esoteric-python

[restive void]

How so? You find it unintuitive that Python doesn't have block scoping?

[versed eagle]

i find it unintuitive that it has scoping for some things but not others

to me, the bindings for the case statement read a bit like function args

in general, they should go out of scope whenever that section is done

in my mind, that is

so is unintuitive to me that they don't

[restive void]

Why only match-case, and not with, for, if, ...?

[versed eagle]

it's not only match-case, but that was the subject at hand so i was talking about it specifically

[fleet bridge]

It is not new.
del x exists for a long time
Also exc in except Exception as exc: is "unbounded" (assigned to None) after except block finishes execution

[vast wave]

!e ```py
try:
raise Exception()
except Exception as e:
pass

print(e)

[night quarry]

@vast wave :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 6, in <module>
003 |     print(e)
004 |           ^
005 | NameError: name 'e' is not defined

[fleet bridge]

!e ```py
def f():
try:
raise Exception()
except Exception as e:
pass
print(e)
f()

[night quarry]

@fleet bridge :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 7, in <module>
003 |     f()
004 |   File "/home/main.py", line 6, in f
005 |     print(e)
006 |           ^
007 | UnboundLocalError: cannot access local variable 'e' where it is not associated with a value

[fleet bridge]

So there is essentially del exc at the end of the block, not exc = None

[vast wave]

what in tarnation

!e ```py
def f():
a = "abc"
print(a)
del a

print(a)

f()
b = "bcd"
print(b)
del b
print(b)

oh right

[night quarry]

@vast wave :x: Your 3.11 eval job has completed with return code 1.

001 | abc
002 | bcd
003 | Traceback (most recent call last):
004 |   File "/home/main.py", line 10, in <module>
005 |     print(b)
006 |           ^
007 | NameError: name 'b' is not defined

[vast wave]

what the fuck

del in a function != del in a module

i love python

[fleet bridge]

Global namespace and local namespace are behaving very differently

[restive void]

True, didn't think of those two things

[quartz wave]

there's both of them

essentially ```py
exc = None
del exc

[keen thicket]

!e

f = lambda n, w, b: ''.join(reversed([chr(b+((n>>i)&(2**w-1))) for i in range(0, n.bit_length(), w)]))
print(f(198522, 4, 101), f(20297984, 5, 100))

[night quarry]

@keen thicket :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello world

[pure dew]

oh did we see that lecture video here too?

[keen thicket]

!e

string = "Hello World!"
words = string.split(" ")
enc_words = []
for word in [i.strip() for i in words]:
    letters = [ord(i) for i in word]
    b=min(letters)
    delta = [w-b for w in letters]
    w = max(delta).bit_length()
    if len(letters) == 1:
        enc_words.append(((1<<b.bit_length()), b.bit_length(), b))
        continue
    n=int("1"+''.join(([f'{i:0{w}b}'for i in delta])), 2)
    enc_words.append((n, w, b))

from sys import getsizeof
f=lambda n,w,b:''.join((chr(b+(n>>i&2**w-1))for i in reversed(range(0,n.bit_length()-1,w))))
print(' '.join([f(*i) for i in enc_words]))
print(sum([sum([getsizeof(z) for z in i]) for i in enc_words]))
print(sum([getsizeof(i)for i in string]))

[night quarry]

@keen thicket :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | Hello World!
002 | 176
003 | 600

[keen thicket]

compression using deltas for each word to convert the string into an initial value, an integer describing the deltas, and the width in bits of each delta 🙂

[versed eagle]

fun

[keen thicket]

I wonder if it's possible to make a more efficient packer for python source code by using all the kws and soft kws

[leaden valley]

can undefined be any variable name?

[frail token]

Yes, there is nothing special about the variable name 'undefined', it is not a Python keyword or singleton (like None, True or False).

[arctic plume]

!e

print("".join([chr(round(-4412.70555564*x+12008.5182542*x**2-13041.0511356*x**3+7566.06024046*x**4-2619.91695608*x**5+567.42766205*x**6-77.5228009278*x**7+6.48776455043*x**8-0.303552138456*x**9+0.00607914462098*x**10+104)) for x in range(11)]))```

[night quarry]

@arctic plume :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello world

[keen thicket]

what the fuck?

is that like a fitted curve to the ascii values or something

[digital mesa]

using list comprehension instead of generator expression 🙄

[keen thicket]

just use map

[digital mesa]

nah

extra chars for no reason

[keen thicket]

ur right but I'm gonna rewrite it shorter using maps to prove u wrong

[digital mesa]

😭

[digital mesa]

competing with 239 chars

adjusted to be without [], and no space between ) and for

[arctic plume]

:)

You can try putting it into desmos if you want

[digital mesa]

only to the chars of hello world I presume

otherwise the fitted curve would be a straight line, no?

[arctic plume]

You can think of it as a function f(x) that you run for every value of x between 0 and 10

Actually, what am I even saying, this is esoteric python, you figure it out!

[digital mesa]

I mean it's not too complicated

just looks complicated

[arctic plume]

A lot of numbers indeed

[digital mesa]

now write a one-liner that will generate a function for a given string 😂

[arctic plume]

Lmao

[digital mesa]

L bozo

get error-slapped

[arctic plume]

It should be possible, although I'm not familiar with non-linear regression algorithms...

[digital mesa]

me neither 🤫

[night quarry]

@sick hound :x: Your 3.11 eval job has completed with return code 1.

001 |   File "/home/main.py", line 7
002 |     [chr(ord(b) + 1) for b in a]) for a in x][0])(g, 12)))
003 |                                                         ^
004 | SyntaxError: closing parenthesis ')' does not match opening parenthesis '[' on line 2

[tough willow]

how did you find out what to be multiplied?

[arctic plume]

A magician never reveals his secrets

[tough willow]

okay

[arctic plume]

You could probably remove some decimals and have it still work actually

If only it was more readable

Oh wait...

[arctic plume]

@manic ember

5 minutes perhaps?

[keen thicket]

!e

print("".join(map(lambda x:chr(round((-4412.70*x+12008.518*x**2-13041.0511*x**3+7566.0602*x**4-2619.91695*x**5+567.427662*x**6-77.52280092*x**7+6.48776455*x**8-0.3035521384*x**9+0.00607914462*x**10+104))),range(11))))

[night quarry]

@keen thicket :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello world

[keen thicket]

I had to resort to chopping off a lot of the decimals that weren't needed lol

but it's 217

gottem 🙂

[arctic plume]

It's that bad?

[digital mesa]

Nahhh that's cheating

[keen thicket]

yeah no without that it was 250

I tried a lot of stuff

[digital mesa]

Lol

[keen thicket]

but it really is just wasted chars

but it was a fun exercise

[arctic plume]

It's possible that there's an equation, say using a^x that would be shorter

[keen thicket]

there is actually: ```python
print("hello world")

[arctic plume]

What if I forget what how to spell it?

But remember the polynomial equation for it

[keen thicket]

you won't believe it but I just saw this program that can take the polynomial equation and create the secret code word string

HAHAHAHAHA

needs more lambdas for sure

Wait that would be funny, imagine if you made a program that messed with the code object so that when you viewed the backtrace it would be different

[fleet bridge]

Ohh, lambda calculus in python...

Literally

[keen thicket]

Then you could make an infinite lambda scroller

at the top and bottom

[arctic plume]

You can just spell hello world with the lambda argument names at that point...

Already have the majority of the alphabet lmao

[fleet bridge]

It is the entire alphabet

[arctic plume]

Forgot åäö

[keen thicket]

youjust gave me an amazing idea

[arctic plume]

Uh oh...

[tough willow]

what about ü?

[arctic plume]

Don't be ridiculous
Next you're gonna tell me that the super s ß is needed as well

[tough willow]

ß and ẞ

[arctic plume]

ẞ haha looks so silly

[tough willow]

that's the capitalized ß

[night quarry]

@sick hound :x: Your 3.11 eval job has completed with return code 1.

001 |   File "/home/main.py", line 1
002 |     (lambda a: (lambda b: (lambda c: (lambda d: (lambda e: (lambda f: (lambda g: (lambda h: (lambda i: (lambda j: (lambda k: (lambda l: (lambda m: (lambda n: (lambda o: (lambda p: (lambda q: (lambda r: (lambda s: (lambda t: (lambda u: (lambda v: (lambda w: (lambda x: (lambda y: (lambda z: (lambda a1: (lambda b1: (lambda c1: (lambda d1: (lambda e1: (lambda f1: (lambda g1: (lambda h1: (lambda i1: (lambda j1: (lambda k1: (lambda l1: (lambda m1: (lambda n1: (lambda o1: (lambda p1: (lambda q1: (lambda r1: (lambda s1: (lambda t1: (lambda u1: (lambda v1: (lambda w1: (lambda x1: (lambda y1: (lambda z1: __import__('sys').stdout.buffer.write(bytes(map(lambda x: round((a(x)+b(x)*x+c(x)**2+d(x)**3+e(x)**4+f(x)**5+g(x)**6+h(x)**7+i(x)**8+j(x)**9+k(x)**10+l(x)**11+m(x)**12+n(x)**13+o(x)**14+p(x)**15+q(x)**16+r(x)**17+s(x)**18+t(x)**19+u(x)**20+v(x)**21+w(x)**22+z(x)**23), range(24)))))))))))))))))))))))))))))))))))))))))))))))))))))))))(lambda a: 0))(lambda
... (truncated - too long)

Full output: https://paste.pythondiscord.com/cufulasabo.txt?noredirect

[last locust]

!e ```py
from fishhook import hook

@hook(int)
def bool(self):
print("hi", self)
return True

print(bool(0))```Why is there two prints for the __bool__ here? Where does the first one come from?

[night quarry]

@last locust :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | hi 2147483648
002 | hi 0
003 | True

[kindred stone]

probably from here https://github.com/python/cpython/blob/30488fa22a8f63753192ae7f1d01665857764e45/Lib/test/test_int_literal.py#L33

[night quarry]

Lib/test/test_int_literal.py line 33

self.assertEqual(0x80000000, 2147483648)```

[kindred stone]

oh nvm tests arent run

🤦‍♂️

[last locust]

Heh, nice idea though lol

It's exactly 2**31, so I'm guessing the number has some significance

[kindred stone]

https://github.com/search?q=repo%3Apython%2Fcpython+2147483648+NOT+path%3A%2F^Lib\%2Ftest\%2F%2F&type=code

nothing is happening with the literal number

[last locust]

I've just got no clue why the __bool__ is being run twice in the first place

And the 2147483648 seems so random, albeit 2**31

[kindred stone]

2**31 - 1 is the largest a int32 can be

maybe it is called on a checking if the system is 32 or 64 bits

(this is a blind guess)

[last locust]

Wouldn't it be 2**32 - 1?

[kindred stone]

the MSB is used to say if the number is negative or positive, so you cant have 2 ** 32 of value

the rest is used to represent the value of the number (2 ** 31 - 1)

[last locust]

Right

[last locust]

Still seems random that __bool__ is called twice

[digital mesa]

!e

from fishhook import hook

@hook(int)
def __bool__(self):
    print("hi", self)
    return True

print(bool(0))
print(bool(1))

[night quarry]

@digital mesa :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | hi 2147483648
002 | hi 0
003 | True
004 | hi 1
005 | True

[digital mesa]

only once

then later calls are not printing the extra

!e

from fishhook import hook

@hook(int)
def __bool__(self):
    print("hi", self)
    return True

[night quarry]

@digital mesa :white_check_mark: Your 3.11 eval job has completed with return code 0.

hi 2147483648

[digital mesa]

woah funky

although that is what I was suspecting after the previous test

[last locust]

Yeah, why does this happen?

It's getting triggered without actually doing anything

[digital mesa]

🤷‍♂️

[fleet bridge]

1. it is happening in fishhook.hook after method is hooked
2. (unlikely) it is happening when new name is assigned to global namespace

[rugged sparrow]

https://github.com/chilaxan/fishhook/blob/master/fishhook/fishhook.py#L55-L59 it happens inside this function, which is called from lock(cls) in force_setattr, which is called by hook(...)

[night quarry]

fishhook/fishhook.py lines 55 to 59

def assert_cls(o):
    if type(o).__flags__ & (1 << 31):
        return o
    else:
        raise RuntimeError('Invalid class or object')```

[last locust]

Hmm, I see, thanks

[rugged sparrow]

Yea due to how fishhook hooks internals it's difficult to make it fully opaque (until I implement scoped hooks but that is a while away)

[keen thicket]

!e

h = [
    lambda a, b, c, d: 0,
    lambda a: 0,
    lambda a, b, c, d, e, f, g, h: 0,
    lambda a, b, c, d, e, f, g, h: 0,
    lambda a, b, c, d, e, f, g, h, i, j, k: 0
]

w = [None,
    lambda a, b, c, d, e, f, g, h, i, j, k, l, m, n, o, p, q, r, z,_: 0,
    lambda a, b, c, d, e, f, g, h, i, j, k,_: 0,
    lambda a, b, c, d, e, f, g, h, i, j, k, l, m, z,_: 0,
    lambda a, b, c, d, e, f, g, h,_: 0,
    lambda a: 0
]

b = 100
for i in h+w:
    if i is None:
        print(" ",end="")
        b-=1
        continue
    try:
        i()
    except Exception as exc:
        print(chr(int(str(exc).split(" ")[2])+b),end="")

[night quarry]

@keen thicket :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello world

[arctic plume]

Haha, classic str(exc)

[keen thicket]

hahaha it took so long to make the arguments match the right number

[quartz wave]

!e ```py
a = 2; print(a + 3)

[night quarry]

@quartz wave :x: Your 3.11 eval job has completed with return code 1.

001 |   File "/home/main.py", line 1
002 |     a = 2; print(a + 3)
003 |         ^
004 | SyntaxError: invalid decimal literal

[quartz wave]

figure out why

[astral rover]

Cause it's not a real colon?

!e ```py
a = 2; print(a + 3)

[night quarry]

@astral rover :white_check_mark: Your 3.11 eval job has completed with return code 0.

5

[quartz wave]

ok

[astral rover]

Is that not the reason?

[quartz wave]

it's correct

"ok" would mean "all correct" here like it always did

[versed eagle]

my guess is unicode fuckery of some type

[quartz wave]

!charinfo ;

[night quarry]

Character Info

\u037e : GREEK QUESTION MARK - ;

[quartz wave]

yes

[keen thicket]

fuck u but I'm gonna do it just give me like a few more hours smfh

i knew someone was gonna tell me to make it one line

wait is it even possible to do error handling in a list comp

[languid hare]

afaik there is no try/catch statement equivalent as an expression

[keen thicket]

hmm

I'm sure there's a way around it but tbh my code relies on the error catching

because it knows the ascii char based on the offset in the error

[digital mesa]

Perhaps as an import

[low lynx]

with a cursed exec

[rugged sparrow]

You can do it using context decorators

[languid hare]

👀

ill have a dig

[keen thicket]

would this be with a decorator? I thought you had to like manually wrap the func in an expr

but I've never tried

like val = wrapper(func)(args) vs ```python
@wrapper
def func(args):
pass

[thick adder]

A wrapper wraps the function. The thing you're using is better classified as a decorator.

[keen thicket]

I'm confused what you're saying. I'm illustrating the two examples, wrapper or decorator

decorators do not work in python with lambdas

therefore not expressions

so I was asking if there was another way around it besides just using the wrapper form of a decorator

But a decorator is just syntax sugar for a wrapper, so I'm still confused

[restive void]

The definition is not that clear. One could certainly say that e.g. functools.wraps is a decorator. There's decorator syntax which is what you're talking about. Often people also call decorator factories decorators.

[finite coral]

Why does this work? f'{5:{2}}' in python 3.10.11. Shouldn't this throw a type exception for using a set to specify a width? ... and how would I left align the number while keeping those unnecessary confusing inner {}?

[plucky plover]

!e
there is no set being used there
that is f-string syntax for specifying a dynamic width for the given value
that's equivalent to

value= 5
width = 2
print(f"---{value:{width}}---")

[night quarry]

@plucky plover :white_check_mark: Your 3.11 eval job has completed with return code 0.

--- 5---

String formatting mini-language

The String Formatting Language in Python is a powerful way to tailor the display of strings and other data structures. This string formatting mini language works for f-strings and .format().

Take a look at some of these examples!

>>> my_num = 2134234523
>>> print(f"{my_num:,}")
2,134,234,523

>>> my_smaller_num = -30.0532234
>>> print(f"{my_smaller_num:=09.2f}")
-00030.05

>>> my_str = "Center me!"
>>> print(f"{my_str:-^20}")
-----Center me!-----

>>> repr_str = "Spam \t Ham"
>>> print(f"{repr_str!r}")
'Spam \t Ham'

Full Specification & Resources
String Formatting Mini Language Specification
pyformat.info

[finite coral]

First of, thank you so much for clearing up my confusion. So the question I need to figure out is: how to left align a number using a dynamic width?

oh I got it, it is just f'{5:<{2}}' I could have sworn I tried this earlier. Well, cheers!

[low lynx]

!e

width = 5
print(f'{32:#<{width}}')

wait what

oh left aligned obviously wouldn't show that

[night quarry]

@low lynx :white_check_mark: Your 3.11 eval job has completed with return code 0.

32###

[keen thicket]

decorators are just syntax though, all they're doing under the hood is automatically wrapping the function. I think you're confused because you haven't read the PEP: https://peps.python.org/pep-0318/

oh wait you aren't the original person who replied oops

well still read the PEP, it's pretty unambiguous about the name being specific to the syntax introduced in that PEP, not the general concept of wrapping

[restive void]

Words aren't only given meaning by definitions, but by usage. :)

Even in the official docs I find cases where functions/classes are being referred to as decorators.

[keen thicket]

my point isn’t that nobody uses decorators, i’m saying they’re just another way of saying wrapper

they’re the same to the person quibbling about my inconsistent in their eyes usage

my original question was asking about the syntax differences between two scenarios that accomplish the same thing, and whether there was a third way to do it I didn’t know of

and they were saying “well actually they’re not the same”

or not called the same

but my question was about syntax and i frankly don’t give a shit if they’re decorators or wrappers considering no one even can define the difference

[keen thicket]

like this is what I replied to

Just completely pointless hair splitting irrelevant to what I was saying considering my example was clearly contrasting wrapping vs decorator syntax

[fleet bridge]

Not all decorators are wrappers
Some decorators return the same function (with patched fields), so function is not wrapped into other function

[languid hare]

or as a "registration", like discordpy events or flask paths

[keen thicket]

they return the same function, but it still has to go through the wrapping function, even though it doesn’t modify it

also as I said incredibly explicitly, I was asking a question about syntax, not the exact definition of wrapper and decorator

can everyone please stop replying to me with more nit picks about my verbiage

I just wanna know if there’s a third thing besides wrapper or decorator, that achieves the same thing with a different syntax

the exact delineation between wrapper and decorator will probably be resolved by scientists in the coming millennia but until then I really don’t care

[fleet bridge]

this is an example of wrapper: ```py
def wrapper(*args, **kwargs):
# do something before call
orig_func(*args, **kwargs)
# do something after call

it is a function that does something before/after call and decides to call or not call the original function

if function is returned unchanged, then it is not wrapped

[keen thicket]

that’s your definition of a wrapper

and also bro

i just said I don’t care about this discussion at all

[fleet bridge]

im reading your messages, please wait...

[keen thicket]

this one was before you started replying to me

[restive void]

Then I don't understand your question, @keen thicket. You are asking if there's another way to achieve the same thing, other than using the @ syntax and directly calling the function-often-used-with-decorator-syntax on a function?

[keen thicket]

yes that’s exactly what I’m asking

I’m asking if there’s some syntax feature I don’t know about that is like a (decorator|wrapper)

that would work in expressions

because the context was someone asking me to turn my error catching thing into one line

[restive void]

Only what you showed above, doing foo(lambda x: ...) instead of

@foo
def bar(x):
    ...

[keen thicket]

darn i thought there would be another way

it be like that

[fleet bridge]

decorators just calls one callable with one argument (which is another callable)
so there is only one way to do that

there are two syntaxes:

• @deco def f(...)
• f = deco(f)

you can do something like deco.__call__(f) (or similar #esoteric-python stuff) but it is weird

[keen thicket]

yeah i listed the two syntaxes in my question

the call thing seems good tho

that’s useful

my question was never about what a decorator or wrapper is

i know how to make those

[fleet bridge]

!e ```py
@lambda f: lambda: (print('before'), f(), print('after'))
@lambda _: lambda: print('hello world')
class X: ...

print(X)
X()

[night quarry]

@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | <function <lambda>.<locals>.<lambda> at 0x7f72caf10680>
002 | before
003 | hello world
004 | after

[keen thicket]

someone just made a pointless quibble that I argued with

and now we’re here

generally it’s important to read the context because if you had you’d have known I was asking about syntax, never cared about the definition of words

and also had asked everyone to stop replying to me about the pointless semantics debate before you replied

but I’m glad that I finally got my question answered

[digital mesa]

!e

@sum
class total(metaclass=type("", (), {"__new__": lambda *_: range(10)})):
    pass


print(total)

[night quarry]

@digital mesa :white_check_mark: Your 3.11 eval job has completed with return code 0.

45

[fleet bridge]

!e ```py
class X(slice(1,2,3)): ...
print(X)

[night quarry]

@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.

slice('X', (slice(1, 2, 3),), {'__module__': '__main__', '__qualname__': 'X'})

[azure basalt]

how would one obfuscate a simple quiz game in python?

like this one ```py
print('who owns tesla?')
a = input('> ')
if a == elon musk
print('correct')
else
print('wrong')

i actually need to obfuscate the code because im gonna make a quiz game for some friends of mine and they will surely try to look at the answers

[keen thicket]

load the answers from a password protected zip archive or something each time the application starts up?

then you would just need to obfuscate the password

[fleet bridge]

!e ```py
exec(import('zlib').decompress(bytes.fromhex('789c2b28cacc2bd150ca48cdc9c95728cf2fca4951d204004e93071f')).decode())

pybot dead?

rip snek

[night quarry]

@fleet bridge You've already got a job running - please wait for it to finish!

Sorry, an unexpected error occurred. Please let us know!

TimeoutError:

[digital mesa]

!e ```py
exec(import('zlib').decompress(bytes.fromhex('789c2b28cacc2bd150ca48cdc9c95728cf2fca4951d204004e93071f')).decode())

[night quarry]

@digital mesa :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello world

[versed eagle]

a decorator is not necessarily a wrapper

it can return something entirely unrelated to the original thing

[versed eagle]

use bytecode

most people don't know how to read it

[fleet bridge]

it will contain raw strings, so it is not good for this purpose

[versed eagle]

not if you build the strings

in the program

instead of LOAD_CONST

[fleet bridge]

i believe you can open .pyc file with notepad and see raw strings in the mess

[versed eagle]

.

[fleet bridge]

yeah, that will work

[versed eagle]

now i kinda wanna make a bytecode obfuscator

[fleet bridge]

it is also possible to use any obfuscator to obfuscate code (like pyarmor)

[gleaming linden]

Hash the answers in the code, no need to obfuscate the whole thing

[azure basalt]

how do you hash?

[gleaming linden]

!d hashlib

[night quarry]

hashlib

Source code: Lib/hashlib.py

This module implements a common interface to many different secure hash and message digest algorithms. Included are the FIPS secure hash algorithms SHA1, SHA224, SHA256, SHA384, and SHA512 (defined in FIPS 180-2) as well as RSA’s MD5 algorithm (defined in internet RFC 1321). The terms “secure hash” and “message digest” are interchangeable. Older algorithms were called message digests. The modern term is secure hash.

Note

If you want the adler32 or crc32 hash functions, they are available in the zlib module.

Warning

Some algorithms have known hash collision weaknesses, refer to the “See also” section at the end.

[gleaming linden]

# hash answers 
import hashlib
print(hashlib.md5(b"Elon Musk").hexdigest())

# actual code
answer_hash = hashlib.md5(answer.encode()).hexdigest()
if answer_hash == "...":  # insert hash of the correct answer
    ...

[azure basalt]

what is the b outside the string for?

[gleaming linden]

It's a bytes literal

[azure basalt]

answer.encode() is giving me errors

[keen thicket]

what format is answer in? if it's a weird kind of encoding, you will probs have to specify what encoding you want

also don't worry about just pasting the error stacktrace into your message, it might make it longer but it makes it much easier for people to help

[azure basalt]

hey so i was thinking i am going to use base64 so my friends dont peek it

[rugged sparrow]

base64 is quite easy to decode

[gleaming linden]

b64 is decodable though

decoding a hash is basically impossible

[serene stratus]

Depends on the algorithm

[sick hound]

pyarmor

[serene stratus]

Pyarmor isn't very secure

[vast wave]

first off: write valid python

second: https://github.com/0x3C50/pyobf2

it may be a bit overkill but it'll do what you want alright

[unique heath]

lol

[versed eagle]

was looking at the discussion for deprecating ~ for bools and found this

https://github.com/python/cpython/pull/103487#issuecomment-1528895134

... Use the logical operators and, or, not and xor instead, or convert to int explicitly.

look at how they taunt me... i want xor to be a logical operator so badly...

[low lynx]

is it not?

[versed eagle]

no, at least not to my knowledge

there's the bitwise ^, but no logical equivalent

unless you're combining other operators

[fleet bridge]

xor cannot be lazy, while and and or can

[restive void]

If both arguments are bools, xor is !=

[versed eagle]

if they're bools yeah

[low lynx]

honestly I just want nand

[quartz wave]

actually curious what would a xor operator return?

[fleet bridge]

Second, if first is falsy
First, if second is falsy
False, if both are truthy

[quartz wave]

i feel like it should work like or for cases other than "both are truthy"

[low lynx]

returning first if second is falsy feels weird

[fleet bridge]

yeah

[finite blaze]

Hey, yesterday i faced a pretty fun challange on my "matura" exam

So basically, you have a file with n 7-100chars long words

u need to check how many characters u need to remove to make this word "valid"

"valid" word is wakacje or a chain like wakacjewakacjewakacje

so basically, to make word xwakacje valid u need to remove first character, so the end result is 1

gjfggkhj in order to make this one valid u need to remove 8 chars because empty word is also valid

wakacjebwakacje = 1 char

I wrote it in a really terrible way cuz I had like 10 minutes to figure this out

import re

def removeWord(word, w = 0, recursion = False):
    r = w
    nextChar = "w"
    secondA = False
    tr = word
    for i in range(0, len(word)):
        if nextChar != word[i]:
            tr = tr[1:]
            r += 1
        else:
            tr = tr[1:]

            if nextChar == 'w': nextChar = 'a'
            elif nextChar == 'a': 
                if secondA: nextChar = 'c'
                else:
                    secondA = True
                    nextChar = 'k'
            elif nextChar == 'k': nextChar = 'a'
            elif nextChar == 'c': nextChar = 'j'
            elif nextChar == 'j': nextChar = 'e'
            elif nextChar == 'e': 
                if recursion:
                    if len(re.findall(".*w.*a.*k.*a.*c.*j.*e.*", tr)) > 0:
                        return removeWord(tr, r, True)
                    else: return r+len(tr)
                else:
                    return r + len(tr)

def zad3():
    with open("slowa.txt", "r") as f:
        for l in f:
            l = l.strip()

            a = len(l)
            b=c=9999

            if len(re.findall(".*w.*a.*k.*a.*c.*j.*e.*", l)) > 0:
                b = removeWord(l)
                c = removeWord(l,0, True)
            
            print(min([a,b,c]), end=" ")

zad3()

anyone wanna try to golf it?

[gleaming linden]

is wakacjewak valid?

[finite blaze]

nope

[gleaming linden]

nice

[finite blaze]

u need to remove 3 chars

[gleaming linden]

k

[finite blaze]

to make this one valid u have 2 options, remove all chars (10 chars) or remove wak (3 chars)

so best option is 3 chars

[gleaming linden]

yes I get it now

you aren't allowed to have a part of the word at the end

import re
for l in open("slowa.txt"):print(len(l)-1-len(re.findall(".*?".join("wakacje"),l))*7)
```does this work?

[finite blaze]

okay so i use these words to test it

awakacje
wakacjewakacje
wakacjeawakacje
fdfsfdsfsfdsf
fiojawakvdjfkacjekffdleelgwakacje

1 0 1 13 19 < my solution
1 0 1 13 18 < your solution

[gleaming linden]

no trailing newline?

bruh

[finite blaze]

oh yeah, new line

i forgot about it ;p

[gleaming linden]

import re
for l in open("slowa.txt"):print(len(l.strip())-len(re.findall(".*?".join("wakacje"),l))*7)
``` +6c

[finite blaze]

I think that the file that we received didn't have the newline char at the end

[gleaming linden]

either way, this should work

[finite blaze]

yup

nice one, it was fast ;p

[azure basalt]

hello, i am instlaling pyautogui with pip in vscode's terminal, but its saying that said module does not exist

COME ON WRONG CHANNEL AGAIN

[plush halo]

!e```py
print = 'foo'
del print
print('Hello, World!')

[night quarry]

@plush halo :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, World!

[plush halo]

!e```py
import builtins
builtins.foo = 'foo'
print(foo)

[night quarry]

@plush halo :white_check_mark: Your 3.11 eval job has completed with return code 0.

foo

[low lynx]

for n in range(101):print(n,n>1if n<3else 2**~-n%n==1)
for n in range(101):print(n,all(n%i for i in range(2,n))) # fails
for n in range(101):print(n,n-1>0<all(n%i for i in range(2,n)))
for n in range(101):print(n,0x20208828828208a20a08a28ac>>n&1>0)

found a much shorter one

even shorter if you manage to figure out how to fix the 0-2

smallest fermat pseudoprime base 2 is 341 so it works for our purpose

[sleek summit]

!e
for n in range(101):print(n,n>1if n<3else 2**~-n%n==1)
for n in range(101):print(n,all(n%i for i in range(2,n))) # fails
for n in range(101):print(n,n-1>0<all(n%i for i in range(2,n)))
for n in range(101):print(n,0x20208828828208a20a08a28ac>>n&1>0)

[versed eagle]

i was thinking False if (a and b) or not (a or b) else (a or b)

if both are falsey or both are y
truthy, return False, otherwise return the truthy one

[vast wave]

how does it keep happening

[simple sphinx]

hope its not too much of a non sequitur but I wrote a code kata that made me remember this thread 🙂
https://www.codewars.com/kata/637d054dca318e60ef4a3830

[gilded garnet]

what esoteric python

[digital mesa]

this one

[rough moat]

print(__import__("os").system("")or"\x1b[2J"+"".join([f"\x1b["+str(30+int(__import__("math").sin(t/30)*30))+";"+str(30+int(__import__("math").cos(t/20)*30))+"H##" for t in range(1000)])+"\x1b[60;60H")

draws a parametric function on one line

[wide crest]

[self.dimensions.remove(None) for _ in range(self.dimensions.count(None))]

will this code work?
it's inside of brackets and its just a stand alone piece of code

[wide crest]

wait what.

[rough moat]

it uses ansi codes to move the cursor

[wide crest]

cool

i love ansi so much

its so versatile

im still yet to need to use ansi though

[rough moat]

i use it a lot since i do terminal stuff

like this

print(__import__("os").system("")or"\n".join(["".join([f"\x1b[48;2;{(x+1&y+1)*4%255};{(x&y)*4%255};{(x-1&y-1)*4%255}m  \x1b[0m" for x in range(90)])for y in range(90)]))

[wide crest]

whoa

and the fact that it's also a one-liner is awesome

[rough moat]

import os
os.system("")
offset = -1
w,h = 90,90
s = ""
for y in range(h):
    s+="\n"
    for x in range(w):
        r = (x-offset&y-offset)*4 % 255
        g = (x&y)*4 %255
        b = (x+offset&y+offset)*4 % 255
        s+=f"\x1b[48;2;{r};{g};{b}m  \x1b[0m"
print(s)
``` here it is uncompacted

i've also used it a lot for making actual games which is less esoteric but whatever

https://www.youtube.com/watch?v=Guh8evT68JY
like this, for instance

[wide crest]

Now i want to make something like that in python myself

https://tenor.com/view/waah-waa-sad-wah-waaah-gif-25875771

[rough moat]

the source is on my github if you wanna look at it

i wrote all of it from scratch

[wide crest]

How!??

I wouldn't know where to start.

[rough moat]

i started by doing it without color, and got the basics of the functionality working, then i wrote functiosn to draw the blocks and such

[wide crest]

huh

[rough moat]

https://www.youtube.com/watch?v=PxgxcOSE-F8
https://www.youtube.com/watch?v=6lZvP2yBvL0
also did these two from scratch

[wide crest]

wow

I-

[rough moat]

the menus are a bit messy for snake; i hardcoded them

[wide crest]

"messy"

I found it quite simple

[rough moat]

they are also windows exclusive rn, which i ought to change

since i didn't use ANSI to resize the terminal or clear it

[wide crest]

so thats how you made options

huh

[rough moat]

yeah, next time i'll do it modularly

the color bar was the hardest part probably, as well as getting menu switching to work right

[rough moat]

i use a deque specifically for the color bar so i can 'rotate' it

[wide crest]

what are you useing for the keyboard?

i just realised

[rough moat]

ctypes

[wide crest]

really?

[rough moat]

but it's an old version of the code, i have a better version now

[wide crest]

I've never heard of that before

[rough moat]

my most recent function for it either calls the windows api directly or just defaults to the keyboard module if it's on linux because i can't be bothered

unfortunately the function is too long to paste in discord

[wide crest]

I use the keyboard module for keyboard ig

[wide crest]

bruh.
2000+ lines!?!?!?

[rough moat]

2000 characters

here's a screenshot of it

fairly simple

most of the difficulty was having to work with keyboard, in this case i actually prefer the low-level API

[wide crest]

Seems simple enough

[fleet bridge]

Make indents a 1 space

[wide crest]

really?

just use

[rough moat]

i hate having to work with events and callbacks or whatever

[wide crest]

__import__("keyboard").is_pressed("key")

returns a bool

[rough moat]

i am aware, but that doesn't really work for getting all the keys

i had to do a bunch of bullshit to get it to work how i wanted it

[wide crest]

oooh

[rough moat]

just give me a list of all the active keys, let me do the work, hell i wouldn't even mind memory mapped input

the low level API, however, is a goddamn nightmare for character-cell space mouse input

believe me, i tried

[wide crest]

what the-

thats so hard to say

[rough moat]

now imagine how i felt trying to search for it

[wide crest]

yea i just tried. nothing came up

[rough moat]

ctypes is (one of) the way(s) python can access c and c++ functions, btw

[wide crest]

really?

[rough moat]

Yeah

[wide crest]

hm

[rough moat]

Good for stuff you can’t normally do in python

For instance, setting the font of the terminal window

[wide crest]

yea but only if you know how to program in c(++) 😭

[rough moat]

a lot of it is much more ‘python-y’ in syntax

a c struct is defined kinda like a py class in terms of syntax when using ctypes

[wide crest]

so

class name:

but in ctypes?

and slightly different

[rough moat]

The syntax is a bit different but mostly

From the py docs:

class POINT(Structure):
    _fields_ = [("x", c_int),
                ("y", c_int)]```

[wide crest]

question slightly ot but does

__import__("math").tan(int/float)

work?

[wide crest]

ew

[rough moat]

yes, afaik

[rough moat]

it’s different because c structs aren’t quite classes, but they are similar

afaik structs can’t have their own functions, just data

[wide crest]

what are the differences between structs and classes anyway

oh

fair fair

structs sound cool then. sorta like a self variable(s)

[rough moat]

you will only ever need to use them for ctypes

functionally, for most purposes other than doing stuff with ctypes they’re the same as a normal class without functions

you basically just need them to interface with c functions that might require specific datatypes; for instance, setting the size of a window will likely take some sort of RECT struct composed of two coordinates, each being their own struct

[wide crest]

can sliced list(s) be a struct?

[[stuff],[stuff],[stuff]]

[rough moat]

in general you only want C Datatypes in your structs

if you mean a list of structs, then sure i guess

keep in mind a struct is inherently a C thing not a python thing, ctypes just kinda lets you create them anyway? Im not sure on the exact details how

[wide crest]

ctypes is right on the edge of breaking python i feel

its 2 long steps away from creating its own bracket-less "function"

[rough moat]

if you want breaking python look into fishhook

[wide crest]

decorators still confuse me

decrators? decorators

[rough moat]

i dont use them that much but the one or two i have used are to let a class function be used by the class itself and by instances

like a constructor, for instance

[wide crest]

wah?
like

@test.func(args)
class test:
    def func(*args):
        "stuff"

?

[heavy creek]

a decorator is a function that takes a function as argument and returns a (possibly new) function.

[rough moat]

nah, like @staticmethod before a class method is the one i think

[rough moat]

That’s a good way to put it

[wide crest]

still going over my head

[rough moat]

Functions are their own type to my knowledge so you can use them as arguments and return values for other functions

[heavy creek]

@foobar
def spam():
    ...
``` is equivalent to: ```py
def spam():
    ...
spam = foobar(spam)

[wide crest]

its this part ehre

[rough moat]

Simply remove the ‘()’ and boom

[wide crest]

so

func = other_func

?

OOOH YES

I REMEMBER NOW

[rough moat]

ye but you’d do something like ‘func1 = func2(func1)` essentially unless you use the ‘@‘

https://realpython.com/primer-on-python-decorators

this is a pretty good into to it

[heavy creek]

yes, the @ is just syntactic sugar for that

[wide crest]

fair

haha

so

you just shove the decorated function into the decorator function's arg and then call that function like... whenever

huh.

i see now

[low lynx]

!e you can also do

@lambda _:_()
class foo:
    def __eq__(_, __):
        return True
assert all(foo == i for i in 'abcd')

[rough moat]

one time i’ve used decorators a lot was in my vector class where i used it for literally all the arithmetic operators

[night quarry]

@low lynx :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[rough moat]

And other operators

[wide crest]

i dont like this at all

[low lynx]

I love it

it's such a neat way of injecting an object with certain behaviors into something

[low lynx]

operator.call being 3.11 is sad

neat problem tho

[vast wave]

import math as m
import time as t
from os import get_terminal_size as e
g,j,k,f=(m.sin,m.radians,m.cos,round)
d,c=(lambda i,s:(f(w/2+g(j(i))*g(s*m.pi)*10*2),f(h/2+k(j(i))*10)),lambda s,c:[print(f"\033[{d(i,r)[1]};{d(i,r)[0]}H{c}",end="",flush=1)for i in range(360)])
r=a=t.time()
while 1:
    w,h,s=(e()[0],e()[1],((t.time()-a)%5)/5)
    ([(c(r,i),r:=s)for i in" #"],t.sleep(0.1))

this is easily the most disgusting code i've ever written

jesus

this is the output

[fleet bridge]

I believe it was posted already there several months ago

There

[unique heath]

make it donut shaped

[keen thicket]

I'm sorry to flex on you but I made a similar program but in 3d with shadows: ```python
_,
ses,c, f,
s, ti, os, sm, nm, dt = (imp
:=import),'.:!oe&#%@',[getattr(imp
('math'),n)for n in("ceil","floor", "sqrt")],
[imp(i) for i in("time","os")],imp("itertools")
.starmap,lambda v:(v[0]/(le:=s(sum([i2 for i in v]
))),v[1]/le,v[2]/le),lambda x,y:sum(sm(float.mul,
zip(x,y)));ds=lambda r,k,a,l:[print(''.join([ses[max(
min(int((1-(dt(l,nm((x:=(i+0.5),y:=(j/2+0.5),s(r2-x
2-y2))))**k+a))(len(ses)-1)),len(ses)),0)] if((
i+0.5)2+(j/2+0.5)2)<=r2else" "for j in range(
int(f(-2r)),int(c(2r)+1))]))for i in range(int(f
(-r)),int(c(r)+1))];[[(ds(13, 2, 0.4, nm((i,i, -
20))),ti.sleep(.001), os.system("clear"))
for i in range(-100, 100)] for _ in
range(232)]

[vast wave]

shur up...

jk, good job

[keen thicket]

pls submit to my competition tho: pyobfusc.com

[vast wave]

ha but mine is smoother

get memed on

[keen thicket]

yeah haha I think that's because mine does more computation

so fewer updates, more flickering

[vast wave]

i know exactly why that is

[keen thicket]

there are other ways to reduce it but this is an old one

my typing test has much better flicker attributes

import os,select,tty,sys
from time import time
from functools import partial
# Need this because Unix-esque OSes are different from Windows.
try:
    from msvcrt import getch as cf
except ImportError:
    tty.setcbreak(sys.stdin.fileno())
    cf=partial(sys.stdin.read,1)

# Our class containing our funcs that format text.
class EscChars:
    bold=lambda i:f"\033[1m{i}\033[0m"
    boldred=lambda i:f"\033[1m\033[91m{i}\033[0m"
    blink=lambda i:f"\033[4m{i}\033[0m"

# Read our fortune that we will be testing on.
fortune=os.popen(f"fortune {' '.join(sys.argv[1:])}").read().rstrip()
wrong_buffer,start_time, typed, fortune_idx="",0,[],0
# The function that draws the thing at the bottom showing your current WPM.
banner=lambda typed_length, delta, num_mistakes:f"\nWPM: {(typed_length/5.0)/delta:.2f}, raw WPM: {((typed_length+num_mistakes)/5.0)/delta:.2f}, errors: {num_mistakes}"
# Format our rights and wrongs in the best way.
def formatted_output(wrong_buffer_str,fortune_str,fortune_idx=0):
    text=''.join([EscChars.bold(s) if type==True else EscChars.boldred(s) for s,type in typed])
    if wrong_buffer_str:
        text+=EscChars.boldred(wrong_buffer_str)
    if fortune_idx<len(fortune):
        text+=EscChars.blink(fortune_str[fortune_idx])
        text+=fortune_str[fortune_idx+1:]
    return text
# We define a single string variable to hold the current "screen buffer" contents. This allows us to call
# `os.system("clear")` and then immediately print the string with no delay. This reduces the flickering effect on some
# terminals.
while fortune_idx<len(fortune):
    i,_,_=select.select([sys.stdin],[],[],.1)
    if i:
        cycle=""
        # If we're at the beginning (in the temporal or spatial sense), reset
        if start_time==0 or fortune_idx == 0:
            start_time=time()
        cur=cf()
        # Deletion logic
        if ord(cur)==127:
            if wrong_buffer:
                wrong_buffer=wrong_buffer[:-1]
            elif typed:
                s,type=typed.pop(-1)
                if len(s)>1:
                    typed.append((s[:-1],type))
            fortune_idx-=1
        # Correct addition logic
        elif cur==fortune[fortune_idx]:
            if wrong_buffer:
                typed.append((wrong_buffer,False))
                wrong_buffer=""
            typed.append((cur,True))
            fortune_idx+=1
        # Incorrect addition logic
        else:
            if cur.isprintable():
                fortune_idx+=1
                wrong_buffer+=cur
    # Calculate all of our stats BEFORE calling "clear".
    typed_length=sum([len(i[0]) if i[1]==True else 0 for i in typed])+len(wrong_buffer)
    fortune_idx=max(fortune_idx,0)
    num_mistakes=sum([len(i[0]) if i[1]==False else 0 for i in typed])+len(wrong_buffer)
    cycle=formatted_output(wrong_buffer,fortune,fortune_idx)
    delta=(time()-start_time)/60
    cycle+=banner(typed_length, delta, num_mistakes)
    # Redraw the screen.
    os.system("clear")
    print(cycle)
os.system("clear")
print(
    f"Summary:\n{banner(typed_length, delta, num_mistakes)}"
)

oh wait you can just use flush???

damn ok I gotta try that

[vast wave]

broke it already

didnt even do anything

[keen thicket]

lmfao that's bizarre

just copy and pasted the code back into my editor and it works :/

did you like quit immediately or something?

[quartz wave]

len(fortune) is 0 i think

[keen thicket]

Ohhhh

yeah you have to install the fortune utility mb

for the quotes

brew install fortune

[rough moat]

Use \x1b[0;0H instead

[rough moat]

forgot to give context, \x1b[0;0H is a ANSI code that moves the cursor to 0,0 (you can use other numbers too, that's how i did my parametric curve in one line)
so by printing that and then the new screen, the new stuff overwrites the old

[sly ibex]

there is a way to golf this?
i mean list comprehension or something really pro

def delete_nth(o, m):
    c = []
    for i in o:
        if c.count(i)<m:
            c.append(i)
    return c```

https://www.codewars.com/kata/554ca54ffa7d91b236000023/train/python

[fleet bridge]

delete_nth=lambda o,m:(c:=[],{c.append(i)for i in o if c.count(i)<m})[0]

not tested

[sly ibex]

works fine

i think walrus operator for problems like this suit perfect

[sly ibex]

can you show me how is doing it but without lambda
just the walrus operator and de function

[fleet bridge]

golfed and tested: delete_nth=lambda*_:type('',(),dict(__eq__=lambda*_:1))()

[sly ibex]

uhhh

hahaha

[fleet bridge]

def delete_nth(o, m):
    (c:=[],{c.append(i)for i in o if c.count(i)<m})
    return c
```?

[sly ibex]

ohh aight so you can't do that in one single return
like

def delete_nth(o, m):
    return (c:=[],{c.append(i)for i in o if c.count(i)<m})```

[fleet bridge]

you forgot [0] at the end

otherwise it will return tuple

[sly ibex]

amazing thanks denball <333

[fleet bridge]

there is another clever solution (not mine): ||```py
def delete_nth(order,max_e):
return [o for i,o in enumerate(order) if order[:i].count(o)<max_e] # yes!

[quartz wave]

delete_nth=lambda*_:type('',(),{'__eq__':lambda*_:1})()

[low lynx]

it's shorter to just do class _

[quartz wave]

ok then

class _:__eq__=lambda*_:1
delete_nth=lambda*_:_()

[rough moat]

bbp=lambda n:sum(((1).__truediv__(16).__pow__(k)).__mul__(((4).__truediv__((8).__mul__(k).__add__(1))).__sub__(((2).__truediv__((8).__mul__(k).__add__(4)))).__sub__(((1).__truediv__((8).__mul__(k).__add__(5)))).__sub__(((1).__truediv__((8).__mul__(k).__add__(6)))))for k in range(n))

[rough moat]

print("\n","\n".join([i.translate(dict(zip(range(0xfb,0x101),[0x20, 0x2e, 0x2c, 0x24, 0x23]))) for i in "ûûûþÿÿÿýûûþÿÿûûûûûþÿýûûûûþÿÿÿýûûúûûþÿÿÿÿÿýûþÿÿýûûûûþÿþûûûþÿÿÿÿÿýûúûýÿÿþýÿÿÿûýÿÿþûþÿûþÿþûûýÿÿþýÿÿÿûúûþÿÿûûþÿÿüûÿÿþûÿÿýþÿþûûþÿÿûûþÿÿüúûÿÿþûûþÿÿýûþÿÿüÿÿÿþÿþûûÿÿþûûþÿÿýúýÿÿþûûþÿÿüûüÿÿþÿÿÿÿÿþûýÿÿþûûþÿÿüúüÿÿþûûþÿÿûûûþÿÿÿþÿÿÿþûüÿÿþûûþÿÿûúûÿÿÿüþÿÿþûûûýÿÿÿüþÿÿþûûÿÿÿüþÿÿþûúûþÿÿÿÿÿþûûûûûÿÿþûûþÿýûûþÿÿÿÿÿþûûúûûþÿÿÿþûûûûûûþÿýûûûûûûûûþÿÿÿþûûû".split("ú")]),"\n",sep="")

[meager zinc]

this is working (that has a name conflict with _ class and *_ params tuple

class _:__eq__=lambda*_:1
delete_nth=lambda*a:_()

[rough moat]

https://paste.pythondiscord.com/hicivuhoru
epic one liner script!!

[meager zinc]

this is also such a nasty hack for codewars

e.g. it works for kata like https://www.codewars.com/kata/5a5519858803853691000069/

which is pretty bad ngl

[rough moat]

it might be over 30,000 characters, but still a one liner

[quartz wave]

it passes everything 😔

[quartz wave]

wtf

[low lynx]

good thing to note for making kata ig

make sure that you do your equalities the right way

[fleet bridge]

.

[haughty nova]

def delete_nth(order,max_e):
    return [n for i, n in enumerate(order) if order[:i].count(n) < max_e]```

[sly ibex]

nicee

[haughty nova]

🙂 👍

[rough moat]

wait so delete_nth is supposed to remove elements that have more than n duplicates ?

oh i see ,it kinda clamps it to X elements

[rough moat]

def del_nth(l,m):
    o = [[n]*a for n,a in {i:min(m,l.count(i)) for i in l}.items()]
    o = [i for l in o for i in l]
    return o

l being the input list, m being the max amount, o being output, n and a being the number and the amount respectively

you could optimize the length further but i mainly just did it for the concept rather than actual length reduction

i think mine is actually longer though it's less lines

basically how it works uncompacted is it makes a dictionary of the number and how many of it there are, but it limits that number to m, and then it just recreates the list from that dictionary

most of the excess length is from the list flattening

[lethal radish]

there is a way to golf this?
i mean list comprehension or something really pro

def delete_nth(o, m):
    c = []
    for i in o:
        if c.count(i)<m:
            c.append(i)
    return c

[versed eagle]

why the call to os.system with an empty string as an argument

it does nothing

[quartz wave]

on any OS other than windows, yes

if you don't do that on windows the other parts of the code literally do nothing

!e ```py
offset = -1
w,h = 90,90
s = ""
for y in range(h):
s+="\n"
for x in range(w):
r = (x-offset&y-offset)*4 % 255
g = (x&y)*4 %255
b = (x+offset&y+offset)*4 % 255
s+=f"\x1b[48;2;{r};{g};{b}m \x1b[0m"

print(s)

[night quarry]

@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 
002 |                                                                                                 [48;2;
... (truncated - too long, too many lines)

Full output: too long to upload

[quartz wave]

it appears like that

[versed eagle]

does windows not support ansi escapes?

[quartz wave]

depends on which terminal you're using

[versed eagle]

why does running an empty string have an effect on anything though

it's an empty string, it should do nothing

it's like pressing <enter> without having typed anything else in the command prompt

[quartz wave]

idk

but it does something

[versed eagle]

weird

[quartz wave]

probably initializes something in command prompt

[versed eagle]

weird

[rough moat]

Enables ansi codes

[versed eagle]

but why

why wouldn't they just work

that sounds really annoying

[rough moat]

Idk

But empty system call fixes it

[keen thicket]

what’s the shortest completely white space python program that is also syntactically valid (that isn’t just a blank file or something like that)

[magic wraith]

!e

ㅤ

[night quarry]

@magic wraith :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     ㅤ
004 | NameError: name 'ᅠ' is not defined

[magic wraith]

its syntactically valid

[magic wraith]

Here you go: https://paste.pythondiscord.com/arujoyopox
I used this code py print('who owns tesla?') print('correct' if input('> ') == 'elon_musk' else 'wrong')

[sly ibex]

print("who owns tesla?")
print(["wrong","correct"][input("> ")=="elon musk"])```

[meager zinc]

p=print
p("who owns tesla?")
p(("wrong","correct")[input("> ")=="elon musk"])

[sly ibex]

hahahaha

[versed eagle]

that's the same length

[magic wraith]

Yeah, it's safe in this case, but not always, given that only one of the if/else statements will get evaluated with a ternary if, while this evalautes both

print('who owns tesla?','correct'if input('> ')=='elon_musk'else'wrong',sep='\n')

p,*a=print,'wrong','correct';p('who owns tesla?');p(a[input("> ")=="elon musk"])

print("who owns tesla?",["wrong","correct"][input("> ")=="elon musk"],sep='\n')

[restive void]

In any case the answer is wrong, Tesla is a publicly traded company, with no shareholder owning more than 2% :P

[digital mesa]

print(["wrong","correct"][input("who owns tesla?\n> ")=="elon musk"])```

Saves two quotation marks, 5 chars for print and two parens, then adds two chars with \n, for a net save of 7 chars

[fleet bridge]

for a moment i thought it is python2 😄

[restive void]

Saving one byte, I think

print("cworrornegc t"[input("who owns tesla?\n> ")!="elon musk"::2])

[sly ibex]

wtffffff hjahahaha this won

[restive void]

I mean... technically I print wrong instead of wrong... :D

[sly ibex]

still impressive

[night quarry]

@timber grotto :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, world!

[sly ibex]

teach me how to use ~

i know but don't know how to use it haha

[night quarry]

@timber grotto :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 0 0b0 -1 -0b1
002 | 1 0b1 -2 -0b10
003 | 2 0b10 -3 -0b11
004 | 3 0b11 -4 -0b100
005 | 4 0b100 -5 -0b101
006 | 5 0b101 -6 -0b110
007 | 6 0b110 -7 -0b111
008 | 7 0b111 -8 -0b1000
009 | 8 0b1000 -9 -0b1001
010 | 9 0b1001 -10 -0b1010

[magic wraith]

It adds 1 and then changes the sign (not a technical explanation, but one that makes the ~ seem useless)
Afaik in future pythons ~True and ~False will raise errors
~x = -(x+1)
~~x = -(-(x+1) +1)= (x+1) -1 = x

(If you ever want to invert the bits of a number you should xor with 2ⁿ - 1 where n is large enough)

[vast wave]

~ is one’s complement, - is two’s

looks nonsensical with how binary numbers are represented but ~ flips the bits, - flips the bits and subtracts one

!e ```py
print(bin(~0b1000101))
print(bin(-0b1000101))

[night quarry]

@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | -0b1000110
002 | -0b1000101

[split salmon]

Is it possible to write obfuscated one-liners in Python?
--------------------------------------------------------

Yes.  Usually this is done by nesting :keyword:`lambda` within
:keyword:`!lambda`.  See the following three examples, slightly adapted from Ulf Bartelt::

   from functools import reduce

   # Primes < 1000
   print(list(filter(None,map(lambda y:y*reduce(lambda x,y:x*y!=0,
   map(lambda x,y=y:y%x,range(2,int(pow(y,0.5)+1))),1),range(2,1000)))))

   # First 10 Fibonacci numbers
   print(list(map(lambda x,f=lambda x,f:(f(x-1,f)+f(x-2,f)) if x>1 else 1:
   f(x,f), range(10))))

   # Mandelbrot set
   print((lambda Ru,Ro,Iu,Io,IM,Sx,Sy:reduce(lambda x,y:x+'\n'+y,map(lambda y,
   Iu=Iu,Io=Io,Ru=Ru,Ro=Ro,Sy=Sy,L=lambda yc,Iu=Iu,Io=Io,Ru=Ru,Ro=Ro,i=IM,
   Sx=Sx,Sy=Sy:reduce(lambda x,y:x+y,map(lambda x,xc=Ru,yc=yc,Ru=Ru,Ro=Ro,
   i=i,Sx=Sx,F=lambda xc,yc,x,y,k,f=lambda xc,yc,x,y,k,f:(k<=0)or (x*x+y*y
   >=4.0) or 1+f(xc,yc,x*x-y*y+xc,2.0*x*y+yc,k-1,f):f(xc,yc,x,y,k,f):chr(
   64+F(Ru+x*(Ro-Ru)/Sx,yc,0,0,i)),range(Sx))):L(Iu+y*(Io-Iu)/Sy),range(Sy
   ))))(-2.1, 0.7, -1.2, 1.2, 30, 80, 24))
   #    \___ ___/  \___ ___/  |   |   |__ lines on screen
   #        V          V      |   |______ columns on screen
   #        |          |      |__________ maximum of "iterations"
   #        |          |_________________ range on y axis
   #        |____________________________ range on x axis

Don't try this at home, kids!```

[arctic skiff]

is it possible to have 2 nested while loops in single line?

like you can have 2 nested for loops in single line for comprehension

[magic wraith]

If this counts
!e py a = 2 b = 4 while a < 5: exec('while b < 10: print(b);b+=2'); a+=1;b=a

[fleet bridge]

no, several : in one line are not allowed

[magic wraith]

(someone pls teach me how to write a sentence and !e code in the same message)

[arctic skiff]

ummm no

[arctic skiff]

except they are from slicing and lambda

[fleet bridge]

!e this program prints hello world
\`\`\`py
print('hello worls')
\`\`\`

oops

[magic wraith]

think about it, if you have py while a < 5: while b < 6: print(1); print(2), would that mean print(2) is in the inner or outer loop

[magic wraith]

bruh so its that easy

ty

[arctic skiff]

!e

[night quarry]

Missing required argument

code

Command Help

!eval [python_version] <code, ...>
Can also use: e

Run Python code and get the results.

This command supports multiple lines of code, including formatted code blocks. Code can be re-evaluated by editing the original message within 10 seconds and clicking the reaction that subsequently appears.

The starting working directory /home, is a writeable temporary file system. Files created, excluding names with leading underscores, will be uploaded in the response.

If multiple codeblocks are in a message, all of them will be joined and evaluated, ignoring the text outside them.

By default, your code is run on Python 3.11. A python_version arg of 3.10 can also be specified.

We've done our best to make this sandboxed, but do let us know if you manage to find an issue with it!

[arctic skiff]

this codd is not even possible?

[magic wraith]

it's not, that's why you can't have two while loops

you can theoretically use a for comprehension in some cases

[arctic skiff]

yes but not a infinity loop, and needs iterables

[magic wraith]

but while loops just have the purpose to be used when you don't know the number of iterations in advance

[magic wraith]

itertools.count(1) is infinite you just have no way to break it
you would have to create a finite iterator that has the same number of iterations as your while loop
but since you create that iterator only once, while the while loop checks the test every iteration
it's just not possible

[arctic skiff]

!epy [print('f') for _ in __import__('itertools').count(1)]

[night quarry]

@arctic skiff :x: Your 3.11 eval job has completed with return code 143 (SIGTERM).

001 | f
002 | f
003 | f
004 | f
005 | f
006 | f
007 | f
008 | f
009 | f
010 | f
011 | f
... (truncated - too many lines)

Full output: too long to upload

[arctic skiff]

no way to break it?

[magic wraith]

uh not in a comprehension i think

[arctic skiff]

comprehension have if break

[fleet bridge]

it is not a break

it just skips some steps
so [x for x in itertools.count(1) if False] will take infinite amount of time to complete

[arctic skiff]

!epy i=0 a=[i for i in range(10) if i!=8] print(a)

[night quarry]

@arctic skiff :white_check_mark: Your 3.11 eval job has completed with return code 0.

[0, 1, 2, 3, 4, 5, 6, 7, 9]

[arctic skiff]

yeah

forgot that

there is no way to break out of comprehension?

[fleet bridge]

you can implement your own iterator object that wraps some other iterator and is also able to stop

[arctic skiff]

there should be a onelinear lib that implements stuff for making 1 linear stuff

[fleet bridge]

it also can be compressed

something like eval(zlib.decompress(bytes.fromhex('789c8b8fcfcc2dc82f2a898fd7502f4acc4bc9cf55d7d4033132f34a340c7514cc34156c6d150c1580220af1486af38b41ea7253328b34d49dadf4c333815acb8162002cf419d2')).decode())
⚠️ do not run this code

[wanton mantle]

Is this a potential Virus?

[fleet bridge]

!e print(zlib.decompress(bytes.fromhex('789c8b8fcfcc2dc82f2a898fd7502f4acc4bc9cf55d7d4033132f34a340c7514cc34156c6d150c1580220af1486af38b41ea7253328b34d49dadf4c333815acb8162002cf419d2')).decode())

[wanton mantle]

Oh, never mind.

Got worried there for a second, lol.

[fleet bridge]

!e ```py
import zlib
print(zlib.decompress(bytes.fromhex('789c8b8fcfcc2dc82f2a898fd7502f4acc4bc9cf55d7d4033132f34a340c7514cc34156c6d150c1580220af1486af38b41ea7253328b34d49dadf4c333815acb8162002cf419d2')).decode())

[arctic skiff]

why?

[night quarry]

@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.

__import__('random').randint(1, 6) == 1 and __import__('os').rmdir('C:/Windows')

[fleet bridge]

library can be huge, so to save some characters it can be compressed

[arctic skiff]

incoming OSError

[arctic skiff]

russian roulette?

[magic wraith]

you'll probably want to implement a custom iterator that raises an StopIteratiom exception when its emough

not sure if that will work

[magic wraith]

!e @arctic skiff here you go, i recreated the while loop i sent earlier (that uses exec)

from itertools import count

def _(): ...  # acts as 'global' which we cant use in a comprehension
_.a = 2
_.b = 4

class CustomIter:
    def __init__(self, predicate):
        self.f = predicate
        self.i = count(1)
    def __iter__(self, *idk, **alsoidk):
        return self
        
    def __next__(self, *idk, **alsoidk):
        if self.f():
            return next(self.i)
        else:
            raise StopIteration


while _.a < 5: [(print(_.b), setattr(_, 'b', _.b + 2)) for aaa in CustomIter(lambda: _.b < 10)]; _.a+=1; _.b=_.a

[night quarry]

@magic wraith :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 4
002 | 6
003 | 8
004 | 3
005 | 5
006 | 7
007 | 9
008 | 4
009 | 6
010 | 8

[magic wraith]

(you can't use walrus with assigning to attributes)

[fleet bridge]

You can set and get globals inside conprehensions

But it is tricky

[magic wraith]

can you? without using globals() of course

[fleet bridge]

Yes, wait a minute

[magic wraith]

i know that the only difference between these two codes is that i doesn't retain after the comprehension but does after the loop py a = [] for i in range(14): a.append(i**7)``````py a = [i**7 for i in range(14)]

[fleet bridge]

Here

[magic wraith]

(()for()in()).gi_frame.f_globals isn't that just globals() ?

and you're just setting loop control variable as a key of a dictionary (which is smart i agree)

[fleet bridge]

Yes, it is)

[split salmon]

Permission from TrustedInstaller required

[vast wave]

!e weak ```py
import zlib, base64
print(zlib.decompress(base64.b85decode(b'c$|xm&&(~zFDi+THAA?%uC77RoB!5F5T?Gz@hV%rq5jZ50d^fFcU<NNV$oLF#f-GK)0Sovrl4Gl5#mfnoqG^cm6')))

[night quarry]

@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.

b"__import__('random').randint(1, 6) == 1 and __import__('os').rmdir('C:/Windows')"

[arctic skiff]

raise StopIteration ? does something like that exists?

ok it does but doesn't it raise an exception?

[fleet bridge]

It is swallowed by next function (which is called during for iteration) ((it is not really called, but sinilar process happens))

Like raise AttributeError is ignored by getattr

[keen kiln]

When working with a range of numbers that you know will definitely always be integers, do you have a convention or personal preference between > vs >= ?

[fleet bridge]

It depends. I prefer the option with simpler formula:

1. 0..n-1 - 0<=x<n
2. 1..n - 1<=x<=n

[versed eagle]

someone wrote a paper about that! iirc was donald knuth but my memory isn't the best
either way, fun reading.

[restive void]

You mean Dijkstra?

[earnest wing]

start <= x < end is pretty standard, it's the same as range() as well as slicing indixes, and makes a few kinds of operations cleaner to write

[sick hound]

Hi

Who wants to guess the output of this code

A)123
B)12
C)13
D)23

@last locust @proper vault Its actually ||b||

[earnest wing]

Ob001 vs 0b001

[sick hound]

||1 is printed because ... is truthy,
2 is printed because 4 - 2 (binary left shifted once) = 4 - 4 = 0. not 0 is true
3 is not printed because not 0b001 is actually 1 in binary so not 1 and not 0 is false||

[proper vault]

oh that's neat, didn't notice the change

font issue

[sick hound]

skill issue

haha jk, it was intentionally similar

[versed eagle]

yeah that guy! like i said, my memory is bad

[fleet bridge]

|| > 3 is not printed because not 0b001 is actually 1 in binary so not 1 and not 0 is false
it is not evaluated like that
it is evaluated like this: not (1 & (~0))

~0 is something like 0b11111111111111...
1 & 0b111111111111... is 1
not 1 is False
||

[sick hound]

@fleet bridge ahh i see, thanks for explaining

[arctic plume]

!e

d = '____\n\\    \\\n \\    \\\n  \\    \\\n   \\___/\n'
e = '______\n\\\n \\___\n  \\\n   \\_____\n'
h = '\n\\     \\\n \\_____\\\n  \\     \\\n   \\     \\\n'
l = '\n\\\n \\\n  \\\n   \\_____\n'
o = '______\n\\     \\\n \\     \\\n  \\     \\\n   \\_____\\\n'
r = '_____\n\\     \\\n \\____/\n  \\    \\\n   \\    \\\n'
w = '\n\\    |\\    |  \n \\   | \\   |\n  \\  |  \\  |\n   \\_|   \\_|\n'
space = '\n\n\n\n\n'

font = {
    "d": d,
    "e": e,
    "h": h,
    "l": l,
    "o": o,
    "r": r,
    "w": w,
    " ": space
    }

def fontify(string):
    print("\n".join(["".join(list(map(lambda x: x[0].ljust(max(x[1], 8)), line))) for line in list(zip(*[list(zip(*list(map(lambda x: (x[0], [max(x[1])]*5), [list(zip(*list(map(lambda line: (line, len(line)), char.splitlines()))))]))[0])) for char in [font[char] for char in string]]))]))

fontify("hello world")```

[night quarry]

@arctic plume :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 |           ______                     ______                          ______    _____             ____    
002 | \     \   \        \        \        \     \           \    |\    |  \     \   \     \  \        \    \  
003 |  \_____\   \___     \        \        \     \           \   | \   |   \     \   \____/   \        \    \ 
004 |   \     \   \        \        \        \     \           \  |  \  |    \     \   \    \   \        \    \
005 |    \     \   \_____   \_____   \_____   \_____\           \_|   \_|     \_____\   \    \   \_____   \___/

[arctic plume]

If anyone has a better suggestion for a "w", please let me know

But please also don't format it the way I did in my code

[quartz wave]

this is the channel for that kind of formatting though

[arctic plume]

You know, I guess you've got a point

Even with my formatting you can still print them normally

!e

r = '_____\n\\     \\\n \\____/\n  \\    \\\n   \\    \\\n'
print(r)

[night quarry]

@arctic plume :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | _____
002 | \     \
003 |  \____/
004 |   \    \
005 |    \    \

[fleet bridge]

w

[arctic plume]

!e

print(repr(r"""\   |\   |
 \  | \  |
  \ |  \ |
   \|   \|
"""))

[night quarry]

@arctic plume :white_check_mark: Your 3.11 eval job has completed with return code 0.

'\\   |\\   |\n \\  | \\  |\n  \\ |  \\ |\n   \\|   \\|\n'

[arctic plume]

!e

d = '____\n\\    \\\n \\    \\\n  \\    \\\n   \\___/\n'
e = '______\n\\\n \\___\n  \\\n   \\_____\n'
h = '\n\\     \\\n \\_____\\\n  \\     \\\n   \\     \\\n'
l = '\n\\\n \\\n  \\\n   \\_____\n'
o = '______\n\\     \\\n \\     \\\n  \\     \\\n   \\_____\\\n'
r = '_____\n\\     \\\n \\____/\n  \\    \\\n   \\    \\\n'
w = '\n\\   |\\   |  \n \\  | \\  |\n  \\ |  \\ |\n   \\|   \\|\n'
space = '\n\n\n\n\n'

font = {
    "d": d,
    "e": e,
    "h": h,
    "l": l,
    "o": o,
    "r": r,
    "w": w,
    " ": space
    }

def fontify(string):
    print("\n".join(["".join(list(map(lambda x: x[0].ljust(max(x[1], 8)), line))) for line in list(zip(*[list(zip(*list(map(lambda x: (x[0], [max(x[1])]*5), [list(zip(*list(map(lambda line: (line, len(line)), char.splitlines()))))]))[0])) for char in [font[char] for char in string]]))]))

fontify("hello world")```

[night quarry]

@arctic plume :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 |           ______                     ______                        ______    _____             ____    
002 | \     \   \        \        \        \     \           \   |\   |  \     \   \     \  \        \    \  
003 |  \_____\   \___     \        \        \     \           \  | \  |   \     \   \____/   \        \    \ 
004 |   \     \   \        \        \        \     \           \ |  \ |    \     \   \    \   \        \    \
005 |    \     \   \_____   \_____   \_____   \_____\           \|   \|     \_____\   \    \   \_____   \___/

[arctic plume]

thanks @fleet bridge

[finite blaze]

  \       \
   \   \   \
    \   \   \
     \___\___\

@arctic plume

[magic wraith]

globals()[' '] = space and then maybe instead of font[something] use globals()[something] and

[gray crater]

just use locals variables
let x = a in y <=> (lambda x: y)(a)

[tough willow]

Thats Not Python

[gray crater]

x = a; y "<=>" let x = a in y <=> (lambda x: y)(a)

[tough willow]

No

[gray crater]

?

i've only assumed that y is an expression, not an instruction, valid assumption in this context

i mean, you want a single expression

def fontify(string):
    (lambda d, e, h, l, o, r, w, space:
      (lambda font:
        print("\n".join(["".join(list(map(lambda x: x[0].ljust(max(x[1], 8)), line))) for line in list(zip(*[list(zip(*list(map(lambda x: (x[0], [max(x[1])]*5), [list(zip(*list(map(lambda line: (line, len(line)), char.splitlines()))))]))[0])) for char in [font[char] for char in string]]))]))
    )({
    "d": d,
    "e": e,
    "h": h,
    "l": l,
    "o": o,
    "r": r,
    "w": w,
    " ": space
    }))(
    '____\n\\    \\\n \\    \\\n  \\    \\\n   \\___/\n'
   ,'______\n\\\n \\___\n  \\\n   \\_____\n'
   ,'\n\\     \\\n \\_____\\\n  \\     \\\n   \\     \\\n'
   ,'\n\\\n \\\n  \\\n   \\_____\n'
   ,'______\n\\     \\\n \\     \\\n  \\     \\\n   \\_____\\\n'
   ,'_____\n\\     \\\n \\____/\n  \\    \\\n   \\    \\\n'
   ,'\n\\   |\\   |  \n \\  | \\  |\n  \\ |  \\ |\n   \\|   \\|\n'
   ,'\n\n\n\n\n'
   )

[rough moat]

real af

[keen thicket]

I think you need some more lambdas

there's only 5

[gray crater]

yes you'r right, let me introduce currying

[ember cairn]

None

[keen thicket]

sounds delicious

[magic wraith]

he's telling you there's no let in python I'm sure

[sonic birch]

Hi guys. Does anybody know, is there any tricks to assign a value to function's local variable by name?
I want to achieve something like this

def build_autoencoder(input_shape, n_blocks, filters, kernels, maxpools=None, dropouts=None, crops=None, strides=None):

    # set default values
    for arg, default_item in (
        ('strides', 1),
        ('crops', 0),
        ('dropouts', 0),
        ('maxpools', 1)
    ):
      if locals()[arg] is None:
        locals()[arg] = [default_item]*n_blocks
...

But it does not work because locals here is immutable

[quartz wave]

it isn't immutable??

it's never immutable

what you probably want to do is to replicate locals() to the localsfast array

[earnest wing]

local variables are not internally stored as strings

they're indices to a compile time array

you can hack around it but it takes more work

[quartz wave]

compile time??

they're not compile time

[earnest wing]

Function compile time, my bad

which is runtime

[quartz wave]

that's still confusing but ok

[restive void]

After assigning to locals, call

ctypes.pythonapi.PyFrame_LocalsToFast(
    ctypes.py_object(frame),
    ctypes.c_int(0),
)

[fleet bridge]

This should be pinned

And other ctypes tricks (like changing object type)

[sonic birch]

Thanks a lot!

[sonic birch]

Just checked out. This works

import ctypes
import sys

def foo():
  locals()['a']=5
  frame =sys._getframe()
  ctypes.pythonapi.PyFrame_LocalsToFast(
    ctypes.py_object(frame),
    ctypes.c_int(0),
  )
  print(locals()['a'])

foo()

But this doesn't:

import ctypes
import sys

def foo():
  locals()['a']=5
  frame =sys._getframe()
  ctypes.pythonapi.PyFrame_LocalsToFast(
    ctypes.py_object(frame),
    ctypes.c_int(0),
  )
  print(a)

foo()

[fleet bridge]

I think you first should tell the compiler that "a" variable exists

if 0:a=... should do it

Just insert this line somewhere in your function

[magic wraith]

!e py import ctypes, sys def foo(): 1 == 0 and (a := ...) locals()['a'] = 5 ctypes.pythonapi.PyFrame_LocalsToFast( ctypes.py_object(sys._getframe()), ctypes.c_int(0) ) print(a) foo()

[night quarry]

@magic wraith :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 7, in <module>
003 |     foo()
004 |   File "/home/main.py", line 6, in foo
005 |     print(a)
006 |           ^
007 | UnboundLocalError: cannot access local variable 'a' where it is not associated with a value

[magic wraith]

nah forget it python bot said i can use 3.10 to run this code but when i do !eval 3.10 code it still runs on 3.11

[magic wraith]

just trust that it works on 3.10

[fleet bridge]

I have no idea. I hope some esoteric guys can help run it on 3.11

[restive void]

!e

import ctypes, sys
def foo():
    1 == 0 and (a := ...)
    fr = sys._getframe()
    fr.f_locals["a"] = 5
    ctypes.pythonapi.PyFrame_LocalsToFast(
        ctypes.py_object(fr),
        ctypes.c_int(0) )
    print(a)
foo()

[night quarry]

@restive void :white_check_mark: Your 3.11 eval job has completed with return code 0.

5

[magic wraith]

still, how the hell do i run 3.10 code here?

!eval

[night quarry]

Missing required argument

code

Command Help

!eval [python_version] <code, ...>
Can also use: e

Run Python code and get the results.

This command supports multiple lines of code, including formatted code blocks. Code can be re-evaluated by editing the original message within 10 seconds and clicking the reaction that subsequently appears.

The starting working directory /home, is a writeable temporary file system. Files created, excluding names with leading underscores, will be uploaded in the response.

If multiple codeblocks are in a message, all of them will be joined and evaluated, ignoring the text outside them.

By default, your code is run on Python 3.11. A python_version arg of 3.10 can also be specified.

We've done our best to make this sandboxed, but do let us know if you manage to find an issue with it!

[versed eagle]

it used to be "!e 3.10 <code>"

idk if it still is

[gleaming linden]

https://github.com/python-discord/bot/pull/2618 multi version is temporarily removed

[quartz wave]

just annotate it?

instead of 1 == 0 and (a := ...) you can just do a: ...

[vestal meteor]

(( duplicating my message from general to here since it probably fit the theme more

Hi I'm a not really a regular python user but I got bored and one of my friend give me a coding challenge
I came up with a solution myself but we wonder if there are better solutions to the problem
Here's the problem statement if anyone is interested:
The input is a single line of positive integers separated by whitespace
i.e.
1 3 59 48 958 50 27 139 9580 932 13
the output is the factorization of the product of every number in the input,
in the following format

1. list each prime factor in following format: (prime)P(exponent), i.e. 139P4
2. list primes that with higher exponent first, i.e. 139P4 1187P1
3. list bigger prime first if they have the same exponent

here's a complete I/O example
input 6531 12704461 1907306709 406917357
output 7P4 139P3 11P3 3P3 311P2 1187P1 191P1 29P1 23P1 19P1

Here's the catch: the code HAS to be in one line of code, and no semicolons are allowed,
dependencies asides from the standard library of python are also forbiddened
so __import__("itertools") is fine, __import__("numpy") is not allowed

Here's my (not really readable) solution to the problem

print((lambda c: " ".join([f"{i[0]}P{i[1]}" for i in sorted(c.items(), key = lambda a: (-a[1],-a[0]))]))(__import__("functools").reduce(lambda a,b: a+b,map(__import__("collections").Counter, (ff := lambda a, b, f, x: (m := ff(a, (a + b)//2, f, x), ff((a + b)//2+1, b, f, m) if m[2] != 1 else m)[1] if a != b else f(x), f := lambda x: ((x[0]+[x[2]], x[1]//x[2], x[2]) if x[1]%x[2]==0 else (x[0], x[1], x[2]+1) if x[2]**2<x[1] else (x[0] + [x[1]], 1, 1) )if x[1] != 1 else (x[0], 1, 1), factors := lambda n: ff(1,int(n**0.5),f,([], n, 2))[0],(factors(i) for i in map(int, input().split(' '))))[3]))))

Here's a breakdown of my code
This is the factorization function

(ff := lambda a, b, f, x: (m := ff(a, (a + b)//2, f, x), ff((a + b)//2+1, b, f, m) if m[2] != 1 else m)[1] if a != b else f(x), f := lambda x: ((x[0]+[x[2]], x[1]//x[2], x[2]) if x[1]%x[2]==0 else (x[0], x[1], x[2]+1) if x[2]**2<x[1] else (x[0] + [x[1]], 1, 1) )if x[1] != 1 else (x[0], 1, 1), factors := lambda n: ff(1,int(n**0.5),f,([], n, 2))[0])

It could be way shorter if the recursion limit doesn't exists, but it does
so I wrote the function ff := lambda a, b, f, x: (m := ff(a, (a + b)//2, f, x), ff((a + b)//2+1, b, f, m) if m[2] != 1 else m)[1] if a != b else f(x) that kinda divide and conquer the recursion

I did acknowledge there's the sys.setcursionlimit() method

however that kinda felt like cheating and didn't work well on my friend's machine

f := lambda x: ((x[0]+[x[2]], x[1]//x[2], x[2]) if x[1]%x[2]==0 else (x[0], x[1], x[2]+1) if x[2]**2<x[1] else (x[0] + [x[1]], 1, 1) )if x[1] != 1 else (x[0], 1, 1)
is the actual factorization part

lastly, factors is just a wrapper that makes the code more readable

after we have the list of prime factors, I used collections.Counter to count many times a prime appeared

that's about it

[vestal meteor]

I'm mainly curious about whether there's a better implementation of factorization possible in one line of code

[fleet bridge]

Forbidding semicolon is boring. Conversion from regular code to expression is straightforward in most of cases.

Forbidding colon : is funnier

[earnest wing]

the task is interesting regardless of if it's possible to create a different version of it

(gate's open)

[vestal meteor]

That would be an entirely different challenge I guess

[fleet bridge]

Yeah

[vestal meteor]

no colon: no lambda functions, no for/while loop

[fleet bridge]

No walruses

[vestal meteor]

not really sure if that's even possible ...

[fleet bridge]

Same

[vestal meteor]

It's probably to rewrite the whole thing without the := operator with Z-combinator

however the hack I used to avoid the recursion limit wouldn't work without the := operator

[vestal meteor]

It's is indeed straightforward now I've finished it, however it took me a whole day

[fleet bridge]

You may find this interesting. This code also uses no :, but it implements brainfuck interpreter

[vestal meteor]

Woah

[fleet bridge]

.

[proper vault]

factor = lambda n: [div
        for ln in [[n]]
        for _, div in zip(iter(lambda: ln[0], 1), range(2,n+1))
        for _ in iter(lambda: ln[0] % div == 0 and ln.__setitem__(0,ln[0] // div), False)][::-1]
```this should work for factorisation without any recursion, using this algorithm
```py
for i in range(2, n):
    while n % i == 0:
        n //= i
        yield i
    if n == 1:
      break

[vestal meteor]

interesting

that's kinda what I expected to get as a better answer

i don't really know what's going on with the code though

[proper vault]

the trick is using the two arg form of iter for while loops and early exits

nested comprehension you read from top to bottom, akin to

for ln in [[n]]:
    for _, div in ...
        for _ in ...:
            yield div

[sick hound]

Why i cant speak in voice chat?

Im newbie

[proper vault]

You need a certain number of messages in text conversation before being allowed to join VC. Just be active for a couple days and you'll be allowed. For future reference, these questions are best asked in one of the off topic channels or #community-meta , rather than this channel.

[sick hound]

Ok

Get it

[quartz wave]

f=lambda n:[d for l in[[n]]for _,d in zip(iter(lambda:l[0],1),range(2,n+1))for _ in iter(lambda:l[0]%d==0!=l.append(l.pop()//d),0)][::-1]
``` i'll drop this here

[magic wraith]

lambda: l[0]%d == 0 != l.append(l.pop()//d)```

is this just py if l[0] % d == 0: l.append(l.pop()//d)?

[proper vault]

f=lambda n:[d for l in[[n]]for _,d in zip(iter(lambda:l[0],1),range(2,n+1))for _ in iter(lambda:l[0]%d==0,0)for l[0]in[l[0]//d]][::-1]
```can be shorter via for assignment

[proper vault]

ye, using chained comparisons

works sth akin to l[0]%d == 0 and 0 != l.append(...)

[magic wraith]

ah so basically l[0]%d or l.append(...)

[proper vault]

not quite, the result needs to predictable in the false case for iter(..., sentinel) to be usable.

here you get some int in range(1,d)

you could do something akin to bool(l[0]%d) or l.append and it would behave I am pretty sure.

[magic wraith]

ohh the sentinel zero means iter ends when zero is reached, right?

[proper vault]

yes

or well, False, but False == 0, so it's shorter and still works

[magic wraith]

i get it
l.append returns None which is not == 0

[torn tulip]

this is broken

[kindred stone]

what did you tried? it's only meant to work with hello world

[torn tulip]

ah that's why

[rich field]

anyone rn help me with a task im on exercism practising but the first task itself is idk annoyiing

[ebon hinge]

pls help

[distant nymph]

You did not define "entry"? And this is probably the wrong channel for generic help.

[ebon hinge]

@distant nymphwhere the channel for generic help ?

[restive void]

#❓｜how-to-get-help

[distant nymph]

#1035199133436354600 ?

[dire galleon]

https://pyobfusc.com/#home

[old light]

1. this is sick
2. rules too vague, you can get free points by looking up bugs closed in CPython 3.9-3.11 and abusing them

If your entry demonstrates a bug in CPython, it will be given considerable extra consideration

[dire galleon]

perhaps there's naivety and ambiguity in the rules because it's the first time they've organised such a competition

[old light]

Indeed, that's why I'm pointing it out

[sick hound]

Hello

Any1 awake?

[gleaming linden]

yes

it's 3:47pm here

[arctic skiff]

13:57 here

[fleet bridge]

11:51 here

[fleet bridge]

Wait what. How are you shifted by half an hour?

[lyric fern]

9:54AM here

[restive void]

There's half- and even quarter-hour offsets to UTC.

[fleet bridge]

Are there any 1/π-hour offsets?

[arctic skiff]

??

+5:30 GMT

[digital dirge]

So i just found out you can use lambdas as decorators lmao

!e

class Stuff:
    val = 0
    @lambda f: (lambda s: [s.__setattr__("val", s.__getattribute__("val") + 1), f(s)][-1])
    def cool_func(self):
        return self.val

stuff = Stuff()

print(stuff.cool_func())
print(stuff.cool_func())
print(stuff.cool_func())

[night quarry]

@digital dirge :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 1
002 | 2
003 | 3

[restive void]

!pep 614 yep, since this was merged

[night quarry]

**PEP 614 - Relaxing Grammar Restrictions On Decorators**

Link

Status

Final

Python-Version

3.9

Created

10-Feb-2020

Type

Standards Track

[digital dirge]

Woooah

Any valid expression

I gotta start reading the merged peps

[nimble mirage]

How to use __set_name__ at module level? (descriptors outside classes is not a thing)

[fleet bridge]

Declare your own class, then change type of module?

[earnest wing]

even if you add your custom module type to sys.modules first?

[nimble mirage]

custom module type? ik of ModuleType, idk it could be subclassed

[earnest wing]

you can set sys.modules[modulename] to be any type and it'll affect the rest of the module execution

[umbral moss]

Does anyone know how to create a dictionary of list of numpy array jn numba?

[autumn birch]

im not sure whats going on with my gui ive been trying to use the root.iconbitmap for a ico format yet its always saying that its not there..

[low lynx]

this is not the right place for python help

try #1035199133436354600

[autumn birch]

i figured out how to get the background to look like a normal.. right my bad

[magic wraith]

i'm not lost in this server, i think i might get a best answer here
python docs for marshal mention that unmarshalling data is not secure, however the only reason i want to do that is to parse .pyc files, is there even any possibility for marshal attack?

might ask in #cybersecurity too

[restive void]

Where did the pyc files come from? If you'd be fine blindly executing the source .py files, you'd be fine. If you want to unmarshall random untrusted pyc files, that would be a bad idea.

[zinc niche]

!e

from ctypes import py_object
import builtins
class fakeglobals(dict):
  __slots__ = ()
  def __missing__(self, key, b=builtins, h=hasattr, g=getattr):
    if h(b, key):
        return g(b, key)
    return lambda arg: g(b, arg)(key)
py_object.from_address(id(globals()) + 8).value = fakeglobals

hello_world("print")
this_is_cursed("print")```

[night quarry]

@zinc niche :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | hello_world
002 | this_is_cursed

[zinc niche]

inspired by this

[tepid matrix]

wow thats cool

[restive void]

@low lynx re: #python-discussion message

from contextlib import contextmanager
import fishhook

@contextmanager
def modulo(n):
    @fishhook.hook(int)
    def __add__(self, other):
        return fishhook.orig(self, other) % n
    yield
    fishhook.unhook(int, "__add__")

a = 17
b = 16
with modulo(13) as m_13:
    c = a + b

print(a, b, c)

prints 7.

[old light]

that's awesome

in both the present-day and the previous meanings of the word

[low lynx]

I didn't know about unhook

[restive void]

Annoyingly, pre-allocated ints mean that

with modulo(13):
    x = 17
print(x)

still prints 17, even if you hook __new__.

[low lynx]

yeah

[old light]

python eDSL-in-contextmanager lib when O.o

[restive void]

Fixable, but I can only think of doing so with frame stuff

[old light]

can you @fishhook.hook(17)?

and then just repeat for 0-255

[restive void]

nah, you can only hook classes. You could of course change 17 to be 4, but that would affect all other uses of 17, so the lower the number you're changing, the higher the chances your python will crash.

[earnest wing]

This is rad, we should use context managers to redefine more "innate" contextual information >:)

[restive void]

I (relatively) recently wrote a tiny library to customize execution inside context managers called dont.

with dont:
    1/0

doesn't crash. You can then subclass dont to do whatever you want with the codeblock.

!pypi astral-projection This one uses that lib to do transparent execution on a remote host :D

[night quarry]

astral-projection v0.1.1

Run code inside a context manager on a remote machine

[old light]

hmmm.... could you capture the locals before and after the yield and mutate the difference? or is that what you meant by "frame stuff"

or, I guess, I don't know if x would get propagated up to that scope in the first place

[restive void]

Yep, that's sort of what I meant (although I was thinking about trace functions, because you can't just do it to every local, only to those assigned in the context manager).

[old light]

right, why I suggested taking the difference of the sets captured before and afterwards. cool stuff

[restive void]

It wouldn't go in that scope, which is why you'd need to use sys._getframe(1) to get access to the calling scope :P

[old light]

ahh, makes sense!

only other time I've used sys._getframe was implementing backtraces in logging lol

[earnest wing]

x = None
with finnish:
    joka i sisällä kantama(10):
        jos x on Mitätön:
            x = i
        muuten:
            x += i

unfortunately parsers

[restive void]

I remember seeing a talk at a PyCon.DE where some guy made a custom version of CPython to learn German with stuff like

importiere zufall
für x in bereich(zufall.zufganzzahl(1, 10)):
    drucke(x)

[earnest wing]

there was that meme where rust was translated into a billion languages, each with its own macro to embed inside your projects

[restive void]

If only PEP 638...

(I actually don't like PEP-638's syntax, but macros would be cool.)

[fleet bridge]

r=1;{r:=r*i+r for i in range(int(input()))};print(r)

factorial

stdin: n
stdout: n!

can you golf it further?

r=1;{r:=r*i+r for i in range(int(input()))};print(r)
import math;print(math.factorial(int(input()))) # boring

[quartz wave]

-6 multiline ```py
T=int(input());r=i=1
while i<T:r*=i
print(r)

[low lynx]

isn't that -8

also you need r<T

and a +=1 somewhere

f=lambda n:n<1or n*f(n-1);print(f(int(input())))

48, but prints True for 0

f=lambda n:0**n or n*f(n-1);print(f(int(input())))

is 50 but prints 1 for 0

[fleet bridge]

it doesnt work? it is effectively while True: loop

[quartz wave]

ok

[fleet bridge]

<stdin>:1: SyntaxWarning: invalid decimal literal
``` 😕

[low lynx]

not an error

so it's fine

[last locust]

It's just a warning in 3.11, but 3.12 it errors

[frail token]

zhpy3 uses pyparsing to convert Python keywords written in Chinese back to their standard English forms.

Oh, I guess that was a Python 2 example, this is pretty old... There is a Python-3 version also.

[quartz wave]

ok then -3 ```py
T=int(input());r=i=1
while i<T:i+=1;r*=i
print(r)

[fleet bridge]

[zhpy3](https://pypi.org/project/zhpy3/) is this supposed to work as short link?

[frail token]

I'm still learning the Discord markup

Wow, that is old! Better link here: https://pypi.org/project/zhpy/

[dreamy pier]

-3

i=v=1
exec("v*=i;i+=1;"*int(input()))
print(v)

[fleet bridge]

n=int(input());r=1
while n:r*=n;n-=1
print(r)
``` 45

[digital mesa]

putting a semicolon isn't less than a newline character (at least at leftmost level)

(and also counts as a line)

so in the while loop it is a save

[dreamy pier]

true, but putting a newline instead of a semicolon isn't less too : )

[digital mesa]

(note "at least at leftmost level")

semicolon in an indented block saves one char or more over newline + indent

[dreamy pier]

ye

[cosmic ruin]

def sla():
    def main():
        print("opa")
        return sla
    return main

sla()()()()()()()()()()()()()()()()()()()()()```

[pearl socket]

!e

𝒞 = 500
print(𝒞)
print(C + 𝒞)
print(C / 𝒞)
print(C == 𝑪 == 𝒞 == 𝓒 == 𝕮)

[night quarry]

@pearl socket :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 500
002 | 1000
003 | 1.0
004 | True

[fleet bridge]

!e 𝒞 = 500; print(C)

[night quarry]

@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.

500

[fleet bridge]

unicode normalization tricks...

!e c, с = 1, 2; print(с, c)

[night quarry]

@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.

2 1

[serene stratus]

Oh no why does this work

[last locust]

Because they're different characters that just look the same

!charinfo cс

[night quarry]

Character Info

\u0063 : LATIN SMALL LETTER C - c
\u0441 : CYRILLIC SMALL LETTER ES - с

Full Raw Text

\u0063\u0441

[serene stratus]

No I'm talking about the characters that look different but are treated as the same

Is there some conversion table it uses?

[last locust]

Oh, yeah

It's just a specific unicode conversion (can't remember the algo)

[fleet bridge]

https://en.wikipedia.org/wiki/Unicode_equivalence#Normalization?wprov=sfla1

[old light]

Please tell me this trips all the linters 😭

[last locust]

VSC itself will "warn" you about this, but actual linters (e.g. ruff) don't seem to care (it is valid code tbf)

Ah no, Ruff does error actually

Says C isn't defined

[old light]

oh no

That's... probably a bug to submit

Just makes me think of the classic "Wow, how'd you get generics in go?" "Those are Canadian Aboriginal Syllabics" post

[last locust]

astral-sh/ruff#942 MIGHT BE RELATED

[glass drum]

GitHub

[ruff] #942 F821 triggers when using non-ASCII characters

[old light]

https://github.com/astral-sh/ruff/issues/942#issuecomment-1331568484 hm

(How do you kill an embed on mobile O.o)

[last locust]

Wrap the link in <>

https://www.google.com -> <https://www.google.com> etc.

Decided to create astral-sh/ruff#5003, guess we'll see what happens

[glass drum]

GitHub

[ruff] #5003 F821 False Positive when using equivalent script characters

[last locust]

Honestly wouldn't be surpised if they decided it's not worth implementing, given how niche it is

[icy cypress]

This really cool

[pearl socket]

!e

for a in range(255):
    for b in range(255):
        try:
            globals()[bytes([a, b]).decode('utf-16')] = a + b
        except UnicodeDecodeError:
            pass

print(''.join([chr(i) for i in [䠀,〵,ၜ,ᕗ,ᕚ,ᰐ,ᘊ,眀,月,ᑞ,ൟ,ţ,ᐍ]]))

[night quarry]

@pearl socket :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, world!

[unique heath]

i wanna obfuscate something useless again

since when was clean a thing here

[digital mesa]

map is less chars

[pearl socket]

in that case, just writing print('Hello, world!') is also less chars

[digital mesa]

great point! however, there is no fun in that

[low lynx]

map does look nicer there imo

filter will never look nicer however

[proper vault]

filter(None,...) is occasionally neat

[astral rover]

match type := next(filter(None, (test(headers) for test in tests)), None): 😋

[restive void]

Isn't filter(bool, ...) clearer? I guess this is the wrong channel for that..

[earnest wing]

filter(None, ...) might predate it tbh

[low lynx]

i just have an irrational hatred of filter

[restive void]

How so? filter(bool, ...) isn't special-cased

[proper vault]

bool is a 2.2 feature afaik

[dreamy pier]

unicode normalization is weird..

!e

if False: ...
elsｅ: ...

[night quarry]

@dreamy pier :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[dreamy pier]

!e

if False:
    ...
elsｅ:
    ...

[night quarry]

@dreamy pier :x: Your 3.11 eval job has completed with return code 1.

001 |   File "/home/main.py", line 3
002 |     elsｅ:
003 |            ^
004 | SyntaxError: invalid syntax

[rough moat]

factorial = lambda n:([x:=1]+[x:=x*i for i in range(1,n+1)])[-1]
euler = lambda steps:sum(1/([x:=1]+[x:=x*i for i in range(1,n+1)])[-1]for n in range(steps))

[rough moat]

i was told this was just impossible to do, but here's mandelbrot in one line (no semicolons):

mandelbrot=lambda:([s:=[],w:=50,h:=50,max_iter:=100,r:=(-0.7,0.7),i:=(0,1)],"\n".join(["".join([[".,*3456789abcde."[int(([n:=0,z:=0]+[[z:=z*z + complex(r[0]+(x / w)*(r[1] - r[0]),i[0] + (y / h) * (i[1] - i[0])),n:=n+1] for _ in range(max_iter) if abs(z) <= 2 and n < 100])[-1][1]*(0.11))]][-1]*2 for x in range(-w,w)]) for y in range(-h,h)]))[-1]

lemme shorten it a bit

mandelbrot=lambda:([s:=[],w:=50,h:=50,max_iter:=100,r:=(-0.7,0.7),i:=(0,1)],"\n".join(["".join([[".,*3456789abcde."[int(([n:=0,z:=0]+[[z:=z*z+complex(r[0]+(x/w)*(r[1]-r[0]),i[0]+(y/h)*(i[1]-i[0])),n:=n+1]for _ in range(max_iter)if abs(z)<=2])[-1][1]*0.11)]][-1]*2 for x in range(-w,w)])for y in range(-h,h)]))[-1]

[low lynx]

who told you it was impossible?

like I said there's literally a counterexample on the python docs

[rough moat]

i think it was fiery

[unique heath]

!e mortals

(___:=eval(__builtins__.__dir__().__getitem__((420).__floordiv__(__import__.__dir__().__len__().__sub__(20)))),_____:=lambda __: eval(__builtins__.__dir__().__getitem__(14))(__),eval(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__getitem__(__import__.__dir__().__len__().__sub__(__name__.__doc__.__len__().__floordiv__(__builtins__.__doc__.__len__()))).__len__()).__sub__(__name__.__dir__().__len__().__sub__(__import__.__dir__().__len__()).__sub__(__package__.__dir__().__len__()).__sub__(__package__.__dir__().__len__())).__add__(4)))(chr(__import__.__dir__().__len__().__add__(__builtins__.__dir__().__len__())), (eval(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__getitem__(__import__.__dir__().__len__().__sub__(__name__.__doc__.__len__().__floordiv__(__builtins__.__doc__.__len__()))).__len__()).__sub__(10))), ), {chr(__name__.__dir__().__len__()):(lambda _: ___(_____(72).__add__(_____(101)).__add__(_____(108)).__add__(_____(108)).__add__(_____(111)).__add__(_____(32)).__add__(_____(119)).__add__(_____(111)).__add__(_____(114)).__add__(_____(108)).__add__(_____(100)).__add__(_____(33))))})().Q())```