack 3.4248 (1.0) 7.7949 (1.0) 3.6266 (1.0) 0.3293 (1.0) 3.5815 (1.0) 0.0310 (1.0) 4;34 275.7434 (1.0) 240 1
test_dataclass_msgpack 140.2774 (40.96) 143.6087 (18.42) 141.3847 (38.99) 1.0038 (3.05) 141.1823 (39.42) 0.7304 (23.60) 2;1 7.0729 (0.03) 8 1
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
datetime
ormsgpack serializes datetime.datetime objects to
RFC 3339 format,
e.g., "1970-01-01T00:00:00+00:00". This is a subset of ISO 8601 and is
compatible with isoformat() in the standard library.
>>> import ormsgpack, datetime, zoneinfo
>>> ormsgpack.packb(
... datetime.datetime(2018, 12, 1, 2, 3, 4, 9, tzinfo=zoneinfo.ZoneInfo('Australia/Adelaide'))
... )
b'\xd9 2018-12-01T02:03:04.000009+10:30'
>>> ormsgpack.unpackb(_)
'2018-12-01T02:03:04.000009+10:30'
>>> ormsgpack.packb(
... datetime.datetime.fromtimestamp(4123518902).replace(tzinfo=datetime.timezone.utc)
... )
b'\xb92100-09-02T00:55:02+00:00'
>>> ormsgpack.unpackb(_)
'2100-09-02T00:55:02+00:00'
>>> ormsgpack.packb(
... datetime.datetime.fromtimestamp(4123518902)
... )
b'\xb32100-09-02T00:55:02'
>>> ormsgpack.unpackb(_)
'2100-09-02T00:55:02'
datetime.datetime supports instances with a tzinfo that is None,
datetime.timezone.utc, a timezone instance from the python3.9+ zoneinfo
module, or a timezone instance from the third-party pendulum, pytz, or
dateutil/arrow libraries.
datetime.time objects must not have a tzinfo.
>>> import ormsgpack, datetime
>>> ormsgpack.packb(datetime.time(12, 0, 15, 290))
b'\xaf12:00:15.000290'
>>> ormsgpack.unpackb(_)
'12:00:15.000290'
datetime.date objects will always serialize.
>>> import ormsgpack, datetime
>>> ormsgpack.packb(datetime.date(1900, 1, 2))
b'\xaa1900-01-02'
>>> ormsgpack.unpackb(_)
'1900-01-02'
Errors with tzinfo result in MsgpackEncodeError being raised.
To disable serialization of datetime objects specify the option
ormsgpack.OPT_PASSTHROUGH_DATETIME.
To use "Z" suffix instead of "+00:00" to indicate UTC ("Zulu") time, use the option
ormsgpack.OPT_UTC_Z.
To assume datetimes without timezone are UTC, use the option ormsgpack.OPT_NAIVE_UTC.
enum
ormsgpack serializes enums natively. Options apply to their values.
>>> import enum, datetime, ormsgpack
>>> class DatetimeEnum(enum.Enum):
... EPOCH = datetime.datetime(1970, 1, 1, 0, 0, 0)
...
>>> ormsgpack.packb(DatetimeEnum.EPOCH)
b'\xb31970-01-01T00:00:00'
>>> ormsgpack.unpackb(_)
'1970-01-01T00:00:00'
>>> ormsgpack.packb(DatetimeEnum.EPOCH, option=ormsgpack.OPT_NAIVE_UTC)
b'\xb91970-01-01T00:00:00+00:00'
>>> ormsgpack.unpackb(_)
'1970-01-01T00:00:00+00:00'
Enums with members that are not supported types can be serialized using
default:
>>> import enum, ormsgpack
>>> class Custom:
... def __init__(self, val):
... self.val = val
...
>>> def default(obj):
... if isinstance(obj, Custom):
... return obj.val
... raise TypeError
...
>>> class CustomEnum(enum.Enum):
... ONE = Custom(1)
...
>>> ormsgpack.packb(CustomEnum.ONE, default=default)
b'\x01'
>>> ormsgpack.unpackb(_)
1
float
ormsgpack serializes and deserializes double precision floats with no loss of precision and consistent rounding.
int
ormsgpack serializes and deserializes 64-bit integers by default. The range supported is a signed 64-bit integer's minimum (-9223372036854775807) to an unsigned 64-bit integer's maximum (18446744073709551615).
numpy
ormsgpack natively serializes numpy.ndarray and individual
numpy.float64, numpy.float32, numpy.float16,
numpy.int64, numpy.int32, numpy.int16, numpy.int8,
numpy.uint64, numpy.uint32, numpy.uint16, numpy.uint8,
numpy.uintp, numpy.intp, numpy.datetime64, and numpy.bool
instances.
numpy.datetime64 instances are serialized as RFC 3339 strings.
ormsgpack is faster than all compared libraries at serializing
numpy instances. Serializing numpy data requires specifying
option=ormsgpack.OPT_SERIALIZE_NUMPY.
>>> import ormsgpack, numpy
>>> ormsgpack.packb(
... numpy.array([[1, 2, 3], [4, 5, 6]]),
... option=ormsgpack.OPT_SERIALIZE_NUMPY,
... )
b'\x92\x93\x01\x02\x03\x93\x04\x05\x06'
>>> ormsgpack.unpackb(_)
[[1, 2, 3], [4, 5, 6]]
The array must be a contiguous C array (C_CONTIGUOUS) and one of the
supported datatypes.
If an array is not a contiguous C array or contains an supported datatype,
ormsgpack falls through to default. In default, obj.tolist() can be
specified. If an array is malformed, which is not expected,
ormsgpack.MsgpackEncodeError is raised.
Performance
---------------------------------------------------------------------------------- benchmark 'numpy float64': 2 tests ---------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_numpy_ormsgpack[float64] 77.9625 (1.0) 85.2507 (1.0) 79.0326 (1.0) 1.9043 (1.0) 78.5505 (1.0) 0.7408 (1.0) 1;1 12.6530 (1.0) 13 1
test_numpy_msgpack[float64] 511.5176 (6.56) 606.9395 (7.12) 559.0017 (7.07) 44.0661 (23.14) 572.5499 (7.29) 81.2972 (109.75) 3;0 1.7889 (0.14) 5 1
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------- benchmark 'numpy int32': 2 tests -------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_numpy_ormsgpack[int32] 197.8751 (1.0) 210.3111 (1.0) 201.1033 (1.0) 5.1886 (1.0) 198.8518 (1.0) 3.8297 (1.0) 1;1 4.9726 (1.0) 5 1
test_numpy_msgpack[int32] 1,363.8515 (6.89) 1,505.4747 (7.16) 1,428.2127 (7.10) 53.4176 (10.30) 1,425.3516 (7.17) 72.8064 (19.01) 2;0 0.7002 (0.14) 5 1
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------- benchmark 'numpy int8': 2 tests ---------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_numpy_ormsgpack[int8] 107.8013 (1.0) 113.7336 (1.0) 109.0364 (1.0) 1.7805 (1.0) 108.3574 (1.0) 0.4066 (1.0) 1;2 9.1712 (1.0) 10 1
test_numpy_msgpack[int8] 685.4149 (6.36) 703.2958 (6.18) 693.2396 (6.36) 7.9572 (4.47) 691.5435 (6.38) 14.4142 (35.45) 1;0 1.4425 (0.16) 5 1
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------- benchmark 'numpy npbool': 2 tests --------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_numpy_ormsgpack[npbool] 87.9005 (1.0) 89.5460 (1.0) 88.7928 (1.0) 0.5098 (1.0) 88.8508 (1.0) 0.6609 (1.0) 4;0 11.2622 (1.0) 12 1
test_numpy_msgpack[npbool] 1,095.0599 (12.46) 1,176.3442 (13.14) 1,120.5916 (12.62) 32.9993 (64.73) 1,110.4216 (12.50) 38.4189 (58.13) 1;0 0.8924 (0.08) 5 1
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------- benchmark 'numpy uint8': 2 tests ---------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_numpy_ormsgpack[uint8] 133.1743 (1.0) 134.7246 (1.0) 134.2793 (1.0) 0.4946 (1.0) 134.3120 (1.0) 0.4492 (1.0) 1;1 7.4472 (1.0) 8 1
test_numpy_msgpack[uint8] 727.1393 (5.46) 824.8247 (6.12) 775.7032 (5.78) 34.9887 (70.73) 775.9595 (5.78) 36.2824 (80.78) 2;0 1.2892 (0.17) 5 1
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
uuid
ormsgpack serializes uuid.UUID instances to
RFC 4122 format, e.g.,
"f81d4fae-7dec-11d0-a765-00a0c91e6bf6".
>>> import ormsgpack, uuid
>>> ormsgpack.packb(uuid.UUID('f81d4fae-7dec-11d0-a765-00a0c91e6bf6'))
b'\xd9$f81d4fae-7dec-11d0-a765-00a0c91e6bf6'
>>> ormsgpack.unpackb(_)
'f81d4fae-7dec-11d0-a765-00a0c91e6bf6'
>>> ormsgpack.packb(uuid.uuid5(uuid.NAMESPACE_DNS, "python.org"))
b'\xd9$886313e1-3b8a-5372-9b90-0c9aee199e5d'
>>> ormsgpack.unpackb(_)
'886313e1-3b8a-5372-9b90-0c9aee199e5d
Pydantic
ormsgpack serializes pydantic.BaseModel instances natively, with
duck-typing.
This is equivalent to serializing
model.model_dump(serialize_as_any=True) with Pydantic V2 or
model.dict()with Pydantic V1.
Performance
-------------------------------------------------------------------------------- benchmark 'pydantic': 2 tests ---------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_pydantic_ormsgpack 4.3918 (1.0) 12.6521 (1.0) 4.8550 (1.0) 1.1455 (3.98) 4.6101 (1.0) 0.0662 (1.0) 11;24 205.9727 (1.0) 204 1
test_pydantic_msgpack 124.5500 (28.36) 125.5427 (9.92) 125.0582 (25.76) 0.2877 (1.0) 125.0855 (27.13) 0.2543 (3.84) 2;0 7.9963 (0.04) 8 1
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Latency
Graphs
Data
----------------------------------------------------------------------------- benchmark 'canada packb': 2 tests ------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_packb[canada] 3.5302 (1.0) 3.8939 (1.0) 3.7319 (1.0) 0.0563 (1.0) 3.7395 (1.0) 0.0484 (1.0) 56;22 267.9571 (1.0) 241 1
test_msgpack_packb[canada] 8.8642 (2.51) 14.0432 (3.61) 9.3660 (2.51) 0.5649 (10.03) 9.2983 (2.49) 0.0982 (2.03) 3;11 106.7691 (0.40) 106 1
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------- benchmark 'canada unpackb': 2 tests --------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_msgpack_unpackb[canada] 10.1176 (1.0) 62.0466 (1.18) 33.4806 (1.0) 18.8279 (1.0) 46.6582 (1.0) 38.5921 (1.02) 30;0 29.8680 (1.0) 67 1
test_ormsgpack_unpackb[canada] 11.3992 (1.13) 52.6587 (1.0) 34.1842 (1.02) 18.9461 (1.01) 47.6456 (1.02) 37.8024 (1.0) 8;0 29.2533 (0.98) 20 1
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------- benchmark 'citm_catalog packb': 2 tests -----------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_packb[citm_catalog] 1.8024 (1.0) 2.1259 (1.0) 1.9487 (1.0) 0.0346 (1.0) 1.9525 (1.0) 0.0219 (1.0) 79;60 513.1650 (1.0) 454 1
test_msgpack_packb[citm_catalog] 3.4195 (1.90) 3.8128 (1.79) 3.6928 (1.90) 0.0535 (1.55) 3.7009 (1.90) 0.0250 (1.14) 47;49 270.7958 (0.53) 257 1
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------ benchmark 'citm_catalog unpackb': 2 tests ------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_unpackb[citm_catalog] 5.6986 (1.0) 46.1843 (1.0) 14.2491 (1.0) 15.9791 (1.0) 6.1051 (1.0) 0.3074 (1.0) 5;5 70.1798 (1.0) 23 1
test_msgpack_unpackb[citm_catalog] 7.2600 (1.27) 56.6642 (1.23) 16.4095 (1.15) 16.3257 (1.02) 7.7364 (1.27) 0.4944 (1.61) 28;29 60.9404 (0.87) 125 1
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------- benchmark 'github packb': 2 tests -----------------------------------------------------------------------------------
Name (time in us) Min Max Mean StdDev Median IQR Outliers OPS (Kops/s) Rounds Iterations
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_packb[github] 73.0000 (1.0) 215.9000 (1.0) 80.4826 (1.0) 4.8889 (1.0) 80.3000 (1.0) 1.1000 (1.83) 866;1118 12.4250 (1.0) 6196 1
test_msgpack_packb[github] 103.8000 (1.42) 220.8000 (1.02) 112.8049 (1.40) 4.9686 (1.02) 113.0000 (1.41) 0.6000 (1.0) 1306;1560 8.8649 (0.71) 7028 1
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------- benchmark 'github unpackb': 2 tests -----------------------------------------------------------------------------------
Name (time in us) Min Max Mean StdDev Median IQR Outliers OPS (Kops/s) Rounds Iterations
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_unpackb[github] 201.3000 (1.0) 318.5000 (1.0) 219.0861 (1.0) 6.7340 (1.0) 219.1000 (1.0) 1.2000 (1.0) 483;721 4.5644 (1.0) 3488 1
test_msgpack_unpackb[github] 289.8000 (1.44) 436.0000 (1.37) 314.9631 (1.44) 9.4130 (1.40) 315.1000 (1.44) 2.3000 (1.92) 341;557 3.1750 (0.70) 2477 1
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------- benchmark 'twitter packb': 2 tests ---------------------------------------------------------------------------------------
Name (time in us) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_packb[twitter] 820.7000 (1.0) 2,945.2000 (2.03) 889.3791 (1.0) 78.4139 (2.43) 884.2000 (1.0) 12.5250 (1.0) 4;76 1,124.3799 (1.0) 809 1
test_msgpack_packb[twitter] 1,209.3000 (1.47) 1,451.2000 (1.0) 1,301.3615 (1.46) 32.2147 (1.0) 1,306.7000 (1.48) 14.1000 (1.13) 118;138 768.4260 (0.68) 592 1
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------ benchmark 'twitter unpackb': 2 tests -----------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR Outliers OPS Rounds Iterations
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_ormsgpack_unpackb[twitter] 2.7097 (1.0) 41.1530 (1.0) 3.2721 (1.0) 3.5860 (1.03) 2.8868 (1.0) 0.0614 (1.32) 4;38 305.6098 (1.0) 314 1
test_msgpack_unpackb[twitter] 3.8079 (1.41) 42.0617 (1.02) 4.4459 (1.36) 3.4893 (1.0) 4.1097 (1.42) 0.0465 (1.0) 2;54 224.9267 (0.74) 228 1
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Reproducing
The above was measured using Python 3.7.9 on Azure Linux VM (x86_64) with ormsgpack 0.2.1 and msgpack 1.0.2.
The latency results can be reproduced using ./scripts/benchmark.sh and graphs using
pytest --benchmark-histogram benchmarks/bench_*.
Questions
Why can't I install it from PyPI?
Probably pip needs to be upgraded to version 20.3 or later to support
the latest manylinux_x_y or universal2 wheel formats.
Will it deserialize to dataclasses, UUIDs, decimals, etc or support object_hook?
No. This requires a schema specifying what types are expected and how to handle errors etc. This is addressed by data validation libraries a level above this.
Packaging
To package ormsgpack requires Rust 1.70
or newer and the maturin build
tool. The default feature unstable-simd enables the usage of SIMD
operations and requires nightly Rust. The recommended build command
is:
maturin build --release --strip
ormsgpack is tested on Linux/amd64, Linux/aarch64, Linux/armv7, macOS/amd64 and Windows/amd64.
There are no runtime dependencies other than libc.
License
orjson was written by ijl <ijl@mailbox.org>, copyright 2018 - 2021, licensed under both the Apache 2 and MIT licenses.
ormsgpack was forked from orjson by Aviram Hassan and is now maintained by Emanuele Giaquinta (@exg), licensed same as orjson.