MessagePack is an efficient binary serialization format. It lets you exchange data among multiple languages like JSON. But it's faster and smaller. This package provides CPython bindings for reading and writing MessagePack data.
$ pip install msgpack
The extension module in msgpack (msgpack._cmsgpack) does not support PyPy.
But msgpack provides a pure Python implementation (msgpack.fallback) for PyPy.
If you can't use a binary distribution, you need to install Visual Studio or the Windows SDK on Windows. Without the extension, the pure Python implementation on CPython runs slowly.
Use packb for packing and unpackb for unpacking.
msgpack provides dumps and loads as aliases for compatibility with
json and pickle.
pack and dump pack to a file-like object.
unpack and load unpack from a file-like object.
>>> import msgpack
>>> msgpack.packb([1, 2, 3])
'\x93\x01\x02\x03'
>>> msgpack.unpackb(_)
[1, 2, 3]Read the docstring for options.
Unpacker is a "streaming unpacker". It unpacks multiple objects from one
stream (or from bytes provided through its feed method).
import msgpack
from io import BytesIO
buf = BytesIO()
for i in range(100):
buf.write(msgpack.packb(i))
buf.seek(0)
unpacker = msgpack.Unpacker(buf)
for unpacked in unpacker:
print(unpacked)It is also possible to pack/unpack custom data types. Here is an example for
datetime.datetime.
import datetime
import msgpack
useful_dict = {
"id": 1,
"created": datetime.datetime.now(),
}
def decode_datetime(obj):
if '__datetime__' in obj:
obj = datetime.datetime.strptime(obj["as_str"], "%Y%m%dT%H:%M:%S.%f")
return obj
def encode_datetime(obj):
if isinstance(obj, datetime.datetime):
return {'__datetime__': True, 'as_str': obj.strftime("%Y%m%dT%H:%M:%S.%f")}
return obj
packed_dict = msgpack.packb(useful_dict, default=encode_datetime)
this_dict_again = msgpack.unpackb(packed_dict, object_hook=decode_datetime)Unpacker's object_hook callback receives a dict; the
object_pairs_hook callback may instead be used to receive a list of
key-value pairs.
NOTE: msgpack can encode datetime with tzinfo into standard ext type for now.
See datetime option in Packer docstring.
It is also possible to pack/unpack custom data types using the ext type.
>>> import msgpack
>>> import array
>>> def default(obj):
... if isinstance(obj, array.array) and obj.typecode == 'd':
... return msgpack.ExtType(42, obj.tostring())
... raise TypeError("Unknown type: %r" % (obj,))
...
>>> def ext_hook(code, data):
... if code == 42:
... a = array.array('d')
... a.fromstring(data)
... return a
... return ExtType(code, data)
...
>>> data = array.array('d', [1.2, 3.4])
>>> packed = msgpack.packb(data, default=default)
>>> unpacked = msgpack.unpackb(packed, ext_hook=ext_hook)
>>> data == unpacked
TrueAs an alternative to iteration, Unpacker objects provide unpack,
skip, read_array_header, and read_map_header methods. The former two
read an entire message from the stream, respectively deserializing and returning
the result, or ignoring it. The latter two methods return the number of elements
in the upcoming container, so that each element in an array, or key-value pair
in a map, can be unpacked or skipped individually.
Early versions of msgpack didn't distinguish string and binary types. The type for representing both string and binary types was named raw.
You can pack into and unpack from this old spec using use_bin_type=False
and raw=True options.
>>> import msgpack
>>> msgpack.unpackb(msgpack.packb([b'spam', 'eggs'], use_bin_type=False), raw=True)
[b'spam', b'eggs']
>>> msgpack.unpackb(msgpack.packb([b'spam', 'eggs'], use_bin_type=True), raw=False)
[b'spam', 'eggs']To use the ext type, pass a msgpack.ExtType object to the packer.
>>> import msgpack
>>> packed = msgpack.packb(msgpack.ExtType(42, b'xyzzy'))
>>> msgpack.unpackb(packed)
ExtType(code=42, data='xyzzy')You can use it with default and ext_hook. See below.
When unpacking data received from an unreliable source, msgpack provides two security options.
max_buffer_size (default: 100*1024*1024) limits the internal buffer size.
It is also used to limit preallocated list sizes.
strict_map_key (default: True) limits the type of map keys to bytes and str.
While the MessagePack spec doesn't limit map key types,
there is a risk of a hash DoS.
If you need to support other types for map keys, use strict_map_key=False.
CPython's GC starts when the number of allocated objects grows.
This means unpacking may trigger unnecessary GC.
You can use gc.disable() when unpacking a large message.
A list is the default sequence type in Python.
However, a tuple is lighter than a list.
You can use use_list=False while unpacking when performance is important.
The package name on PyPI was changed from msgpack-python to msgpack in 0.5.
When upgrading from msgpack-0.4 or earlier, do pip uninstall msgpack-python before
pip install -U msgpack.
-
Python 2 support
-
The extension module no longer supports Python 2. The pure Python implementation (
msgpack.fallback) is used for Python 2. -
msgpack 1.0.6 drops official support of Python 2.7, as pip and GitHub Action "setup-python" no longer supports Python 2.7.
-
-
Packer
- Packer uses
use_bin_type=Trueby default. Bytes are encoded in the bin type in MessagePack. - The
encodingoption is removed. UTF-8 is always used.
- Packer uses
-
Unpacker
- Unpacker uses
raw=Falseby default. It assumes str values are valid UTF-8 strings and decodes them to Python str (Unicode) objects. encodingoption is removed. You can useraw=Trueto support old format (e.g. unpack into bytes, not str).- The default value of
max_buffer_sizeis changed from 0 to 100 MiB to avoid DoS attacks. You need to passmax_buffer_size=0if you have large but safe data. - The default value of
strict_map_keyis changed to True to avoid hash DoS. You need to passstrict_map_key=Falseif you have data that contain map keys whose type is neither bytes nor str.
- Unpacker uses
