Understanding Monkey Patching in Python: Replacing json with ujson for Faster Serialization

What is a monkey patch? Monkey patch (Monkey Patch) is a technique that dynamically replaces attributes or modules at runtime, effectively adding or changing program functionality.

Monkey Patch functionality overview allows modifying a class or module while the program is running.

Example shows replacing the standard json module with ujson using a simple monkey patch.

<code>"""
file:json_serialize.py
"""
import time
import json

# 时间测试装饰器
def run_time(func):
    def inner(*args, **kwargs):
        start_time = time.time()
        result = func(*args, **kwargs)
        end_time = time.time()
        print(f'程序用时：{end_time - start_time}')
        return result
    return inner

@run_time
def json_dumps(obj):
    return json.dumps(obj)

# 生成测试字典
test_dict = {i: 1 for i in range(1, 10000001)}
</code>

Running the original program with json.dumps is relatively slow. The following script measures the time taken to serialize a large dictionary using the standard json module.

<code>"""
file:run.py
"""
from json_serialize import json_dumps, test_dict

print(f'json.dumps编码用时:', end='')
r1 = json_dumps(test_dict)
</code>

By applying a monkey patch, json.dumps is replaced with ujson.dumps , which significantly speeds up serialization.

<code>"""
file:run.py
"""
import json
import ujson
from json_serialize import json_dumps, test_dict

def monkey_patch_json():
    json.dumps = ujson.dumps

monkey_patch_json()
print(f'使用猴子补丁之后json.dumps编码用时:', end='')
json_dumps(test_dict)
</code>

After the patch, the json module in the project is effectively swapped for ujson, demonstrating how a monkey patch can globally affect a process and improve performance.