如何在Jupyter Notebook中重定义Python3 print函数使flush默认设为True？

Your initial attempt at redefining print will cause infinite recursion (since your custom function calls itself directly), but there are two reliable ways to achieve what you want:

Method 1: Use functools.partial #

This is the simplest approach—we can create a wrapped version of the built-in print that sets flush=True as the default:

from functools import partial

# Override print with our version that flushes by default
print = partial(print, flush=True)

Now any call to print() will automatically include flush=True unless you explicitly set flush=False in the arguments.

Method 2: Create a Custom Wrapper Function #

If you want more flexibility (like letting users override the flush default when needed), write a wrapper that preserves access to the original print function:

# Save a reference to the original print before overriding it
original_print = print

def print(*args, **kwargs):
    # Set flush to True only if it's not already specified
    kwargs.setdefault('flush', True)
    original_print(*args, **kwargs)

This way, if someone calls print("Hello", flush=False), the custom function will respect that explicit False value, while using True by default.

Why This Fixes Your Jupyter Notebook Issue #

The out-of-order output between print() and input() is a classic buffer problem. By default, Python buffers standard output—meaning it holds onto printed text until a newline or buffer flush occurs. In Jupyter, this can lead to race conditions where input() triggers a flush before print() has written its content, messing up the order. Setting flush=True forces the print statement to send its content to the output immediately, ensuring your messages appear in the expected sequence every time.

内容的提问来源于stack exchange，提问作者Jo Simon