lambdser 0.1.2

fast lambda expression serializer

lambdser

A lambda expression serializer for python. Can be used to make pickle eat lambdas with closures.

A typical use case is serializing lambdas for multiprocessing. Using lambdser in front, let multiprocessing eat the lambda expression.

Install

pip install lambdser

or install it from github

pip install git+https://github.com/cloasdata/lambdser.git

or just clone it

todo

I did not find a way to register lambdser es pickler in the pickle module. It would be really useful if somebody can help me. However, one can use the LambdserPickler class to overwrite the default behaviour of pickle.Pickler (?) or use LamdserPickler as the one pickler.

I also did not test for particular edge cases. But feel free to contribute such tests.

usage

Example 1: module namespace

using it in module namespace.

    import pickle
    import lambdser
    
    
    two = "2"
    expression = lambda x: "number" + x + two
    
    result1 = expression("4")
    ser = lambdser.dumps(expression)
    # now pickle can dump!
    s = pickle.dumps(ser)
    ser = pickle.loads(s)
    
    expression = lambdser.loads(ser)
    result2 = expression("4")
    assert result1 == result2

Example 2: Using closure

Make a proxy of what you want to spawn in a separate process.

    import lambdser
    import multiprocessing as mp
    
    
    def make_proxy(para, *funcs):
        # make proxy for the mp
        ser_list = []
        for f in funcs:
            ser_list.append(lambdser.dumps(f))
        return para, ser_list
    
    
    def processor(*ser):
        # unzip the proxy and to the work
        para, funcs = ser
        funcs = [lambdser.loads(ser) for ser in funcs]
        res = None
        for f in funcs:
            res = f(para)
        print(res)
        return res
    
    
    def do_stuff():
        two = "2"
        ser = make_proxy("4", lambda x: x + two)
        mp.Process(target=processor, args=ser).start()

    do_stuff()

performance

it is around 100 times faster than dill. This was one reason to develop this package.

py .\test\profiling.py

Results dumps
lambdser: 0.012459
dill:     1.485589
times:    119.236291

Copyright © 2022 Simon Bauer

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.