esoteric-python

[marble gorge]

h

[royal whale]

what dark being does this conjure

and how

[quartz wave]

@rapid sparrow managed to fix if statements (with and without else clauses)
can you review the PR cause i have to sleep now

[unreal echo]

given a list of floats, return the indices of the 3 largest

i could write something for this, but i gotta do something meanwhile

[burnt pasture]

with_indexes = [(f, i) for i, f in enumerate(the_floats)]; with_indexes.sort(); return [i for f, i in with_indexes[:3]]

[unreal echo]

ned ⁉️

in #esoteric-python ⁉️

[burnt pasture]

this question wasn't esoteric 🙂

[unreal echo]

well, i wanted a one liner

so it had the potential

np.array([i[1] for i in sorted(enumerate(arr), key=lambda x: x[1], reverse=True)[:num]])came up with something already anyway

[burnt pasture]

one-liners are overrated, and it's easy to make anything into a one-liner.

[floral meteor]

I have been summoned

[minor topaz]

h😀😀😀

[earnest wing]

reduce max-and-shift using a 3-length queue initialized with -Infinity

[marble gorge]

death.py? i wonder what that does

[floral meteor]

fatal error

it's in the name

[rugged sparrow]

[i for i, f in sorted(enumerate(floats), key=lambda t:t[1])][::-1][:3]

[floral meteor]

although sometimes a blue screen of death

[marble gorge]

ok

souds safe

[floral meteor]

If you run it, your computer probably won't blow up

[marble gorge]

ok

what are the system requirements?

[floral meteor]

maybe

Windows 11

or 10

[marble gorge]

i have 2370M cpu, Win 10, 8G ram, intel hd graphics 3000

[floral meteor]

idk maybe win7 or linux might work

the more ram, the longer it might take

[marble gorge]

ok

[floral meteor]

opposite of the conventional more ram=less time

[marble gorge]

a = "."
for a in range(30)
  a += a

file = open("file.txt","a")
for b in range(1000):
  file.write(a)
file.close()

I think your computer would like it
Disclaimer: Don't run

[burnt pasture]

I think my computer will raise a TypeError

[marble gorge]

i always make the mistake

i do for loop with a always for some reason

[burnt pasture]

now it raises SyntaxError, then it will raise TypeError 🙂 details are hard!

[marble gorge]

ok

i mean good thing that it does not work

[quaint gull]

that gives me a type error ;-;

[marble gorge]

it could also damage your hard drive

[bronze agate]

yum, 1000 delicious GiB for my ssd if i'm not mistaken

[earnest wing]

it could damage your drive if it's 25 years old and in heavy use

[marble gorge]

well, yes

1000 GB is nothing

[bronze agate]

it sounds like you're able to have 2, maybe 3 triple a titles installed at the same time, well done

[marble gorge]

wdym

i have just one drive that has just 140 GB

i meant 1000 GB is nothing if you want to damage the drive

[bronze agate]

true, it's repeated write cycles and not one wide write that hurts for ssds, i'm not sure what would be harmful for hdds though, aside from dropping them on the floor

[languid night]

dropping them on the floor then stepping on them

[modern fiber]

print = open('3d_engine/objs/faced_suzanne.obj', 'r').readlines()
from itertools import chain
import turtle as open
open.tracer(0, 0)
for input in tuple(tuple(tuple([list(map(float,_[2:].split()))for _ in print if _[:2]=='v '][dict])for dict in len)for len in[[int(dict.split("/")[0])-1 for dict in len]for len in[_[2:].split()for _ in print if _[:2]=='f ']]):open.penup();open.goto((input[0][0]*-200*1/2**(1/2)-input[0][2]*(10*2**(1/2))**2*1/2**(1/2),(input[0][0]*-200*1/2**(1/2)+input[0][2]*(10*2**(1/2))**2*1/2**(1/2)+input[0][1]*40**2/8*2**(1/2))/3**(1/2)));open.pendown();[open.goto((_[0]*-200*1/2**(1/2)-_[2]*(10*2**(1/2))**2*1/2**(1/2),(_[0]*-200*1/2**(1/2)+_[2]*(10*2**(1/2))**2*1/2**(1/2)+_[1]*40**2/8*2**(1/2))/3**(1/2)))for _ in input]
open.hideturtle()
open.update()
open.exitonclick()

visualize .obj files wireframe in isometric view

using turtle

[valid delta]

why do you redefine print

[modern fiber]

i wanted to see if i could push that limit

[valid delta]

and then import turtle as open

[modern fiber]

same reason

[valid delta]

kinda weird

[modern fiber]

ikr

screenshot of the output

[serene stratus]

Damn good job

Looks really good

[modern fiber]

thanks

[sick hound]

!e

f'{(hw:="++++++++++[>+++++++>++++++++++>+++>+<<<<-]>++.>+.+++++++..+++.>++.<<+++++++++++++++.>.+++.------.--------.>+.>.")}{(bf:=lambda prgm:print(((pc:=0),(ti:=0),(t:=[0]*30000),(rs:=[]),(o:=""),(*(((char:=prgm[pc]),(((ti:=ti+1),(pc:=pc+1))if char==">"else(((ti:=ti-1),(pc:=pc+1))if char=="<"else(((t.__setitem__(ti,(t[ti]+1)&0xFF)),(pc:=pc+1))if char=="+"else(((t.__setitem__(ti,(t[ti]-1)&0xFF)),(pc:=pc+1))if char=="-"else(((o:=o+chr(t[ti])),(pc:=pc+1))if char=="."else(((t.__setitem__(ti,ord(input())%0xFF)),(pc:=pc+1))if char==","else(((rs.append(pc),(pc:=pc+1))if t[ti]!=0 else((bc:=1),(*(((pc:=pc+1),((bc:=bc+1)if prgm[pc]=="["else(bc:=bc-1)if prgm[pc]=="]"else...),(pc:=pc+1))for _ in iter(lambda:bool(bc),False)),)))if char=="["else((pc:=rs.pop())if char=="]"else...)))))))))for _ in iter(lambda:pc<len(prgm),False)),))[-1][-1][-1][0]))}{(bf(hw))}'

[night quarry]

@sick hound :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello World!

[sick hound]

one line brainf*** hello world i guess

"but you could use globals().__setitem__ instead of walrus!!" i enjoy walrus operator.

[bronze agate]

the funny operator

[earnest wing]

I also like the ^_^ operator

[fleet bridge]

I wanted to say that this is the invalid syntax, but then I realized...

[finite blaze]

Heyo

another golfing challange

Miller Rabin test (it checks if the given number is prime

!e

import random
def miller_rabin(n, k):
    if int(repr(n)[-1]) % 2 == 0:
        return False
    else:
        d = n - 1
        s = 0
        while d % 2 == 0:
            d = d // 2
            s += 1
            
        for i in range(k):
            r = random.randint(1, n - 1)
            for j in range(s):
                x = pow(r, d * 2**j, n)
                if ((j == 0 and x == 1)
                or (x == n - 1)):
                    break
                elif (j == s - 1):
                    return False
        return True
print(miller_rabin(11,7))

[night quarry]

@finite blaze :white_check_mark: Your 3.11 eval job has completed with return code 0.

True

[royal whale]

repr() to str()- one character less

[proper vault]

import random
def miller_rabin(n, k):
 if~n&1:return False
 d=n-1;s=0
 while~d&1:s+=1;d//=2
 while k:
  k-=1;r=random.randint(1,n-1)
  for j in range(s):
   x=pow(r,d<<j,n)
   if x==n-1or j==0and x==1:break
   elif j==s-1:return False
 return True
assert miller_rabin(11,7)
```haven't done code golf in a while, but this should be equivalent

lots of room for improvement still

[finite blaze]

mr= lambda t:[n:=12,k:=7,r:=(0 if int(repr(n)[-1])%2==0 else 1)]and None

I wonder how can I return r from this lambda?

[proper vault]

and rshould be fine afaict

[finite blaze]

yes

perfect

~n&1

whats this?

[proper vault]

check if number is even, honestly don't remember if it's the best way to do that

n&1 checks if last bit is 1, that is is number odd, and ~n inverts all bits

[finite blaze]

i am doing int(str(n)[-1])%2==0 cuz I was thinking that it might be faster than n % 2==0 if the number is realllllyyyy big

[proper vault]

it won't be

[finite blaze]

oke

[proper vault]

allocating the buffer for the string will be far slow than any amount of bigint math

[finite blaze]

d<<j = d*2**j

interesting

is it possible to use while inside lambda?

[restive void]

No, too bad there are no while comprehensions..

[proper vault]

you can do a listcomp over a iter(lambda: cond, False)

[finite blaze]

so i guess i'll just do a for loop instead

mr= lambda t:[n:=11,k:=7,r:=(False if ~n&1 else d:=n-1,s:=0, [d=d//2,s+=1]for _ in iter(~d&1, False))]and r

so it is telling me that d is not defined

[d=d//2,s+=1] here

but everything is okay with s+=1

[floral meteor]

what does t do?

[last locust]

I think you might even be able to remove the space in j==s-1 for

Like you did with the ~n&1 else

[floral meteor]

yes, this cheap syntax highlighting makes that less obvious

[last locust]

Yeah lol

Tbf even idle doesn't like it

[floral meteor]

It takes too long to run, there's an infinite loop somewhere

[floral meteor]

[-1]

assert(m:=lambda n,k:0if~n&1else(lambda d,s,R,k=[k]:([(s:=s+1,d:=d>>1)for _ in iter(lambda:d&1,1)],not any([(k.__setitem__(0,k[0]-1),r:=R(1,n-1),any([j==s-1for j in range(s)if not((x:=pow(r,d<<j,n))==n-1or j==0and x==1)]))[-1]for _ in iter(lambda:k,[0])]))[-1])(n-1,0,__import__('random').randint))(11,7)

mildly tested

[quartz wave]

 not((x:=pow(r,d<<j,n))==n-1or j==0and x==1)
``` might be ```py
~(((x:=pow(r,d<<j,n))==n-1)|(j==0)&(x==1))

[floral meteor]

those are only bitwise though, replacing stuff that compares to zero

[quartz wave]

what?

[floral meteor]

!e ```py
if not 2:print('derp')
if~2:print('derp')

[night quarry]

@floral meteor :white_check_mark: Your 3.11 eval job has completed with return code 0.

derp

[floral meteor]

only one derp

so they're not the same thing

[earnest wing]

the arguments are boolean

just because something is in general not equivalent doesn't mean it's not useful here

[floral meteor]

I just read it properly and you are right the arguments are boolean

[devout juniper]

this is so weird but i cant stop watching it

[floral meteor]

the channel?

[floral meteor]

This channel is for the mad scientists of python

[fleet bridge]

!e class X(range(10)): ...

!e ```py
class X(slice(10, 20, 30)): ...
print(X)

[night quarry]

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

slice('X', (slice(10, 20, 30),), {'__module__': '__main__', '__qualname__': 'X'})

[fleet bridge]

I dont understand...

1. Get metaclass: type(base) [it is class slice in this case]
2. Call metaclass: slice(clsname, bases, ns)

Is it how it works?

It is very very weird and counterintuitive

[floral meteor]

lmao you made the cursed slice class

[fleet bridge]

It isnt a class
It is instance of builtin slice

I will try to break mypy using this behavior

Is this way of using class statement documented?

[restive void]

yes

Also works with map and zip

[quartz wave]

combined with (c)python scientists (people who know the internals)

[vague cairn]

But I can't seem to use X for an instance o f slice.

[quartz wave]

wdym by that?

[vague cairn]

after I say:
class X(slice(10, 20, 30)): ...
I cannot seem to use X for anything, except printing

[earnest wing]

metaclass=slice should also work, right?

just with empty bases tuple

[vague cairn]

I cannot call/instantiate an X i cannot use it for an already instantiated slice, what good is it?

[quartz wave]

so like you're saying you wanna use an example/test for something else other than testing?

[vague cairn]

I mean it would be lovely to use class X(slice(something)):...
in place of x=slice(something)
So that when I need to do an inordinate number of
for y in Y[something]:
I can do
for y in Y[X] instead.
but class X(slice(something)):... is so much more essoteric than x=slice(something) so that would be better, but it doesn't seem to work that way, so I'm trying to understand how it is working.

[earnest wing]

that's because it's different

class X(slice(...)) is equivalent to
X = slice(class name, bases tuple, namespace)

[vague cairn]

got it.

[earnest wing]

and since slice(...) appears inside the bases, it's part of the tuple xd

[vague cairn]

!e ```py
x = list(range(100))
class Z(slice(3,6,1)): ...
print(x[Z.stop[0]])

[night quarry]

@vague cairn :white_check_mark: Your 3.11 eval job has completed with return code 0.

[3, 4, 5]

[floral meteor]

The Forbidden Slice

[quartz wave]

!e ```py
@lambda :print(list(range(100))[.stop[0]])
class Z(slice(3,6,1)):...

[night quarry]

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

[3, 4, 5]

[quartz wave]

same thing

[vague cairn]

Much better, yes.

!e```py
#Or I think this is what I'm looking for:
@lambda :.stop[0]
class Z(slice(3,6,1)):...
#and much later...
print([_ for _ in range(100)[Z] ])

[night quarry]

@vague cairn :white_check_mark: Your 3.11 eval job has completed with return code 0.

[3, 4, 5]

[floral meteor]

_ for _ in can be replaced by *

[near gust]

i broke the rpython type annotator https://media.discordapp.net/attachments/702883266998566976/1015433619537416213/unknown.png

[floral meteor]

noice

90% of my blue screens of death happen when I'm running a python script I've written

A game of snake? unfortunately Windows has stopped due to HYPERVISOR_ERROR.
Pretty matrix rain? Nope, irql decides to be not_less_than_or_equal

hehehehehehe

[potent flare]

one line .env parser, can someone do better? :P
parse_env=lambda:(f:=open(".env","r"),{(t:=l.split("="))[0].strip():t[1].strip()for l in f.read().splitlines()},f.close())[1]

[restive void]

For once, you could just not close the file.

[potent flare]

b- but

leak

memowy weak 🥺

[restive void]

not really

[potent flare]

can i just use a context manager

ffs

call __enter__ and __exit__ manually

[proper vault]

parse_env=lambda:{(t:=l.split("="))[0].strip():t[1].strip()for l in open('.env')}

[restive void]

f is only visible from within the lambda, so it will anyways get garbage collected

[potent flare]

damn

ok

[proper vault]

though it does leak the file

[potent flare]

didn't know you could iterate directly from a file objet

[restive void]

why would it?

[proper vault]

in CPython it won't, but I am pretty sure the file stays open even if the iterator ends

[restive void]

Right, okay, not guaranteed by the language, that's true.

parse_env=lambda:dict(map(str.strip,l.split("="))for l in open('.env'))

[earnest wing]

fiwe descwiptow weak

[night quarry]

@mellow salmon :white_check_mark: Your 3.11 eval job has completed with return code 0.

['__class__', '__delattr__', '__dict__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__getstate__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__module__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__']

[fleet bridge]

parse_env=lambda:dict(map(lambda s:map(str.strip,s.split('=')),open('.env')))
parse_env=lambda:dict(map(lambda s:s.split('='),open('.env'))) # without stripping

[soft vapor]

!e

import os
files = [f for f in os.listdir(os.cwdb) if os.path.isfile(f)]
print(f"Files\n{files}")
print(os.listdir(os.cwdb))

[night quarry]

@soft vapor :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 2, in <module>
003 | AttributeError: module 'os' has no attribute 'cwd'

[soft vapor]

!e

import os
path = os.getcwdb() + "/config"
files = [f for f in os.listdir(path) if os.path.isfile(f)]
print(f"Files\n{files}")
print(os.listdir(path))

[night quarry]

@soft vapor :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | Files
002 | []
003 | [b'requirements', b'config', b'user_base']

[soft vapor]

!e

import os
path = os.getcwdb() + b"/config/snekbox.cgf"
#files = [f for f in os.listdir(path) if os.path.isfile(f)]
#print(f"Files\n{files}")
#print(os.listdir(path))
file = open(path, "r")
print(file.read())

[night quarry]

@soft vapor :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | Files
002 | []
003 | [b'snekbox.cfg', b'gunicorn.conf.py', b'__pycache__']

[soft vapor]

!e

import os
path = os.getcwdb() + b"/config/snekbox.cgf"
#files = [f for f in os.listdir(path) if os.path.isfile(f)]
#print(f"Files\n{files}")
#print(os.listdir(path))
file = open(path, "r")
print(file.read())

[night quarry]

@soft vapor :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 6, in <module>
003 | FileNotFoundError: [Errno 2] No such file or directory: b'/snekbox/config/snekbox.cgf'

[potent flare]

!e

print(open("/snekbox/config/snekbox.cfg").read())

[night quarry]

@potent flare :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | name: "snekbox"
002 | description: "Execute Python"
003 | 
004 | mode: ONCE
005 | hostname: "snekbox"
006 | cwd: "/snekbox"
007 | 
008 | time_limit: 6
009 | 
010 | keep_env: false
011 | envar: "LANG=en_US.UTF-8"
... (truncated - too many lines)

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

[potent flare]

!e

import os

print(os.listdir("/snekbox/user_base/lib/python3.11/site-packages"))

[night quarry]

@potent flare :white_check_mark: Your 3.11 eval job has completed with return code 0.

['sniffio-1.2.0.dist-info', 'sympy', 'dateutil', 'pendulum', 'pytzdata', 'toml', 'async_generator-1.10.dist-info', 'beautifulsoup4-4.11.1.dist-info', 'python_dateutil-2.8.2.dist-info', 'pendulum-2.1.2.dist-info', 'arrow', 'idna', 'numpy', 'PyYAML-6.0.dist-info', 'six-1.16.0.dist-info', 'trio-0.21.0.dist-info', 'fishhook', '_yaml', 'arrow-1.2.2.dist-info', '__pycache__', 'anyio', 'numpy-1.23.1.dist-info', 'bs4', 'fuzzywuzzy-0.18.0.dist-info', 'pytz', 'sympy-1.10.1.dist-info', 'outcome', 'mpmath', 'sniffio', 'attrs-22.1.0.dist-info', 'pandas', 'outcome-1.2.0.dist-info', 'pytz-2022.2.dist-info', 'anyio-3.6.1.dist-info', 'typing_extensions-4.3.0.dist-info', 'pytzdata-2020.1.dist-info', 'async_generator', 'networkx', 'attr', 'soupsieve', 'sortedcontainers-2.4.0.dist-info', 'fuzzywuzzy', 'sortedcontainers', 'isympy.py', 'forbiddenfruit-0.1.4.dist-info', 'more_itertools-8.14.0.dist-info', 'more_itertools', 'idna-3.3.dist-info', 'toml-0.10.2.dist-info', 'lark', 'fishhook-0.1.4.dist-info', 'lar
... (truncated - too long)

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

[potent flare]

that's a lot

[last locust]

!e
import os

print(sorted([x for x in os.listdir("/snekbox/user_base/lib/python3.11/site-packages") if ".dist-info" not in x]))

[night quarry]

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

['__pycache__', '_yaml', 'anyio', 'arrow', 'async_generator', 'attr', 'attrs', 'bs4', 'dateutil', 'fishhook', 'forbiddenfruit', 'fuzzywuzzy', 'idna', 'isympy.py', 'lark', 'more_itertools', 'mpmath', 'multidict', 'networkx', 'numpy', 'outcome', 'pandas', 'pendulum', 'pytz', 'pytzdata', 'six.py', 'sniffio', 'sortedcontainers', 'soupsieve', 'sympy', 'toml', 'trio', 'typing_extensions.py', 'tzdata', 'yaml', 'yarl']

[potent flare]

that's not a lot

and wtf is "fuzzywuzzy"

fuzzywuzzy uwu 🥺 👉 👈

[cloud fossil]

!e

import fuzzywuzzy
help(fuzzywuzzy)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 2, in <module>
003 | NameError: name 'help' is not defined

[cloud fossil]

¯_(ツ)_/¯

[proper vault]

fuzzywuzzy is a library for fuzzy matching text IIRC

[cloud fossil]

Why did my Gboard have an auto complete suggestion for it

[versed eagle]

!e
import fishhook

[night quarry]

@versed eagle :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[versed eagle]

yay

they added fishhook to 3.11

finally

[dense marlin]

Hi, I'm trying to compute a property of a class which depends on the classes __mro__ I tried doing something like this

class Foo(Bar):
    x = __mro__[::-1]

but I get x = __mro__[::-1]
NameError: name '__mro__' is not defined
Is there a way to do something like this? at what point are things like __mro__ actually defined / available on classes?

[restive void]

No, they are class attributes, but not available during class definition.

[restive void]

Class decorator would be the easiest way

!e

def orm_as_x(cls):
    cls.x = cls.__mro__[::-1]
    return cls

class Foo:0

@orm_as_x
class Bar(Foo):0

print(Bar.x)

[night quarry]

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

(<class 'object'>, <class '__main__.Foo'>, <class '__main__.Bar'>)

[restive void]

You can do it inside the class definition as well, with a self-replacing descriptor, but if this is enough, I'd use this ^

[potent flare]

no, it's just pre-installed packages with snekbox

[rigid bolt]

You can also use a metaclass or __init_subclass__ but the methods above are preferable.

[fleet bridge]

n = 128;
Graphics3D[Table[Point[{x, y, BitXor[x, y]}], {x, 0, n - 1}, {y, 0, n - 1}]]
``` code for wolfram mathematica

view from top

n = 128;
Image[Table[BitXor[x, y]/n, {x, 0, n - 1}, {y, 0, n - 1}]]
``` code to generate 2d images

here you can see relative speed of euclidean algorithm for computing gcd(x, y)
white = slower, black = faster

set of coprime integers
white = x and y are coprime
black = not coprime

riemann function plot
f(x) = 1/q if x = p/q (and p and q are coprime)
f(x) = 0 if x is irrational

every line = binary representation of 1/x number (x is line number)
black = digit 0, white = digit 1

1. count number of 1's in binary representation of x and y

#x > #y - white
#x == #y - grey
#x < #y - black

same, but white for == and black otherwise

for all rational numbers p/q (p and q are coprime) draw point (p/q, q)

i have wolfram mathematica code for all of these pictures
if you want, i can share it

[night quarry]

Hey @fleet bridge!

It looks like you tried to attach file type(s) that we do not allow (.bmp). We currently allow the following file types: .gif, .jpg, .jpeg, .mov, .mp4, .mpg, .png, .mp3, .wav, .ogg, .webm, .webp, .flac, .m4a, .csv, .json.

Feel free to ask in #community-meta if you think this is a mistake.

[fleet bridge]

another picture without explanation

[versed eagle]

yes, ik what it is
but it used to be that fishhook wasnt in the 3.11 packages

so to use it you had to do !e 3.10

[potent flare]

ah ok, just confused with the wording

[versed eagle]

sorry!
my grammar is bad when im tired

[royal whale]

Wow, nice
'Tis sad that Wolfram Mathematica isn't free

[fleet bridge]

you can use online interpreter, it is free

and you can get 1 week (or 1 month, i dont remember) trial for desktop wolfram mathematica

[burnt pasture]

these pictures are interesting, but what do they have to do with esoteric Python?

[rich osprey]

Hello

import sys;a,b=sys.stdin.read().split();a,b=int(a),int(b);print(min(a,b),max(a,b))```

Is there any way anyone can find to shorten this

by even one character

[next flame]

not sure why you need to handle input separated by arbitary whitespace (instead of just space separated, or just newline separated), but:

a,b=open(0).read().split();print(min(c:=int(a),d:=int(b)),max(c,d))

print(*sorted(map(int,open(0).read().split())))

[rich osprey]

holy smokes

[next flame]

and if you ignore the loose input format from the original code and just assume 2 ints separated by newline:

print(*sorted(map(int,open(0))))

for a total of 32 bytes
(for reference, original code is 83 bytes)

[floral meteor]

I think I corrupted my python installation lmao

If the current directory is desktop, the terminal host crashes with ntstatus division by zero when attempting to run python.

[versed eagle]

huh

[rich osprey]

thank you! I had to change the format bc of how we have to submit the solution but that was really helpful!

[floral meteor]

Never admit to the question being homework when asking for help, because people are less likely to help.

[quartz wave]

!e to whoever asked how to make hello world with type, here you go ```py
d = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4, 2, 0, 0, 0, 0, 0, 0, 0, 0, 5, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 10, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]

from opcode import opmap

globals().update(opmap)

print_hello_world = type(f:=lambda:0)(type(f.code)(0, 0, 0, 0, 2, 1, bytes([
RESUME, 0,
PUSH_NULL, 0,
LOAD_CONST, 1,
LOAD_CONST, 0,
IMPORT_NAME, 0,
LOAD_ATTR, 1,
*[CACHE, 0]*d[LOAD_ATTR],
PRECALL, 0,
CACHE, 0,
CALL, 0,
*[CACHE, 0]*d[CALL],
POP_TOP, 0,
LOAD_CONST, 0,
RETURN_VALUE, 0,
]), (None, 0), ("hello", "main"), (), name, "<manual-code>",
f"{name}.<manual-code>", 1, b"", b""), globals())

print_hello_world()

[night quarry]

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

Hello world!

[versed eagle]

you cant open an int, has to be a string

[languid night]

0 is for stdin? That works

[versed eagle]

>>> print(open(0, "r"))
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: open: path should be string, bytes or os.PathLike, not int

[languid night]

I do not get that error

[versed eagle]

huh

weird

[languid night]

!e print(open(0))

[night quarry]

@languid night :white_check_mark: Your 3.11 eval job has completed with return code 0.

<_io.TextIOWrapper name=0 mode='r' encoding='utf-8'>

[versed eagle]

from python docs:

(talking about open arguments)

file is a path-like object giving the pathname (absolute or relative to the current working directory) of the file to be opened or an integer file descriptor of the file to be wrapped. (If a file descriptor is given, it is closed when the returned I/O object is closed unless closefd is set to False.)

pathlike object:

An object representing a file system path. A path-like object is either a str or bytes object representing a path, or an object implementing the os.PathLike protocol. An object that supports the os.PathLike protocol can be converted to a str or bytes file system path by calling the os.fspath() function; os.fsdecode() and os.fsencode() can be used to guarantee a str or bytes result instead, respectively. Introduced by PEP 519.

doesn't say int is valid

weird

[languid night]

0, 1 and 2 are file descriptors

stdin, stdout and stderr

!e print(open(3))

[night quarry]

@languid night :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 1, in <module>
003 | OSError: [Errno 9] Bad file descriptor

[languid night]

It can accept FDs 0-2

[versed eagle]

ik what they are lmao

[languid night]

So you should understand opening 0 is referring to stdin

I'm not sure why it doesn't work for you

[astral rover]

it accepts more than 0-2 they just need to be open

In [9]: sock = socket.socket()
In [12]: open(sock.fileno())
Out[12]: <_io.TextIOWrapper name=12 mode='r' encoding='UTF-8'>

[versed eagle]

sorry i wasn't there for a bit, i had to go

[versed eagle]

it might be a version thing then?

my sys.version is this: 3.11.0rc1 (main, Aug 8 2022, 18:31:54) [GCC 9.4.0]

[rugged owl]

Whats the point in passing in mode, even though default is “r” already? Or just common practice

[versed eagle]

intent is more understandable

[versed eagle]

turned out my python was messed up

instead of calling __builtins__.open, it was calling os.open, which has different requirements for whats passed to it

¯_(ツ)_/¯

user@host:~$ python3
Python 3.11.0rc1 (main, Aug  8 2022, 18:31:54) [GCC 9.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> open(0)
<_io.TextIOWrapper name=0 mode='r' encoding='UTF-8'>
>>> import os; os.open(0)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: open() missing required argument 'flags' (pos 2)
>>> import os; os.open(0, "r")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: open: path should be string, bytes or os.PathLike, not int 

so what i was getting was a completely different thingy

so at least that got fixed

[unreal echo]

given an array of floating point numbers (any length) return a tuple (ordered greatest to least based on the floating point number) of the index of the item and the floating point number. i.e ```py
def func():...

arr = (0.1, 0.2, 0.05, 0.5, 0.3, 0.8)
print(func(arr))# should output ((0.8, 5), (0.5, 3), (0.3, 4)) ....```

can anyone make a func like that (golfed as much as possible)

[next flame]

lambda x:sorted(zip(x,range(len(x))),reverse=1))

[unreal echo]

thanks

[dry mirage]

!e

import inspect

print(inspect.stack())

[night quarry]

@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.

[FrameInfo(frame=<frame at 0x7fbb336f45c0, file '<string>', line 3, code <module>>, filename='<string>', lineno=3, function='<module>', code_context=None, index=None, positions=Positions(lineno=3, end_lineno=3, col_offset=6, end_col_offset=21))]

[dry mirage]

!e

import inspect

print(5)

x = inspect.stack()[-1]
print(x)
print(x.code_context)

[night quarry]

@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 5
002 | FrameInfo(frame=<frame at 0x7f62742ac680, file '<string>', line 6, code <module>>, filename='<string>', lineno=5, function='<module>', code_context=None, index=None, positions=Positions(lineno=5, end_lineno=5, col_offset=4, end_col_offset=19))
003 | None

[dry mirage]

curious, why no code context in -c

[vague cairn]

Well done, I was thinking:
lambda x:[a,b for b,a in sorted(enumerate(x),key=lambda y:y[1]),reversed=1)] which isn't nearly as good.

and lambda x:sorted(map(reversed,enumerate(x)),reverse=1) hasn't worked since python2

[neat elk]

Hi friends,
How are you?
By the way, I would like to make a birthday wish for my best friend tomorrow, it's her birthday. So I would like to wish her birthday using pyhton (turtle package or any other package)
Would you like to help me?

[vague cairn]

https://www.facebook.com/groups/329660547115351/permalink/5617354591679227/

Like this? 😉

[floral meteor]

lambda x:sorted(zip(x,range(len(x))))[::-1]

golfeder

[earnest wing]

Do you ever write code that seems normal but is in fact terrifying?

def inner(fn: Callable[[T, U], V]) -> InfixOperator[T, U, V]:
    ctx = context(fn)
    if ctx == FunctionContext.Local:
        raise TypeError(
            f"Custom operator '{op}' for '{fn.__qualname__}' cannot be defined in a local scope"
        )
    operator = InfixOperator(op, flipped, fn)
    if ctx == FunctionContext.Global:
        if BUILTIN_OPERATORS.get(op) is not None:
            raise TypeError(
                f"Builtin operator '{op}' for '{fn.__qualname__}' cannot be defined in a global scope\n"
                "Module-scoped operators must not be the same as built-in ones"
            )
        _bind_to(sys.modules[fn.__module__], operator)
    return operator

[quartz wave]

BUILTIN_OPERATORS.get(op) is not None
you know in exists, right?

[earnest wing]

sure

That was a remnant from when it actually did different logic for builtin ops

[frigid wharf]

lambda x:sorted(zip(x,range(9**99)))[::-1]```

[humble rune]

[print(i)for i in range(101)if(lambda n,p=1:[p:=0for i in range(2,n)if n%i<1]*0 or n>1&p>0)(i)]

made this 95 character prime generaator

very pog

[quartz wave]

82 ```py
[print(i)for i in range(101)if(lambda n:n>1>any(1for i in range(2,n)if n%i<1))(i)]

[quartz wave]

78 ```py
[print(i)for i in range(101)if(lambda n:n>1>any(n%i<1for i in range(2,n)))(i)]

[quartz wave]

65 ```py
[print(n)for n in range(101)if n>1>any(n%i<1for i in range(2,n))]

[humble rune]

what the fuck

[proper vault]

61

[print(n)for n in range(2,101)if all(n%i for i in range(2,n))]

[restive void]

!e 79 (not winning, but style points)

[print(i)for i in range(2,101)if not __import__("re").match(r"(11+)\1+$","1"*i)]

[night quarry]

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

001 | 2
002 | 3
003 | 5
004 | 7
005 | 11
006 | 13
007 | 17
008 | 19
009 | 23
010 | 29
011 | 31
... (truncated - too many lines)

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

[versed eagle]

its shorter to do:

import re;[print(i)for i in range(2,101)if not re.match(r"(11+)\1+$","1"*i)]

i like the regex, its not every day that you see regex get used for anything except trying to relearn regex for the 18th time

[restive void]

It's my all-time favorite regex, because famously checking for primal length is impossible with a regular language :)

[proper vault]

yeah, you can't fit a backref that holds an unbounded string into a finite automaton

but that is super cool

[vague cairn]

Ok, yeah, you definitely get style points.

[woven pivot]

also 61

x=2
while 999>x:0<all(x%i for i in range(2,x))==print(x);x+=1

[twin reef]

also 61, but finishes with style

def f(x):all(x%y for y in range(2,x))and print(x);f(x+1)
f(2)

[woven pivot]

55

x=2
while 101>x:103**~-x%x+107**~-x%x==2==print(x);x+=1

[quartz wave]

54 ```py
x=2
while 101>x:103**~-x%x+107**~-x%x<3==print(x);x+=1

[woven pivot]

in that case this works too
50

x=2
while 101>x:4**~-x%x+3**~-x%x<3==print(x);x+=1

[grave rover]

back to working on this cursed project, this time actually being somewhat productive and keeping code clean

(this outputs (1, 2, 9999))

Types are now also easier to define: ```py
class PyTupleObject(PyVarObject):
def init(self, addr: int):
super().init(addr)
self.array("ob_item", lambda _addr: Pointer(_addr, PyObject), self.ob_size) # since ob_size is a property that can be read, this array is of dynamic size :D

[grave rover]

Strings are a pain because their structs aren't 100% consistent

[sick hound]

maybe maybe maybe

[unreal echo]

If I ever think I'm good at python, I just visit this channel

Thank

[rugged sparrow]

that looks really similar to my native_ctypes project from a while back

i did a lot more screwery with metaclasses tho ```py
class TupleObj(PyVarObject):
ob_item: field(lambda inst:py_object*inst.ob_size)

```py
>>> from native_ctypes import *
>>> x = (0,1,2)
>>> t = TupleObj.from_address(id(x))
>>> print(t)
TupleObj({'ob_refcnt': 1, 'ob_base': <class 'tuple'>, 'ob_size': 3, 'ob_item': [0, 1, 2]})
>>> t.ob_item
(py_object*3)([0, 1, 2])
>>> t.ob_item[0] = 1
>>> x
(1, 1, 2)
>>> ```

[serene stratus]

!e 3.10

x = 1 +3
+4
print(x)

[night quarry]

@serene stratus :white_check_mark: Your 3.10 eval job has completed with return code 0.

4

[grave rover]

I think we were working on it around the same time; You also made a getmem impl but that one didn't work for me anymore

[rugged sparrow]

Ah yea I never got around to making the getmem impl swappable

[grave rover]

How did you end up implementing your repr? I'm torn between yours and mine ```py

my_tuple_repr.struct_str()

struct PyTupleObject {
Py_ssize_t ob_refcnt;
struct PyTypeObject* ob_type;
Py_ssize_t ob_size;
struct PyObject*[3] ob_item;
}

my_tuple_repr.repr()

PyTupleObject(
ob_refcnt=3,
ob_type=PyTypeObject(
ob_refcnt=229,
ob_type=PyTypeObject(
ob_refcnt=87,
ob_type=...,
ob_size=0,
tp_name='type',
...
),
ob_size=0,
tp_name=*'tuple',
...
),
ob_size=3,
ob_item=[
..., # small bug in recursion detection, whoops
*PyObject(
ob_refcnt=830,
ob_type=*PyTypeObject(...),
)
PyObject(
ob_refcnt=2,
ob_type=...
)
]
)```

[soft vapor]

Hey does anyone know a good way of creating a package registry in python

(I need it for my esolang)

[rugged sparrow]

my objects have a pretty repr, my structures don't yet, I need to implement that still

[grave rover]

ended up making a config flag to shorten PyTypeObjects ```py
PyTupleObject(
ob_refcnt=3,
ob_type=*<type 'tuple'>,
ob_size=3,
ob_item=[
PyObject(ob_refcnt=1692, ob_type=<type 'int'>),
PyObject(ob_refcnt=830, ob_type=<type 'int'>),
PyObject(ob_refcnt=2, ob_type=<type 'int'>)
]
)

[quartz wave]

esoteric python is either filled with code gore or filled with people showing the progress they made on a project that has been attempted before

[grave rover]

yup

though tbf I'm mostly just rewriting what to my knowledge was the first version of this

[hasty wind]

anyone have nice license system for python scripts ?

[grave rover]

allow toggling pretty print:

[grave rover]

what do you mean? In terms of software licensing or license distribution/verification similar to Jetbrains or what

[sick hound]

there is an hidden null terminator haha

See related issue: https://github.com/python/cpython/issues/96670

[serene stratus]

cool

[grave rover]

TIL the __repr__ function on tuples supports recursion despite the fact that that shouldn't ever be a concern in normal situations

[cloud fossil]

Tuples' __str__ and __repr__ call their objects' __repr__ I think

[proper vault]

!e

class X:
  def __repr__(self):
    return f'<{self.v}>'
a = X()
u = a,
a.v = u
print(repr(u))
``` it can come into play (though if you don't use the tuple you do just get a RecursionError)

[night quarry]

@proper vault :white_check_mark: Your 3.11 eval job has completed with return code 0.

(<(...)>,)

[grave rover]

I see

[grave rover]

god I hate unicode objects

I can't manage to get the string :(

[rugged sparrow]

Repr does, hash doesn't

!e ```py
from ctypes import *

x = (0,)
py_object.from_address(id(x)+tuple.basicsize).value = x
print(x)
print(hash(x))```

[night quarry]

@rugged sparrow :x: Your 3.11 eval job has completed with return code 139 (SIGSEGV).

((...),)

[grave rover]

@rugged sparrow did you ever figure out strings? (Currently on 3.9.7 but implementing strings the way they're defined doesn't seem to work)

[rugged sparrow]

I think I did once, but it was in the repl just messing around so it's lost forever lmao. I can take a look in about 30

[grave rover]

Lol

I used to have it but my old repo is down cuz the craptop hosting it died and I have no clue how I ever did it

The biggest problem seems to be the bitfield; I don't get consistent results

I wonder if I messed up in alignment somewhere

Because that's the only thing I'm not checking for

[rugged sparrow]

yea i dont think i bothered to implement bitfields yet

[grave rover]

oh no my design doesn't have a good way to handle aligns

[rugged sparrow]

(another thing i didnt bother to add)

[grave rover]

why is the actual text this far away

[grave rover]

I think my MEMORY.seek isn't working reliably

oh wait I see

thank you contextmanagers

[rugged sparrow]

yea thats why i prefer my getmem(addr, size, fmt='c') that just returns a memoryview with the given format and size

[grave rover]

yeah it was just that I was seeking again in a function call :P

now to figure out how to turn a bytes object containing a wchar_t array into a byte array

oh wait I can just use utf-16 can't I

the only painful part now is figuring out sizeof(wchar_t) from python, ideally without use of ctypes

[rugged sparrow]

You could divide the length of the bytes by the length of the string right?

[grave rover]

the problem is that wstr_length is the only value I have as reference

[grave rover]

anyways, memhax for 3.9 is now on pypi for all to enjoy

[restive void]

no link to a repo there?

[grave rover]

https://github.com/Martmists-GH/memhax

pushed to pypi before making the repo

[serene stratus]

Is that an anime background in your IDE?

[grave rover]

yes

Slime: Rimuru theme from The Doki Theme

[serene stratus]

How is that useful

[grave rover]

idk I mostly just enjoy the color scheme it adds

[quartz wave]

16 bits, 2 bytes

if you're using only windows then that's the value

[earnest wing]

i don't think "figuring out the size" implies hard coding

[quartz wave]

well probably not

[grave rover]

On some systems wchar_t is 32 bits

That's the problem

[quartz wave]

try:
    import array
    array.array('u', ['\U00010000'])
except TypeError:
    sizeof_wchar_t = 2
else:
    sizeof_wchar_t = 4

that's one way to figure out

[obtuse elk]

Is it possible to use .out files from cpp as imports in python scripts?

[soft prairie]

Technically yes, by compiling it as a shared library (remember to use extern "C" { ... }) and renaming it to something like libname.linux-x86_64-cpython-310, then importing it with import libname should work. But you'd rather stick with setuptools for the sake of simplicity and avoiding dealing with compiler options. Here's the official python documentation for that: https://docs.python.org/3/extending/extending.html#writing-extensions-in-c

[soft prairie]

class CursedInt(int):
  def __init__(self, num: int):
    self.c = 0
    super().__init__()

  def __neg__(self):
    self.c += 1
    self.c %= 2
    if self.c:
      return super().__neg__()
    return -(self - 1)

x = CursedInt(10)
assert  -x == -10
assert --x == 9

[quartz wave]

this is one of the most basic ones 😔

[sick hound]

overengineered imo

the trick here is just to use a \0 lmfao

aka null terminator

[quartz wave]

!e ```py
from fishhook import hook, orig
from sys import _getframe
from opcode import opmap, opname

EXTENDED_ARG = opmap["EXTENDED_ARG"]

def get_arg(codestring, idx):
arg = codestring[idx + 1]
try:
if codestring[idx - 2] is EXTENDED_ARG:
arg |= codestring[idx - 1] << 8
if codestring[idx - 4] is EXTENDED_ARG:
arg |= codestring[idx - 3] << 16
if codestring[idx - 6] is EXTENDED_ARG:
arg |= codestring[idx - 5] << 24
except IndexError:
pass
return arg

@hook(int)
def neg(self, UNARY_NEGATIVE=opmap["UNARY_NEGATIVE"],
NAME_OPS={"LOAD_NAME": ("co_names", lambda x: x),
"LOAD_GLOBAL": ("co_names", lambda x: x >> 1),
"LOAD_FAST": ("co_varnames", lambda x: x),
"LOAD_CLOSURE": ("co_varnames", lambda x: x),
"LOAD_DEREF": ("co_varnames", lambda x: x),
"LOAD_CLASSDEREF": ("co_varnames", lambda x: x)}):
f = _getframe(1)
code = f.f_code
co_code = code.co_code[:f.f_lasti]
if co_code[-2] is UNARY_NEGATIVE is not (instr := co_code[-4]):
if tup := NAME_OPS.get(opname[instr]):
attr, func = tup
name = getattr(code, attr)[func(get_arg(co_code, -4))]
f.f_locals[name] = res = f.f_locals[name] - 1
return res
else:
return orig(self) - 1
return orig(self)

x = 5
print(--x, x)

[night quarry]

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

4 4

[quartz wave]

@soft prairie better
it actually changes the variable

[wheat river]

now make x-- 🔫

[quartz wave]

can't it's on the other PC

[quartz wave]

if you want to test it out, clone and build
https://github.com/thatbirdguythatuknownot/cpython/tree/modified-again

[quartz wave]

@tiny knot try now ```py
(i:=import,w:=(255,)*3,P:=i("pygame"),L:=i("pygame.locals",1,1,'a'),d:=P.draw.rect,V:=P.Vector2,R:=i("random").randint,S:=P.display.set_mode((600,)*2),C:=P.time.Clock(),sp:=V(10,10),sb:=[],di:=V(10,0),n:=lambda:globals().update(fo=V(R(0,59)*10,R(0,59)*10)),n(),[(sb:=sb[1:]+[V(sp)],sp.iadd(di),(sb:=[],sp.update(10,10))if sp in sb or not 0<=sp.x<=600>=sp.y>=0else None,(sb:=[V(-10,-10)]+sb,n())if sp==fo else None,S.fill((0,)*3),d(S,(255,0,0),(fo.x,fo.y,10,10)),d(S,w,(sp.x,sp.y,10,10)),[d(S,w,(p.x,p.y,10,10))for p in sb],[1/0if ev.type==L.QUIT else di.update({L.K_w:(0,-10),L.K_s:(0,10),L.K_a:(-10,0),L.K_d:(10,0)}.get(ev.key,di))if ev.type==L.KEYDOWN else None for ev in P.event.get()],P.display.update(),C.tick(10))for _ in iter(int,1)])

[tiny knot]

TypeError: can only concatenate list (not "pygame.math.Vector2") to list

[quartz wave]

edited

[tiny knot]

it does work.. but you can't teleport in the edges

oh

seems the original program also has the problem

[quartz wave]

ok good

because i just decreased 30 characters

[tiny knot]

this is 162 less than original

nice.

my head isn't working after just 91

[polar plover]

It's a feature 😠

[quartz wave]

here's a notable downsize py not(0<=sp.x<=600and 0<=sp.y<=600) to ```py
not 0<=sp.x<=600>=sp.y>=0

[tiny knot]

cool

[tiny knot]

that the snake break dances on the edge of the screen? 😄

[polar plover]

...... Why wouldn't you want that xD

[tiny knot]

everyone wants their snake to dance

instead of 0/1 you can use x to quit (undefined)

[soft vapor]

0
3cell_pos:1
3cell1:0
3cell2:0
3cell3:0
3cell4:0
3input:+>+>+>+<++<-
6n\12\3index:^n-1;3letter:@input.index;?letter=>|3cell_pos=^cell_pos+1|?letter=<|3cell_pos-1|?letter=+|?cell_pos=1|3cell1:^cell1+1|?cell_pos=2|3cell2:cell2+1|?cell_pos=3|3cell3:cell3+1|?cell_pos=4|3cell4:cell4+1|2error
1

added 3 instructions of bf in this interpreter >, <, +

[soft prairie]

This is crazy, I love it!

[tiny knot]

numberscript with any space/newlines given there right

[polar plover]

🤦‍♂️ ofc

[soft vapor]

This interpreter is broken

there is a bug some where

0
3cell_pos:1
3cell1:0
3cell2:0
3cell3:0
3cell4:0
3input:+>+>+>+<++<-
6n\12\3index:^n-1;3letter:@input.index;?letter=>|3cell_pos=^cell_pos+1|?letter=<|3cell_pos:^cell_pos-1|?letter=+|?cell_pos=1|3cell1:^cell1+1|?cell_pos=2|3cell2:cell2+1|?cell_pos=3|3cell3:cell3+1|?cell_pos=4|3cell4:^cell4+1|2error
1

Can anyone fix this (mess)?

Error: Error: Variable Content not defined

oh waity

I placed = then :

sily old me

wait it still isn't working

0
3cell_pos:1
3cell1:0
3cell2:0
3cell3:0
3cell4:0
3input:+>+>+>+<++<-
6n\12\3index:^n-1;3letter:@input.index;?letter=>|3cell_pos:^cell_pos+1|?letter=<|3cell_pos:^cell_pos-1|?letter=+|?cell_pos=1|3cell1:^cell1+1|?cell_pos=2|3cell2:cell2+1|?cell_pos=3|3cell3:cell3+1|?cell_pos=4|3cell4:^cell4+1|2error
2cell1
2cell2
2cell3
2cell4
1

any idea?

[wheat river]

ah ok, nice

[grave rover]

Is there a way to preserve type annotations on an object? e.g, with a simple generic class: py print(Pointer) # -> <class 'memhax2.native_complex.Pointer'> print(Pointer[Char]) # -> memhax2.native_complex.Pointer[memhax2.native.Char] print(Pointer[Char](0).__class__) # -> <class 'memhax2.native_complex.Pointer'> # <- We don't want to lose type info here

[quartz wave]

hmm yeah idk what're you asking for

do you like mean having Pointer[Char](0).__class__ return Pointer[Char]?

[grave rover]

yeah

I guess something similar to __orig_class__ which existed before 3.7 (which did have that information)

no wait it does stll exist

but not always?

[astral rover]

__orig_class__ is available on any class after init has finished if it subclasses typing.Generic or a builtin generic (class with a __class_getitem__ = classmethod(types.GenericAlias))

[grave rover]

oh so only after __init__ has finished?

[astral rover]

and it has the slot for it if required (not in the case of classes like list/tuple)

[grave rover]

that's gonna make things a little less fun but still doable

[astral rover]

yeah cause guido thinks types should be erased at runtime

https://github.com/python/cpython/issues/90899

[grave rover]

god, I'm rewriting it again inspired by chilaxan but now I've run into an issue where scoping is a pain

class PyTupleObject(PyVarObject, Struct[tuple]):
    ob_item: PropertySizeArray[Pointer[PyObject], lambda self: self.ob_size()]  # how am I gonna pass `self` [PyTupleObject] to this function

[astral rover]

whats the issue with that?

i think this is totally do-able with descriptors

[grave rover]

hmm, how would you approach this? because it's only used in PropertySizeArray

And in some cases it's even worse, here it's wrapped inside a Pointer: ```py
class PyListObject(PyVarObject, Struct[list]):
ob_item: Pointer[PropertySizeArray[Pointer[PyObject], lambda self: self.allocated()]]
allocated: Py_ssize_t

[astral rover]

    ob_item: PropertySizeArray[Pointer[PyObject]] = PropertySizeArray(callback=lambda self: self.ob_size())

[grave rover]

oh no like

passing the callback like this already work

[astral rover]

and then on access to ob_item it just calls callback with the instance argument from __get__

[grave rover]

but passing self to the array in a way that doesn't require a breaking change is gonna be tough

[astral rover]

you dont have a stability policy or are at v1 why do you care?

[grave rover]

I mean I'd like to not overcomplicate the code, which is the biggest problem

__get__ would work if it was only direct children

but the self always refers to the class level which isn't always the parent

[grave rover]

also __get__ requires being a property on the class, which doesn't work here

[grave rover]

My alignment code is wrong

                                            PyMemberDef(
                                                name=INVALID_STRING
                                                type=208
                                                offset=1
                                                flags=0
                                                doc='__base__'
                                            )
                                            PyMemberDef(
                                                name='__base__'
                                                type=6
                                                offset=256
                                                flags=1
                                                doc=NULL
                                            )
```hmm

[grave rover]

        PyObject(
            ob_refcnt=1000000009,
            ob_type=<class 'str'>
        ),
```another bug(?) found while debugging, how in the world is it getting this for refcnt? do some values just have an inherent high refcount?

this is for "a" btw

[astral rover]

cant you just check sys.getrefcount("a")?

[grave rover]

hmm, I get a slight mismatch but that's likely just the code

yeah for some reason it's just massive in the IDE and not in the REPL lol

no wait

it's on 3.11 where it's crazy high

[astral rover]

it might be immortal?

[grave rover]

¯_(ツ)_/¯

[astral rover]

check ram usage?

oh wait that wouldnt show up

[grave rover]

no it's definitely consistent behavior in the interpreter

[quartz wave]

3.11 immortal objects

at some point they decided to make the cached values have an obscenely high reference count

at the end of a program the immortal object must have a reference count of exactly 999999999

[grave rover]

I see

found the pep, but definitely isn't anywhere in the docs or in the changelog

[astral rover]

do you get a segfault otherwise?

[astral rover]

i remember seeing the change in the git commit list just after it was pushed

i can probably find it

[plush halo]

!e ```py
i = 0
while i < 10:
print(i)
i++ # this used to work in C

[night quarry]

@plush halo :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 4
002 |     i++ # this used to work in C
003 |         ^^^^^^^^^^^^^^^^^^^^^^^^
004 | SyntaxError: invalid syntax

[plush halo]

!e ```py
i = 0
while i < 10:
print(i)
++i # works in python but..

[night quarry]

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

001 | 0
002 | 0
003 | 0
004 | 0
005 | 0
006 | 0
007 | 0
008 | 0
009 | 0
010 | 0
011 | 0
... (truncated - too many lines)

Full output: too long to upload

[grave rover]

looks like it yeah

I wonder if I can wrap a memoryview in a file-like object 🤔

[astral rover]

as it turns out i cant find the commit

[grave rover]

turns out there's no proper way to do this

[grave rover]

and my crappy impl seems to also just fail and segfault for no reason

[quartz wave]

fatal error

[grave rover]

Huh, I wonder what a missing decref on my end did to cause the segfault then

[quartz wave]

wdym by "used to"

[plush halo]

it still works in C but not python... just lamenting xD

[quartz wave]

you mean "this only works in C" (works in C only) instead of "this used to work in C" (works in C some time ago, but now it doesn't)

[quartz wave]

https://github.com/python/cpython/commit/121f1f893a39d0b58d3d2b5597505c154ecaac2a

started with that

[quartz wave]

https://bugs.python.org/issue45953

[plush halo]

I'm seeing things from python's point of view... maybe it didn't carry the clear message I thought it would

[sonic ginkgo]

>>> def f():
...     def g():
...         return g
...     return g
...
>>> f()
<function f.<locals>.g at 0x105971cf0>
>>> f()()
<function f.<locals>.g at 0x105971e10>
>>> f()()()
<function f.<locals>.g at 0x105971ea0>

this is a fun one

[sick hound]

How can I do from urllib.parse import urlsplit in an import dunder

[quartz wave]

__import__("urllib.parse",1,1,'a').urlsplit

[near gust]

my proudest piece of code

circumventing a pypy bug

[grave rover]

Nice

[old socket]

Was it intentional to be able to use keywords as variable names when using ast?

I thought I was gonna get an error or something but I guess not

!e ```py
import ast

module = ast.Module(
body=[
ast.Assign(
targets=[ast.Name(id="import", ctx=ast.Store())],
value=ast.Constant(value="Foo")
)
],

type_ignores=[],

)

code = compile(
ast.fix_missing_locations(module),
"<ast>",
"exec"
)

exec(code, d:={})
print(d["import"])

[night quarry]

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

Foo

[sick hound]

Whats the integers and a for?

[quartz wave]

so it works

you can also just do __import__("urllib.parse",fromlist='a').urlsplit

[quartz wave]

they're only checked in the parser

[sick hound]

What does the fromlist do exactly?

[quartz wave]

if you try to use True/False/None(/__debug__?) it'll error

[quartz wave]

makes it work

[old socket]

Yea I was thinking that as well at first I was just confused and realised it only matter in the lexical part

[floral meteor]

Overrides the module dunder all

[magic wraith]

https://stackoverflow.com/questions/2724260/why-does-pythons-import-require-fromlist
TL DR unless you are importing a whole module (you just want urlsplit which is a function), fromlist can be anything

When importing a module from a package, note that __import__('A.B', ...) returns package A when fromlist is empty, but its submodule B when fromlist is not empty.

[lunar moon]

using fishhook, sometimes causes SIGSEGV to occur

[earnest wing]

not surprising

[fickle scarab]

is there a way to create a class with an expresion

without using SimpleNamespace

[proper vault]

@fickle scarab you can call type(or some other metaclass) manually:

class A(Base):x=4
# =
A=type('A', (Base,), {'x'=4})

[fickle scarab]

thank you! after shuffling around my words in the search engine I actually came across this exact same solution a few minutes ago. thank you very much nonetheless

(i initially came here not being able to find any results online)

[fickle scarab]

is there a way to reassign properties as an expression? seems walrus doesn't allow it for some reason

[proper vault]

the setattr builtin

[fickle scarab]

ah thanks

[maiden blaze]

whats the reasoning for walrus not allowing it

like sure setattr works but i dont see that being the reason for going out of the way to forbid it for walrus

[potent comet]

The reason is to push you to use regular assignments, not bloat the scope of walrus to other tasks. It's easy to remove a restriction later, much more difficult to deprecate functionality if it is found to be confusing.

[proper vault]

another possibility is that having the lhs of the assignment create an exception somehow makes the impl harder

[quartz wave]

it was already too controversial using just names

[quartz wave]

it's quite easy to change the grammar for that but there'll be some problems with the symtable when checking for comprehension walruses i think

[proper vault]

Ye, grammar seems fine

My worry is None.attr := value, feels like that could be an annoying thing to compile correctly.

[rugged sparrow]

I mean it would be the same as attr := value except with an extra LOAD_* at the beginning and it would use STORE_ATTRIBUTE instead of STORE_FAST or STORE_GLOBAL

!e ```py
import dis
dis.dis('(a := 1)')
dis.dis('a.attr = 1')

[night quarry]

@rugged sparrow :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 |   0           0 RESUME                   0
002 | 
003 |   1           2 LOAD_CONST               0 (1)
004 |               4 COPY                     1
005 |               6 STORE_NAME               0 (a)
006 |               8 RETURN_VALUE
007 |   0           0 RESUME                   0
008 | 
009 |   1           2 LOAD_CONST               0 (1)
010 |               4 LOAD_NAME                0 (a)
011 |               6 STORE_ATTR               1 (attr)
... (truncated - too many lines)

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

[rugged sparrow]

so the resulting bytecode would need to be this: ```py
001 | 0 0 RESUME 0
002 |
003 | 1 2 LOAD_CONST 0 (1)
004 | 4 COPY
005 | 6 LOAD_GLOBAL 0 (obj)
006 | 8 STORE_ATTR 1 (a)
007 | 10 RETURN_VALUE

>>> x = lambda v:(v or a.a)
>>> dis.dis(x)
  1           0 LOAD_FAST                0 (v)
              2 JUMP_IF_TRUE_OR_POP      4 (to 8)
              4 LOAD_GLOBAL              0 (a)
              6 LOAD_ATTR                0 (a)
        >>    8 RETURN_VALUE
>>> co_code = bytearray(x.__code__.co_code)
>>> co_code[2] = dis.opmap['DUP_TOP']
>>> co_code[6] = dis.opmap['STORE_ATTR']
>>> x.__code__ = x.__code__.replace(co_code=bytes(co_code))
>>> class a:pass
... 
>>> x(5)
5
>>> a.a
5
>>> 
``` ^3.10 example

[quartz wave]

this would be a little annoying to check against iterator assignments in a comprehension

[potent comet]

I think also the issue comes from the fact that the target object would be able to override behaviour, so if __getitem__ and __setitem__ didn't match up you'd get inconsistent results.

[rugged sparrow]

personally i think that if it was implemented to allow (obj.attr := value) it would be the same as obj.attr = value; return value so __getitem__ wouldnt matter

[fickle scarab]

that seems the only logical way to set it up

[fickle scarab]

if 'm reading this bytecode right, you just dup the rvalue before assignment and return it

which seems to be a not-so-difficult way of doing it in bytecode

(i come from java bytecode so maybe it's different in python)

[fickle scarab]

that's how java does it

return someField = someOtherField;

compiles to

getstatic someOtherfield
dup
putstatic someField
ireturn

[finite blaze]

Quick golfing challenge

[wheat river]

list.count(max(list))

[finite blaze]

Yeah i just realized how boring it is

https://c.tenor.com/pagVxAkHfWAAAAAM/my-job-here-is-done-bye.gif

[fickle scarab]

i was so confused by the "word problem" aspect of it for some reason until i looked at the code and realized that it's just "find the mode of the list"

[versed eagle]

# alternative to ternary
((condition)and(_:=expressionIfTrue,)or(_:=expressionIfFalse,),_)[-1]

it's like a ternary but more characters to get the same result
oh and it also assigns the result to the local namespace so that's a thing

[fickle scarab]

Is there a delete-as-expression?

[opal breach]

I don't think so, other than del keyword you have the garbage collector module though, perhaps you can manually GC a variable, __del__ and __delete__ won't do anything either since those are called after the fact, iirc.

[quartz wave]

((condition)and(if_true,)or(if_false,))[0]

[versed eagle]

oh well
i tried

but anyway, i like that way of doing things over python's ternary because it's closer to C/C++, with the condition first and then the results

[fleet bridge]

del a.b
(delattr(a, 'b'))

del a[b]
(a.__delitem__(b))

del a
(a := None) or (a := MISSING)
# what shoult expression return in this case?

i think you can use ctypes to break frame object and assign NULL to your variable
in this case you will get UnboundLocalError, just like after del a

if you want to delete global, you can globals().__delitem__('b')

>>> globals()
{   '__name__': '__main__',
    '__doc__': None,
    '__package__': None,
    '__loader__': <class '_frozen_importlib.BuiltinImporter'>,
    '__spec__': None,
    '__annotations__': {},
    '__builtins__': <module 'builtins' (built-in)>}
>>> globals().__delitem__('__name__')
>>> __name__
'builtins'
``` bruh

>>> globals().clear()
>>> import __main__
>>> __main__
<module '?'>

[twilit mountain]

Would there be a way to do a++ in Python? (a is an int variable)

[fickle scarab]

__import__("ctypes").c_void_p.from_address(id(a)+8)).value=id(a+1)```

(this doesn't actually work sadly)

even though it SHOULD

[restive void]

!e very brittle, breaks if number is too small or too large, but:

import ctypes
a = 1234
ctypes.c_int.from_address(id(a)+24).value += 1
print(a)

[night quarry]

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

1235

[restive void]

and if you mean the syntax, then no, not without the usual syntax hacks (codecs or building your own CPython variant with different grammar)

[fierce echo]

hi mates!

I would have a question about Python, is it allowed to send screenshot here?

[maiden blaze]

see #❓｜how-to-get-help for general python questions

[fierce echo]

@maiden blaze ok thanks

[serene stratus]

As far as I understand you're just setting the value to an id? Don't you either have to remove the + 8 offset or just set a+1 without the id ()?

[vital dirge]

Heh, "Comprehension Walrus" sounds like it should be one of those philosophical animal memes

[quartz wave]

unfortunately that's not how objects work

fortunately i have done this before

unfortunately i can't find the code

[quartz wave]

this can just be used anyway

[royal whale]

Yes

Totally

[vital dirge]

🙂

[unreal echo]

It has been a long time since I've come here

[grave rover]

If sizeof void* is 8 this increases ob_refcnt, if sizeof void* is 4 this increases ob_size

Wait no that's incorrect

You get the gist of it at least

[fickle scarab]

i actually don't really understand that snippet to be honest, but i messed around with it to get it to work

[quartz wave]

it's setting the type of a to a+1

so that should never work

[soft vapor]

well then, welcome back to the work of weirdness and strangeness good

[fickle scarab]

I mean I don't understand how it works. And also, why can't we set a to a+1 if we can set True to an arbitrary class?

[quartz wave]

you're not setting a to a+1
you're setting the type of a to a+1

of course you can set a to a+1

[fickle scarab]

Ah

That makes the snippet whence it came make more sense as well

[quartz wave]

__import__("ctypes").c_void_p.from_address(id(a)).value=id(a+1)
when you instead do this you set the reference count of a to id(a+1)

[fickle scarab]

I see

Ah it all makes sense now

[versed eagle]

# biggest num it works for before overflow
# 0b111111_11111111_11111111_11111111    (1073741823)
# it probably overflows when decrementing on -(thatnumber+1), but I haven't tested that
# works for negative numbers
# breaks on: -1, 0, 1
n=some_integer 
from ctypes import c_longlong as c_ll
c_ll.from_address(id(n)+24).value=c_ll.from_address(id(n+1)+24).value

[quartz wave]

!e ```py
from ctypes import c_uint
for n in range(-1, 2):
print(n, end=" -> ")
c_uint.from_address(id(n)+24).value=n+1&1073741823
print(n)

[night quarry]

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

001 | -1 -> -0
002 | 0 -> 0
003 | 1 -> 2

[quartz wave]

@versed eagle seems to work as intended

you get nothing for 0 because it has a size of 0

[versed eagle]

huh

with -1 it breaks for me

*nvm, im wrong

[quartz wave]

this presents you with the value of -2 when you do it on -1

[versed eagle]

!e

import ctypes
n=1
for i in range(10):
    ctypes.c_longlong.from_address(id(n)+24).value = ctypes.c_longlong.from_address(id(n+1)+24).value
    print(n)

[night quarry]

@versed eagle :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 2
002 | 4
003 | 8
004 | 16
005 | 32
006 | 64
007 | 128
008 | 256
009 | 512
010 | 1024

[versed eagle]

breaks on 1

[quartz wave]

you add 1 with itself

which is 2

then you add 1 (2) with itself

which is 4

and so on

[versed eagle]

why isn't that behavior replicated with 2 then

!e

import ctypes
n=2
for i in range(10):
    ctypes.c_longlong.from_address(id(n)+24).value = ctypes.c_longlong.from_address(id(n+1)+24).value
    print(n)

[night quarry]

@versed eagle :x: Your 3.11 eval job has completed with return code 139 (SIGSEGV).

001 | 3
002 | 4
003 | 5
004 | 6
005 | 7
006 | 8
007 | 9
008 | 10
009 | 11
010 | 12

[versed eagle]

it just counts

[quartz wave]

because you don't add 2 with itself

[versed eagle]

ah
ok

[quartz wave]

ctypes.c_longlong.from_address(id(n+1)+24).value
                                   ^^
``` what do you think this is

[quartz wave]

1 is cached so it's effectively adding itself

[versed eagle]

oh right
python caches 1-256

sorry! I'm dumb

[quartz wave]

it caches [-5, 256]

that's a range not a list

[versed eagle]

I only remember the positive numbers cause I don't use negative numbers lol

[versed eagle]

ik

[versed eagle]

is it possible to free the cache?

[quartz wave]

yes but that breaks everything

[versed eagle]

how so?

would it not just dynamically allocate the ints like it does for numbers bigger than 256

[rugged sparrow]

no, there are hardcoded code paths to grab from the cache

[versed eagle]

huh

[versed eagle]

why delete it

i didnt have time to read it

[rugged sparrow]

it was wrong

[versed eagle]

oh

[rugged sparrow]

!e maybe will work ```py
def new_int(n):
sz = (n.bit_length() + 7) // 8
bts = n.to_bytes(sz, 'little', signed=n<0)
return int.from_bytes(bts, 'little', signed=n<0)

print(id(1), id(new_int(1)))

[night quarry]

@rugged sparrow :white_check_mark: Your 3.10 eval job has completed with return code 0.

139897336873200 139897335828752

[rugged sparrow]

eh it works in 3.10

[magic venture]

hi all, quite a specific question:
has anyone used the PyMultiNest package? I am trying to scan over a large combination of parameter values, although there is a particular parameter which I want to force to be integers only while scanning. does anyone know how to do this?

[restive void]

breaks if number is too small or too large

[quartz wave]

the cache is full of static structs as of 3.11

[versed eagle]

uh
check if type(param) == int, and if not, raise an error?

something like

if type(param) != int: 
    raise TypeError(f"param_name should be an int, but {param} is of type {type(param)}")
# then do stuff here
do_stuff()```

[quartz wave]

use is/is not

[restive void]

use isinstance

(most of the time)

[versed eagle]

my logic for not using is/is not is that with my way, users can inherit from int, and then use a metaclass to overload eq and return true for int

I forgot about isinstance though :(

[keen swan]

How would you go about destructuring a python dictionary?

for example, I have a person dictionary, with name and age

!e

person = {
  "name" : "Bob",
  "age" : 33
}

for values in person.values():
  print(values)

I know this works

[night quarry]

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

001 | Bob
002 | 33

[keen swan]

But in Javascript there's a way to destructure an object

with the spread operator

name, age = ...person

Sort of like that

What's the equivalent in Python?

[rugged owl]

You can do this

from operator import itemgetter
x = {'a': 1, 'b': 2}

a, b = itemgetter('a', 'b')(x)

[keen swan]

hmm have to import itemgetter

Nothing built in?

oh wait

what about **kwargs?

doesn't seem right

So I know *args can spread a list

!e

numbers = [1,2,3,4,5]
print(*numbers)

[night quarry]

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

1 2 3 4 5

[keen swan]

And doing the same will only spread the keys on a dictionary

but won't get the values

[cloud fossil]

!e

x = {"a": 1, "b": 2}
keys = *x
print(keys)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 2
002 | SyntaxError: can't use starred expression here

[cloud fossil]

Oops

!e

x = {"a": 1, "b": 2}
a, b = x.values()
print(a)
print(b)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 1
002 | 2

[cloud fossil]

!e

x = {"a": 1, "b": 2}
a, b = x
print(a)
print(b)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | a
002 | b

[royal whale]

!e

a = {"a":1, "b":2, "c":3}
b, c, d = a.items()
print(b, c, d)

[night quarry]

@royal whale :white_check_mark: Your 3.11 eval job has completed with return code 0.

('a', 1) ('b', 2) ('c', 3)

[keen swan]

hmm

[cloud fossil]

In Python this is called unpacking

[keen swan]

How would I use it in a function?

[cloud fossil]

The process of assigning names to values in a sequence

[keen swan]

Gimme a min

[cloud fossil]

What do you mean by that

You can use it anywhere, it's syntax

[keen swan]

!e

person = {
  "name": "Bob",
  "gender": "Male"
}

def intro( name, age ):
  print(f"Hi {name}. You are a {gender}")

intro(person.values())

Testing

[night quarry]

@keen swan :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 9, in <module>
003 | TypeError: intro() missing 1 required positional argument: 'age'

[keen swan]

hmm

oh wait

sorry I had it as age instead of gender

in func

!e

person = {
  "name": "Bob",
  "gender": "Male"
}

def intro( name, gender ):
  print(f"Hi {name}. You are a {gender}")

intro(*person)

Would this work?

[night quarry]

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

Hi name. You are a gender

[keen swan]

oh I think I'm getting close

[cloud fossil]

*person.values()

Or use the asterisk in the parameter

[keen swan]

!e

person = {
  "name": "Bob",
  "gender": "Male"
}

def intro( name, gender ):
  print(f"Hi {name}. You are a {gender}")

intro(*person.values())

[night quarry]

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

Hi Bob. You are a Male

[keen swan]

Ah yes

much bettter

[cloud fossil]

Which would still require using .values()

[keen swan]

wdym?

ye

[cloud fossil]

!e

person = {
    "name": "Bob",
    "gender": "Male"
}

def intro(*args):
    name, gender = args
    print(f"Hi {name}. You are a {gender}")

intro(*person.values())

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hi Bob. You are a Male

[cloud fossil]

Or

!e

person = {
    "name": "Bob",
    "gender": "Male"
}

def intro(person_values):
    name, gender = person_values
    print(f"Hi {name}. You are a {gender}")

intro(person.values())

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hi Bob. You are a Male

[keen swan]

huh, this function is a bit more readable that *args

Or is args just another word?

!e

person = {
    "name": "Bob",
    "gender": "Male"
}

def intro(*person_values):
    name, gender = person_values
    print(f"Hi {name}. You are a {gender}")

intro(*person.values())

[night quarry]

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

Hi Bob. You are a Male

[cloud fossil]

It's a name

[keen swan]

Yeah its just a name

hmm

wait, whats the diff with or without *?

[cloud fossil]

!e

def add_all(*numbers):
    return sum(numbers)

print(add_all())
print(add_all(1, 2, 3, 4))
print(add_all(1))

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 0
002 | 10
003 | 1

[cloud fossil]

!e

def f(*args, **kwargs):
    print(args, kwargs)

f(1, 2, argument=3)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

(1, 2) {'argument': 3}

[cloud fossil]

!e

def f(a, *b, c):
    print(a, b, c)

f(1, 2, 3)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 4, in <module>
003 | TypeError: f() missing 1 required keyword-only argument: 'c'

[cloud fossil]

I guess it makes the rest of the arguments keyword-only

[keen swan]

hmm

[cloud fossil]

!e

def f(a, *b, c):
    print(a, b, c)

f(1, 2, c=3)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

1 (2,) 3

[keen swan]

I think this helped

[cloud fossil]

!e

def f(a, *b, c):
    print(a, b, c)

f(1, 2, 3, 4, c=5)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

1 (2, 3, 4) 5

[cloud fossil]

Ok

[keen swan]

oh

that example makes this unpredictable

It returned a tuple

Sorry I mean unexpected

[cloud fossil]

Unrelatedly, because we are on the topic of function signatures I want to note that specifically in function signatures, / separates positional-only arguments (those are on the left side of it) and * separates keyword-only arguments (those are on the right side of it)

[cloud fossil]

This time it's called packing

[keen swan]

oh ok

[cloud fossil]

In function signatures, a parameter prefixed with * is just like when you do

a, *b = "hello"  # a = "h"; b = "ello"

[keen swan]

right

a fancier way of slicing

[cloud fossil]

It's called packing

Slicing is when you use slice or [start:stop:step] syntax

[keen swan]

right

so it packed "ello" because positionally, its after the a variable

[cloud fossil]

!e

def f(a=1, /, b):
    print(a, b)

f(b=7)
f(10, b=5)
f(a=4, b=6)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     def f(a=1, /, b):
003 |                   ^
004 | SyntaxError: non-default argument follows default argument

[cloud fossil]

!e

def f(a=1, /, b, *):
    print(a, b)

f(b=7)
f(10, b=5)
f(a=4, b=6)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     def f(a=1, /, b, *):
003 |                   ^
004 | SyntaxError: non-default argument follows default argument

[cloud fossil]

Huh

[keen swan]

first time seeing a / in a function arg

[cloud fossil]

I explained it here

[keen swan]

yeah I know

Just never seen it before

[cloud fossil]

!e

def f(a, /):
    print(a)

f(10)
f(a=10)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 | 10
002 | Traceback (most recent call last):
003 |   File "<string>", line 5, in <module>
004 | TypeError: f() got some positional-only arguments passed as keyword arguments: 'a'

[cloud fossil]

!e

def f(*, a):
    print(a)

f(a=10)
f(10)

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 | 10
002 | Traceback (most recent call last):
003 |   File "<string>", line 5, in <module>
004 | TypeError: f() takes 0 positional arguments but 1 was given

[cloud fossil]

!e

def f(*, a):
    print(a)

f()

[night quarry]

@cloud fossil :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 4, in <module>
003 | TypeError: f() missing 1 required keyword-only argument: 'a'

[cloud fossil]

Arguments after * are passed as keyword-only

Arguments before / are passed as positional-only

[keen swan]

I'm trying to find in the doc for that / arg

I don't know what its called

[cloud fossil]

!pos

[night quarry]

Positional vs. Keyword arguments

Functions can take two different kinds of arguments. A positional argument is just the object itself. A keyword argument is a name assigned to an object.

Example

>>> print('Hello', 'world!', sep=', ')
Hello, world!

The first two strings 'Hello' and world!' are positional arguments.
The sep=', ' is a keyword argument.

Note
A keyword argument can be passed positionally in some cases.

def sum(a, b=1):
    return a + b

sum(1, b=5)
sum(1, 5) # same as above

Somtimes this is forced, in the case of the pow() function.

The reverse is also true:

>>> def foo(a, b):
...     print(a, b)
...
>>> foo(a=1, b=2)
1 2
>>> foo(b=1, a=2)
2 1

More info
• Keyword only arguments
• Positional only arguments
• !tags param-arg (Parameters vs. Arguments)

[cloud fossil]

See More info

Positional only arguments, to be exact

[keen swan]

I see it thanks

Interesting

[cloud fossil]

With assignment it's different

Remember how I couldn't do def f(a, *b, c):?

!e

a, *b, c = "hello"
print(a, b, c)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

h ['e', 'l', 'l'] o

[keen swan]

yup

[cloud fossil]

* and ** are called star expressions I think

[keen swan]

they are

!e
nums = [1,2]
*nums

[night quarry]

@keen swan :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 2
002 | SyntaxError: can't use starred expression here

[cloud fossil]

!e

print({**{"a": 1}})

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

{'a': 1}

[keen swan]

wouldn't that just be exactly calling the dictionary?

Without **

[cloud fossil]

And without a pair of braces

Because what I did is unpack it into a dictionary

[keen swan]

oh yea

alright I think I can do the rest of my code

Thanks @cloud fossil

[cloud fossil]

Once upon a time you would merge dictionaries with this

Now the pipe operator exists

[keen swan]

ok I'm curious

I thought | is just OR bitwise operator

[cloud fossil]

!e

a = {"a": 1}
b = {"b": 2}
print({**a, **b})
print(a | b)

[night quarry]

@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | {'a': 1, 'b': 2}
002 | {'a': 1, 'b': 2}

[cloud fossil]

Even |= is defined

[keen swan]

huh

[cloud fossil]

It's like + and += for lists

And it's defined for sets too

[keen swan]

aren't sets immutable?

[cloud fossil]

frozensets are immutable

And also hashable

sets are neither of these

[keen swan]

hmm

probably for another time

I just need to get around to packing and unpacking dictionaries

[cloud fossil]

Also, | replaces typing.Union on newer versions of python

By "replaces" I mean exists there to use (instead of having to import and maybe for readability too)

[keen swan]

I see

[fleet bridge]

!e

person = {
  "name" : "Bob",
  "age" : 33
}
name, age = person.values()
print(name, age)

[night quarry]

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

Bob 33

[keen swan]

Thanks denball. I've found that I can use that or *args in a function that I needed

[rugged owl]

Only problem:

!e

person = {
  "name" : "Bob",
  "age" : 33,
  "test": 2
}
name, age = person.values()
print(name, age)

[night quarry]

@rugged owl :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 6, in <module>
003 | ValueError: too many values to unpack (expected 2)

[fleet bridge]

it is not a problem

[rugged owl]

Well yeah, but 🤷‍♂️

[restive void]

!e

person = {
    "name": "Bob",
    "age": 33,
    "test": 2,
}
name, age, *_ = person.values()
print(name, age)

[night quarry]

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

Bob 33

[undone helm]

print('\n'.join([''.join([('Engineer'[(x-y)%8 ]if((x*0.05)**2+(y*0.1)**2-1)**3-(x*0.05)**2*(y*0.1)**3<=0 else' ')for x in range(-30,30)])for y in range(15,-15,-1)]))

can anyone explain me the logic behind this code

like what is exactly happenening

[subtle solar]

Seems like its going through a 2d grid of x and y coordinates as it prints, checking whether the current position being checked lies inside of that equation for a heart shape, and if so, prints out a character from the string "Engineer"

[undone helm]

ok hows it chcking tht

[subtle solar]

here's it re-written in a way using regular for loops: ```py
for y in range(15, -15, -1):
for x in range(-30, 30):
if (((x * 0.05) ** 2 + (y * 0.1) ** 2 - 1) ** 3 - (x * 0.05) ** 2 * (y * 0.1) ** 3) <= 0:
print("Engineer"[(x - y) % 8], end="")
else:
print(" ", end="")
print()

[undone helm]

ohk

[undone helm]

thank you very much it will be easier to understand now

[earnest wing]

Smells suspicious

My guess would be it obfuscates a slur

[ruby orbit]

haha it does not

bjokey was right

the equation simplifies to https://www.desmos.com/calculator/pymohatezu and thats also what it prints

def equation_viewer(yes, equation, width, height, no=" "):
    for y in range(height, -height, -1):
        for x in range(-width, width + 1):
            choice = yes if equation(x, y) else no
            print(choice[(x - y) % len(choice)], end="")
        print()

def heart(x, y):
    return ((0.0025*x**2+0.01*y**2-1)**3-0.0000025*x**2*y**3) <= 0

equation_viewer("█#", heart, width=23, height=12, no="_")

[floral meteor]

Wow, I don't need a graphing calculator anymore

Which is faster,
(...)**2
or
(lambda k:k*k)(...)
for any given expression ...?

[quartz wave]

(a:=...)*a?

[floral meteor]

Variable creation is faster than anonymous function?

[quartz wave]

anonymous function still requires function making and calling

[floral meteor]

Huh nice

Which is faster for complex math, using built-in types or making a mimic with custom class and integers

Or functions of pairs of numbers?

[quartz wave]

oops

[quartz wave]

print('\n'.join(''.join(' 'if(.25*x*x+.01*y*y-1)**3-2.5*1e-6*x*x*y**3>0else'Engineer'[x-y&7]for x in range(-30,30)])for y in range(15,-15,-1)))

[floral meteor]

I was almost expecting 1e-6*x*x*y*y*y

[undone helm]

👀

[ruby orbit]

print('\n'.join(''.join('+'[(x-y)%1]if(abs(x) ** abs(y) < 10)else' 'for x in range(-30, 30))for y in range(15, -15, -1)))

def eso_graph_generator(yes, equation_as_str, width, height, no):
    return f"print('\\n'.join(''.join('{yes}'[(x-y)%{len(yes)}]if({equation_as_str})else'{no}'for x in range(-{width}, {width}))for y in range({height}, -{height}, -1)))

[quartz wave]

raise IndexError("string index out of range")
``` golfed to the minimum

[ruby orbit]

!e

print('\n'.join(''.join('+'[(x-y)%1]if(abs(x) ** abs(y) < 10)else' 'for x in range(-30, 30))for y in range(15, -15, -1)))

[night quarry]

@ruby orbit :white_check_mark: Your 3.11 eval job has completed with return code 0.

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

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

[quartz wave]

nvm then

[ruby orbit]

!e

print('\n'.join(''.join('+-'[(x-y)%2]if(abs(x) ** abs(y) < 10)else' 'for x in range(-30, 30))for y in range(5, -5, -1)))

[night quarry]

@ruby orbit :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 |                              +-+                            
002 |                              -+-                            
003 |                             -+-+-                           
004 |                            -+-+-+-                          
005 |                      +-+-+-+-+-+-+-+-+-+                    
006 | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
007 |                      +-+-+-+-+-+-+-+-+-+                    
008 |                            -+-+-+-                          
009 |                             -+-+-                           
010 |                              -+-                            

[royal brook]

https://www.reddit.com/r/Python/comments/xg56eq/prefix_and_postfix_operators_for_python/

have a look lol

[ruby orbit]

hahaha thats hilarious

[royal whale]

Nice
Did you make that?

[royal brook]

yea sorry for the nick lmao

[fleet bridge]

lambda call - 70 ns
assignment - 7 ns
float ** 2 - 50 ns
float * float - 25 ns```

`(lambda k:k*k)(...)`
creates lambda (100 ns) + call it (70 ns) + multiply vars (25 ns) + several assignments and lookups = 200+ ns

`f = lambda k:k*k; f(...)` (create lambda once and call it several times)
lookup (7 ns) + call (70 ns) = 77 ns

`(...)**2`
square float (50 ns) = 50 ns

`(a:=...)*a`
assignment (7 ns) + lookup (7 ns) + multiplication (25 ns) = 40 ns

all calculations are approximate

i have huge benchmark of different operations, i can show it if you want

[quartz wave]

!e ```py
from ctypes import pythonapi as p, py_object
p.PyUnicode_FromObject.argtypes = [py_object]
p.PyUnicode_FromObject.restype = py_object
p.PyUnicode_FromObject(py_object(5))

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "<string>", line 4, in <module>
003 | TypeError: Can't convert 'int' object to str implicitly

[restive void]

Is it possible to catch an exception, and then get access to the frame where it occurred, and continue there (possibly with some bytecode offset)?

Just had a stupid idea, but fiddling with pdb on my phone turns out to be difficult

[serene stratus]

should be possible

[proper vault]

seems fairly straightforward via sys.settrace

just wait for an exception event and create a jump to... IG the next line

[restive void]

No, I want to be able to catch an exception somewhere up higher in the callstack, and then optionally jump back to where / after where the exception happened

I guess I could keep a reference to the frame where the exception got raised with a trace function, and then somehow use that from outside

[proper vault]

ah, that's certainly an idea

still doable

[restive void]

Ideal syntax:

def do_thing():
    wat = raise ValueError(42)
    print("wat:", wat)

do_thing()  # raises ValueError
# prints "wat: lol"
try:
    do_thing()
except ValueError:
    continue with "lol"

But I'll settle for something without special syntax for now :)

[earnest wing]

algebraic effects

[fleet bridge]

You just invented javascript
Whatever happens, you will get no errors and execution will continue

[restive void]

You mean PHP?

[fleet bridge]

I think it is doable with generators
(Idk how to use .send)

def do_thing():
    wat = yield 42
    print("wat:", wat)

do_thing().send('lol')

[restive void]

Right, but that means every function in-between has to "cooperate"

The idea of algebraic effects is that the functions in-between the handler and the "causer", so to speak, don't have to be aware of the effect.

[fleet bridge]

You can implement your custom global context manager

class CManager:
    def __init__(self, handler: Callable): ...

def get_value(): ... # some code here using Cmanager

# usage:

def do_thing():
    wat = get_value(42)
    print(wat)

def foo(f):
    return f()

with CManager(handler=lambda x: 'lol'):
    foo(do_thing)
    

you can track all active context managers and use innermost in get_value

[restive void]

I guess that would work, yes... I could add a bit more sugar and it would almost work

[fleet bridge]

!e```py
from future import annotations
from types import TracebackType
from typing import Any, Callable, ClassVar

class CM:
_s: ClassVar[list[CM]] = []

def __init__(self, handler: Callable[[Any], Any]):
    self.handler = handler

def __enter__(self) -> CM:
    self.__class__._s.append(self)
    return self

def __exit__(
    self,
    exc_type: type[Exception],
    exc_value: type[Exception] | Exception,
    traceback: TracebackType,
) -> bool:
    assert self.__class__._s.pop() is self
    return True

def handle(self, x: Any) -> Any:
    return self.handler(x)

def handle(value: Any) -> Any:
return CM._s[-1].handle(value)

def do_thing() -> None:
print(f'{handle(42) = }') # 42 42\nhandle(42) = 'wat'

with CM(handler=lambda x: (print(x), 'lol')[-1]):
    print(f'{handle(43) = }') # 43\nhandle(43) = 'lol'

print(f'{handle(44) = }') # 44 44\nhandle(44) = 'wat'

def foo(f: Any) -> Any:
return f()

with CM(handler=lambda x: (print(x, x), 'wat')[-1]):
foo(do_thing)

[night quarry]

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

001 | 42 42
002 | handle(42) = 'wat'
003 | 43
004 | handle(43) = 'lol'
005 | 44 44
006 | handle(44) = 'wat'

[next flame]

hey js isn't that bad

isn't that a vb feature

[fleet bridge]

Imo, js is very bad

[rain mountain]

how to obfuscate a python script

wdym

[night quarry]

@mellow salmon :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, world!

[rain mountain]

i dont understand anything

[night quarry]

@mellow salmon :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, world

[rain mountain]

uh

yea

ok

[night quarry]

@mellow salmon :white_check_mark: Your 3.11 eval job has completed with return code 0.

1st

@mellow salmon :white_check_mark: Your 3.11 eval job has completed with return code 0.

2

[rain mountain]

whats that

[void terrace]

So i was trying to figure out the shortest code to find the sum of all even numbers from 0 to N, and we are wondering if there is a code shorter than 32 symbols:

n=int(input())//2
print(n*(n+1))

Is there a better code alternative?

Thanks in advnace

[earnest wing]

n*-~n

or ~-n, I forget which way

[void terrace]

Wait, what does the tilda do?

[earnest wing]

Two's complement

~x is -x-1

[void terrace]

Oh wow, ok

@solemn delta

[solemn delta]

Wow

Thanks

[earnest wing]

sum(range( is unoptimal then?

[solemn delta]

Its larger i think

We tried

[rugged owl]

How do you come up with this weird logic

[tribal wolf]

how do i do for i in range(x): variable = variable(string)

also how do i use python to interact with a website from my computer

also how do i make a kahoot bot spammer

[royal whale]

None of this goes in #esoteric-python except maybe the first one.

[languid night]

!rule 5

[night quarry]

Rules

5. Do not provide or request help on projects that may break laws, breach terms of services, or are malicious or inappropriate.

[restive void]

!e My solution looks similar, but I think my solution should work with threads/async/etc.:

import sys

class Effect:
    def __enter__(self):
        self.local_handlers = sys._getframe().f_back.f_locals.setdefault("_effect_handlers", [])
        self.local_handlers.append(self)

    def __exit__(self, *args):
        self.local_handlers.pop()

    def __call__(self, key, *args, **kwargs):
        if hasattr(self, f"handle_{key}"):
            return getattr(self, f"handle_{key}")(*args, **kwargs)
        elif hasattr(self, "handle"):
            return getattr(self, "handle")(*args, **kwargs)
        return NotImplemented


class Printer(Effect):
    def __init__(self, prefix):
        self.prefix = prefix

    def handle_print(self, val):
        print(self.prefix, val)


def perform(value, *args, **kwargs):
    frame = sys._getframe()
    handled = False
    while frame := frame.f_back:
        if "_effect_handlers" in frame.f_locals:
            for handler in reversed(frame.f_locals["_effect_handlers"]):
                if (retval := handler(value, *args, **kwargs)) is not NotImplemented:
                    return retval
    raise RuntimeError("unhandled effect")


def something():
    perform("print", "hello")

with Printer("outer"):
    something()
    with Printer("inner"):
        something()
    something()

[night quarry]

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

001 | outer hello
002 | inner hello
003 | outer hello

[fleet bridge]

It is a lot better

[tribal wolf]

how do i do for i in range(x): variable = variable(string). i tried fill = builtins.dict['color' + str(x)] but i bet you can guess how that went lol

[tribal wolf]

i have defined variables like var1 var2 etc. i want to cycle through them by going stuff = var[1, 2, etc]. how do i do this.

[rugged sparrow]

!e You shouldn't? You should use a dictionary. But since this is #esoteric-python ```py
for i in range(3):
globals()[f'var{i}'] = i

print(var0, var1, var2)

[night quarry]

@rugged sparrow :white_check_mark: Your 3.11 eval job has completed with return code 0.

0 1 2

[burnt pasture]

seriously, don't do that. Use a dict.

or a list.

[tribal wolf]

i know how to do that. thats why i want to know how to do it the other way.

[burnt pasture]

then i guess use chilaxan's code

[floral meteor]

!e ```py
var0,var1,var2,var3,var4 = range(5,10)
print(*[globals()[f'var{i}']for i in range(5)])

[night quarry]

@floral meteor :white_check_mark: Your 3.11 eval job has completed with return code 0.

5 6 7 8 9

[sick hound]

after creating dictionary, you can do something like this

d = {'a': 1, 'b': 2}

locals().update(d)

print(a)
print(b)

[rapid sparrow]

oops i didn't see this... you still want me to review?

[rapid sparrow]

nice, dict.update is my favorite way of updating locals

[quartz wave]

didn't it already get merged?

how about view pull #16

[granite bloom]

!e

(lambda: [
    ['def'
        for sys in [__import__('sys')]
        for math in [__import__('math')]

        for sub in [lambda *vals: None]
        for fun in [lambda *vals: vals[-1]]

        for echo in [lambda *vals: sub(
            sys.stdout.write(u" ".join(map(unicode, vals)) + u"\n"))]

        for Cylinder in [type('Cylinder', (object,), dict(
            __init__ = lambda self, radius, height: sub(
                setattr(self, 'radius', radius),
                setattr(self, 'height', height)),

            volume = property(lambda self: fun(
                ['def' for top_area in [math.pi * self.radius ** 2]],

                self.height * top_area))))]

        for main in [lambda: sub(
            ['loop' for factor in [1, 2, 3] if sub(
                ['def'
                    for my_radius, my_height in [[10 * factor, 20 * factor]]
                    for my_cylinder in [Cylinder(my_radius, my_height)]],

                echo(u"A cylinder with a radius of %.1fcm and a height "
                     u"of %.1fcm has a volume of %.1fcm³."
                     % (my_radius, my_height, my_cylinder.volume)))])]],

    main()])()

[frigid wharf]

!e ```py
import sys, ctypes

def swap(a,b):
f=sys._getframe(1)
c=f.f_code
g=lambda n:(c.co_names,c.co_varnames)[c.co_code[f.f_lasti-n-1]==124][c.co_code[f.f_lasti-n]]
save=lambda:ctypes.pythonapi.PyFrame_LocalsToFast(ctypes.py_object(f),ctypes.c_int(0))
t=(f.f_locals,f.f_globals)[c.co_code[f.f_lasti-4]==116][g(3)]
(f.f_locals,f.f_globals)[c.co_code[f.f_lasti-4]==116][g(3)]=(f.f_locals,f.f_globals)[c.co_code[f.f_lasti-2]==116][g(1)]
save()
(f.f_locals,f.f_globals)[c.co_code[f.f_lasti-2]==116][g(1)]=t
save()

x,y=10,5
swap(x,y)
print(x,y)```

[night quarry]

@frigid wharf :white_check_mark: Your 3.10 eval job has completed with return code 0.

5 10

[still owl]

Hello what y'all talking about

[vague cairn]

!e

#Syntax isn't as cool, but this actually runs:
(lambda: [
    ['def'
        for sys in [__import__('sys')]
        for math in [__import__('math')]

        for sub in [lambda *vals: None]
        for fun in [lambda *vals: vals[-1]]

        for echo in [lambda *vals: sub(
            sys.stdout.write(" ".join(map(str, vals)) + "\n"))]

        for Cylinder in [type('Cylinder', (object,), dict(
            __init__ = lambda self, radius, height: sub(
                setattr(self, 'radius', radius),
                setattr(self, 'height', height)),

            volume = property(lambda self: fun(
                [self.height * top_area for top_area in [math.pi * self.radius ** 2]
                 ][0],

                ))))]

        for main in [lambda: sub(
            ['loop' for factor in [1, 2, 3] if sub(
                ['def'
                    for my_radius, my_height in [[10 * factor, 20 * factor]]
                    for my_cylinder in [Cylinder(my_radius, my_height)]
                 

                if echo(u"A cylinder with a radius of %.1fcm and a height "
                     u"of %.1fcm has a volume of %.1fcm³."
                     % (my_radius, my_height, my_cylinder.volume)
                     )
                 ]
                ,
            )])]
     if main()],

    ])()

[night quarry]

@vague cairn :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | A cylinder with a radius of 10.0cm and a height of 20.0cm has a volume of 6283.2cm³.
002 | A cylinder with a radius of 20.0cm and a height of 40.0cm has a volume of 50265.5cm³.
003 | A cylinder with a radius of 30.0cm and a height of 60.0cm has a volume of 169646.0cm³.

[rugged owl]

how do you get away with for sys in [__import__('sys')] without : at the end

[vague cairn]

it's a list comprehension.

[rugged owl]

ah yeah, mb didnt see that it was in a list

way too late

[vague cairn]

When I saw advik's code I def did a doubletake before I figured out what was going on, then I had to move the execution bits to inside the def lists.

(Because list comprehension vars are local to the comprehension)

[rugged owl]

Everytime I come into this channel I learn something funky syntax

[rugged owl]

!e
started messing around and came up with this shi

(lambda: [
    ['def'
        for math in [__import__('math')]
        for sys in [__import__('sys')]
        for numpy in [__import__('numpy')]
        for sub in [lambda *vals, **kwds: None]
        for fun in [lambda *vals: vals[-1]]

        for echo in [lambda *vals: sub(fun(
            sys.stdout.write(" ".join(map(str, vals)) + "\n")
        ))]

        for Rectangle in [type('Rectangle', (object, ), dict(
            __init__ = lambda self, length, width, height: sub(
                setattr(self, 'length', length),
                setattr(self, 'width', width),
                setattr(self, 'height', height)),
            

            area = property(lambda self: fun(
                [self.length * self.width * self.height * n for n in [abs(numpy.arcsin(-1)) // numpy.arcsin(1)]][0]
            ))))]
            
        
        for main in [lambda: sub(fun(
            ['loop' for factor in [1, 2, 3] if sub(
                ['def'
                    for my_height, my_length, my_width in [[10 * factor, 20 * factor, 30 * factor]]
                    for my_rect  in [Rectangle(my_length, my_width, my_height)]

                if echo(u"A rectangle with the area of %.1fcm^2 and a height "
                        u"of %.1fcm and a width "
                        u"of %.1fcm and a length "
                        u"of %.1fcm" % (my_rect.area, my_height, my_width, my_length))
                ]
                ,
        )]))]
    if main()],
        
    ])()

[night quarry]

@rugged owl :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | A rectangle with the area of 6000.0cm^2 and a height of 10.0cm and a width of 30.0cm and a length of 20.0cm
002 | A rectangle with the area of 48000.0cm^2 and a height of 20.0cm and a width of 60.0cm and a length of 40.0cm
003 | A rectangle with the area of 162000.0cm^2 and a height of 30.0cm and a width of 90.0cm and a length of 60.0cm

[rugged owl]

Weird how any indentation error, makes the variable not found. Like you have to indent atleast 1 tab to use all the imported libs

[frozen silo]

!e

print((lambda n: str(n) + 'tsnrhtdd'[n%5*(n%100^15>4>n%10)::4])(22))

[night quarry]

@frozen silo :white_check_mark: Your 3.11 eval job has completed with return code 0.

22nd

[frozen silo]

!e

print((lambda n: str(n) + 'tsnrhtdd'[n%5*(n%100^15>4>n%10)::4])(0))

[night quarry]

@frozen silo :white_check_mark: Your 3.11 eval job has completed with return code 0.

0th

[restive void]

!e

print((lambda n: str(n) + 'tsnrhtdd'[n%5*(n%100^15>4>n%10)::4])(-1))

[night quarry]

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

-1th