DEV Community

Dead Simple Python: Lambdas, Decorators, and Other Magic

Jason C. McDonald on August 07, 2019

Like the articles? Buy the book! Dead Simple Python by Jason C. McDonald is available from No Starch Press. Python has a reputation for looking...

Read full post

Anna R Dunster • Aug 21 '20

Curious about "Functions should not have side effects; they should not modify anything external to themselves." If I had a function designed to take a dictionary as an argument, with the intention that other relevant functions would pass in a dictionary generated dynamically at each call, rather than an existing global, how much does it matter if the function mutates the dictionary to produce its results? (In this case, writing to an SQL database.) Should I bulletproof it, maybe by making a full copy of the input dictionary inside the function, warn about the behavior in the docstring, or does it not matter? I'd be surprised if anyone but me ever uses this code, but, who knows XD

Jason C. McDonald • Aug 21 '20

That would still be considered impure, from a functional standpoint. If it makes sense in your code base, you might be able to get away with it, but it's still not recommended, as your unspoken intended rule may not be respected by future-you.

Anna R Dunster • Aug 24 '20

Probably worth doing something to idiot proof it then, who knows who might try to do something with my private project in the future o.o XD Better habit, anyway!

Jason C. McDonald • Aug 24 '20 • Edited

Usually the worst future idiots using our projects are our own overworked selves.

Anna R Dunster • Aug 24 '20

That's exactly who I'm most worried about 😅

Sandor Dargo • Aug 7 '19

Very useful article.

Regarding the default depth, it's interesting. Here and there I read 997, then I went to see the CPython implementation, there it's set to 1000.

Well, it doesn't change a lot of things. But probably it's worth mentioning that you actually change that limit, by using sys.setrecursionlimit.

Jason C. McDonald • Aug 7 '19

Great insight, thanks!

Ornataweaver • Jan 27 '20

There will be no "Transfiguring frog into frog-orange." since you redefined 'partial_transfiguration'.
Thanks for the useful article.

Jason C. McDonald • Jan 27 '20

Where specifically is it redefined?

Ornataweaver • Jan 29 '20

@party_cannon
def partial_transfiguration(target, combine_with):
    return f"{target}-{combine_with}"

😅

Ornataweaver • Jan 29 '20

but before this, it was:

def partial_transfiguration(target, combine_with):
    result = f"{target}-{combine_with}"
    print(f"Transfiguring {target} into {result}.")
    return result

Jason C. McDonald • Jan 29 '20 • Edited

Ah, following you now. I've been working in C++ for the past couple of weeks, so "redefined" had a different connotation for me.

I had indeed removed the print statement in the second version! I'll go back and fix that. Thanks for the catch.

Eric Ahnell • Aug 8 '19

Great explanation of various concepts I see all the time in Python code, but wasn't sure how, or if, I wanted to make use of my own code. Now I can!

Mujeeb Ishaque • Aug 7 '19

I love this guy. Respects.

Ornataweaver • Jan 18 '20

Great article, thanks a lot.
I suggest reading dev.to/yawpitch/the-35-words-you-n... to people who like this article.