esoteric-python

[night quarry]

@unique heath :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello world!

[unique heath]

p.s: cheetsheet

>>> 81 // 34
2
>>> __builtins__.__dir__().__len__().__floordiv__(__import__.__dir__().__len__())
5
>>> __builtins__.__dir__().__len__().__floordiv__(__doc__.__dir__().__len__())
6
>>> __builtins__.__dir__().__len__().__floordiv__(__nam__.__dir__().__len__())
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
NameError: name '__nam__' is not defined. Did you mean: '__name__'?
>>> __builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())
1
>>> __builtins__.__dir__().__len__().__floordiv__(__spec__.__dir__().__len__())
6
>>> __builtins__.__dir__().__len__().__floordiv__(__debug__.__dir__().__len__())
2
>>> __doc__.__dir__().__len__()
25
>>> __builtins__.__dir__().__len__()
158
>>> __builtins__.__dir__().__len__().__floordiv__(15)
10
>>> __import__.__dir__().__len__()
30

mix and match stuff to make esoteric hell

[sick hound]

!e

T=int(input());r=i=1
while i<T:i+=1;r*=i
print(r)

[night quarry]

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

:warning: Note: input is not supported by the bot :warning:

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     T=int(input());r=i=1
004 |           ^^^^^^^
005 | EOFError: EOF when reading a line

[restive void]

Oh, is that a new notice?

!e

import sys, io
sys.stdin = io.StringIO("foo")
print(input())

[night quarry]

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

foo

[vast wave]

!e probably because of ```py
raise EOFError("EOF when reading a line")

[night quarry]

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

:warning: Note: input is not supported by the bot :warning:

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     raise EOFError("EOF when reading a line")
004 | EOFError: EOF when reading a line

[vast wave]

yeah

[last locust]

Apparently at least 3months old

[fleet bridge]

!e ```py
print('EOFError: EOF when reading a line')
import sys; sys.exit(1)

[night quarry]

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

:warning: Note: input is not supported by the bot :warning:

EOFError: EOF when reading a line

[last locust]

Committed March 1st & merged on March 3rd https://github.com/python-discord/bot/pull/2422

[versed eagle]

!e

from posix import *
print(read(0, 11))

[night quarry]

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

b''

[versed eagle]

huh

they should do something to prevent this then, no?

[last locust]

Why? The only reason input() doesn't work is because there's no sys.stdin.

It's not a security thing or anything

[versed eagle]

right but the warnings are inconsistent

thats what i was trying to demonstrate

also, there is sys.stdin

!e

import sys
print(hasattr(sys, "stdin"))

[night quarry]

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

True

[versed eagle]

it just returns EOF when read from

[versed eagle]

(which is why the read call returned 0 bytes)

[last locust]

!e py import sys print(sys.stdin)

[night quarry]

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

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

[last locust]

Fair enough

[last locust]

No readable stdin then

[versed eagle]

so imo they should make it more consistent by making it so that all reads to that file descriptor result in that warning (unless the fd is replaced with something like dup2)

[last locust]

I mean the idea is just to let people who aren't familiar with the error know what's going on

If you use some other obscure method then you know what you're doing

[versed eagle]

thats true, and completely valid! but it still bothers me that the warnings are inconsistent

[last locust]

You can always raise it in #dev-contrib, but you'll probably just get the same response

[versed eagle]

mhm

the only way it would ever get implemented is probably if i implemented it myself

which is sad because i want to do that now but i dont have the time

one day, perhaps

[sudden osprey]

The warning is implemented without requiring adding any code at runtime

It's simple by design

Also still not sure what you mean by the warnings being inconsistent.

[versed eagle]

the warning is meant to tell people they can't read input from the user, right?

but it shows the warning in some cases where that wasnt attempted, and doesnt always show it in cases where it was attempted

inconsistent

hopefully that explanation makes sense? sorry if it doesnt, i dont have enough sleep so i have trouble forming coherent thoughts

[sudden osprey]

Yeah it makes sense

Although if the aim was to be precise as possible, EOFError already does that

[versed eagle]

its not that i want it to be more descriptive, i just want it to be shown in some situations and not shown in others

to be better aligned with what the goal of the warning is

[fleet bridge]

this warning is designed for users that dont realise that you cannot use input() in snekbox
if you are doing this: ```py
from posix import *
print(read(0, 11))

[versed eagle]

yep, that's right

[versed eagle]

see above

[night quarry]

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

Hello, world!

[low lynx]

what does it do?

[serene stratus]

Print Hello world I assume 😄

[low lynx]

also that definitely counts as esoteric

[burnt pasture]

i am also confused about "not esoteric" 😄

[quartz wave]

start deobfuscating here ```py
def (__, , ):
try:
()
except :
...
(lambda: (( := (lambda , __=builtins, =builtins.getattr: (, ))), ( := (lambda : (lambda : ('setitem', (('decode', ('fromhex', ('bytes'))('676c6f62616c73'))('ascii'))())(('decode', ('bytes')())('utf-16'), ('sum')()), ('UnicodeDecodeError'), ))), ( := (lambda : ('join', ('str')())(('map')(('chr'), )))), ( := ((((111, 114, 100)))((((115, 116, 114)))((((115, 117, 112, 101, 114))))[9]) + (((111, 114, 100)))((((115, 116, 114)))((((102, 111, 114, 109, 97, 116))))[7]) + (((111, 114, 100)))((((115, 116, 114)))((((111, 114, 100))))[9]) + (((84, 114, 117, 101))))), (((䨭, 㔽, Ũ, ὕ, 夌)), (((匠, 猁, 堌, Ů, 䌲, 椋)), (((ⰳ, 刍, ⴼ, 怍, 䠨, 弐, ⵅ, 欉, ᙉ, ࡗ)))(((严, 㝂, 㰷)))))(((㠐, 堍, 椃, ⴿ, 搋, ᘖ, ഓ, 昑, Ů, 氆, 朅, 怄, ᰅ, Ȉ))) if {((, __)) for _ in (((114, 97, 110, 103, 101)))(_____) for __ in (_((114, 97, 110, 103, 101)))(_____)} else ''))()

[finite blaze]

def main(arg1, arg2, someFunc):
    try:arg1(someFunc)
    except arg2:...
(
    lambda: (
        (strToFunc := (lambda arg2, arg1=__builtins__, someFunc=__builtins__.getattr: someFunc(arg1, arg2))),
        (
            someFunc := (
                lambda arg2: main(
                    lambda arg2: strToFunc(
                        "__setitem__",
                        strToFunc(strToFunc("decode",b'globals',)("ascii"))(),
                    )(bytes(arg2).decode("utf-16"), sum(arg2)),UnicodeDecodeError,arg2,
                )
            )
        ),
        (more := (lambda arg2: str.join(map(chr, arg2)))),
        (someInt := 255),
        strToFunc("write",strToFunc("stdout",__import__("sys"),),
        )("Hello, world!\n")
        if {
            someFunc((arg2, arg1))
            for arg2 in range(someInt)
            for arg1 in range(someInt)
        }
        else "",
    )
)()

some more cleaning

def main(arg1, arg2, someFunc):
    try:arg1(someFunc)
    except arg2:...
(lambda: (
        (someFunc := (lambda arg2: main(lambda arg2: globals().__setitem__(bytes(arg2).decode("utf-16"), sum(arg2)),UnicodeDecodeError,arg2,))),
        __import__("sys").stdout.write("Hello, world!\n")
        if {
            someFunc((arg2, arg1))
            for arg2 in range(255)
            for arg1 in range(255)
        }else ""))()

First time deobfuscating code, pretty fun

[rough moat]

i posted this in general, but for adding and subtracting 1 to an int, you can do val = -~val and val = ~-val

[brittle cape]

yeag sounds right

[heady cosmos]

I want to encode two esp32s that communicate over an nrf24, but it doesn't work no matter how hard I try. Can anyone send me a short sample micropython code?

[sick hound]

is this pyarmor

[versed eagle]

wrong channel, see #❓｜how-to-get-help or go to a [correct] topic-specific channel

[grave rover]

Does this work on all platforms? I believe older systems don't necessarily use 2s complement which is what this relies on, no?

[gleaming linden]

https://docs.python.org/3/library/stdtypes.html#bitwise-operations-on-integer-types

The result of bitwise operations is calculated as though carried out in two’s complement with an infinite number of sign bits.
All implementations of python have to use 2s complement for bitwise operations, so yes

[grave rover]

I see

[last locust]

Does anyone here happen to know how to raise an error in cpython? I found PyErr_SetString(PyExc_ValueError, "the text of the value error"); but that doesn't seem to actually raise it.

[fleet bridge]

Does PyErr_SetRaisedException work?

Or any other PyErr_Set* function

[fleet bridge]

I think you also can do something like PyExecString("raise Exception") (i dont how this function is named exactly)

[last locust]

I managed to get an answer in #internals-and-peps. I just needed to add an explicit return NULL;

[sick hound]

bro how do y'all obfuscate at this level bro
It got me to question my existence. DO I EVEN KNOW PYTHON?

[serene stratus]

Did you manually do it or was it automated with a tool?

[dusk basin]

Ola

[vast wave]

no one knows python

there is always one more trick, one more standard library package, etc

[unreal echo]

Me saving hours just print("hello world")

[sick hound]

I also think the more you obfuscate your code the faster it will be

[unreal echo]

No not really

[sick hound]

is your name
uint64_t ?

also I think I saw someone

make that test

the more the code was obfuscated the faster it was

[fleet bridge]

Thats not true in almost all cases

Obfuscation usually needs redundant code, so obfuscated code os not faster

[low lynx]

if obfuscated code is faster it's probably not obfuscated for the sake of obfuscation

[versed eagle]

that could be true in some cases but in the vast majority of cases obfuscation will only make things slower

and even if it's faster at first, further obfuscation will make it slower eventually

[fleet bridge]

There are some microoptimization techniques that looks weird

[versed eagle]

i wouldn't call that obfuscation though

[bitter jacinth]

https://youtu.be/t863QfAOmlY this was fascinating, and hilariously esoteric

[torn tulip]

nah

[dreamy pier]

import time
with open("file.py") as file:
    text = file.read()
    time.sleep(1 / len(text))
    exec(text)

: D

(here should be some better formula because it will be too close to zero, but just as an example)

[versed eagle]

lmao

[errant slate]

whats the point of time.sleep() here? just a bit confused

[fleet bridge]

It makes code faster 😄

(by making it slower)

[errant slate]

thank you for the help

[keen thicket]

import os
timestep=1./60
gravity=-9.81
tg=timestep*gravity
s_x,s_y=10,20
u=v=[[.0 for A in range(s_y)]for A in range(s_x)]
pr=[[.0 for A in range(s_y)]for A in range(s_x)]
st=[[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1],[0,1,1,1,1,0,.1,0,0,.1,.1,.1,.1,.1,.1,.1,0,0,0,0],[0,1,1,1,1,1,.1,.1,.1,.1,.1,.1,.1,.1,.1,0,0,0,0,0],[0,1,1,1,1,1,.1,.1,.1,.1,.1,.1,.1,.1,.1,0,0,0,0,0],[0,.1,0,1,1,1,.1,.1,.1,.1,.1,.1,.1,.1,.1,0,0,0,0,0],[0,.1,0,.1,.1,.1,.1,.1,.1,.1,.1,.1,.1,.1,0,0,0,0,0,0],[0,.1,0,.2,.1,.1,.1,0,.1,.1,.1,.1,0,0,0,0,0,0,0,0],[0,.1,0,0,.2,.1,.1,0,.1,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0]]
C=1e3/timestep
loop=[(A,B)for A in range(1,len(u)-1)for B in range(1,len(u[0])-1)]
cl=lambda x,size:0 if x<=0 else size-1 if x>=size else x
sf=lambda i,j,field:((A:=cl(int(i-.5),s_x-1)),(C:=i-.5-A),(E:=cl(A+1,s_x-1)),(B:=cl(int(j-.5),s_y-1)),(D:=j-.5-B),(F:=cl(B+1,s_y-1)),(G:=1.-C),(H:=1.-D),G*H*field[A][B]+C*H*field[E][B]+C*D*field[E][F]+G*D*field[A][F])[-1]
ses='.',':','!','*','o','e','&','#','%','@'
for _ in range(2**10):
    u=[[A+tg for A in A]for A in u];v=[[A+tg for A in A]for A in v]
    for(i,j)in loop:
        l,r,t,b=u[i][j],u[i+1][j],v[i][j],v[i][j+1];ls,rs,ts,bs=st[i-1][j],st[i+1][j],st[i][j-1],st[i][j+1]
        if st[i][j]==0:continue
        s=ls+rs+ts+bs;d=l-r+t-b;p=-d/s*1.9;pr[i][j]+=p*C;u[i][j],v[i+1][j],u[i][j],v[i][j+1]=l+d*ls,r+d*rs,t+d*ts,b+d*bs
    for i in range(len(u)):u[i][0],u[i][s_y-1],v[0][j],v[s_x-1][j]=u[i][1],u[i][s_y-2],v[1][j],v[s_x-2][j]
    new_u=new_v=[[.0 for A in range(s_y)]for A in range(s_x)]
    for(i,j)in loop:
        if st[i][j]!=0 and st[i-1][j]!=0:x,y=i,j+.5;nu,nv=u[i][j],sum((v[i][j],v[i+1][j],v[i][j],v[i][j+1]))/4;x-=timestep*nu;y-=timestep*nv;new_u[i][j]=sf(x//1,y//1,u)
        if st[i][j]!=0 and st[i][j-1]!=0:x,y=i+.5,j;nu,nv=sum((u[i][j],u[i+1][j],u[i][j],u[i][j+1]))/4,v[i][j];x-=timestep*nu;y-=timestep*nv;new_v[i][j]=sf(x//1,y//1,v)
    u,v=new_u,new_v;os.system('clear')
    for line in pr:print(''.join([ses[min(int(abs(A)/C)//10,9)]for A in line]),flush=True)

https://asciinema.org/a/592380

fluid simulation in less than 2kB

[dusk basin]

hi

[fleet bridge]

!e ```py
import itertools
for _ in map(print, itertools.count()):
...

[night quarry]

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

001 | 0
002 | 1
003 | 2
004 | 3
005 | 4
006 | 5
007 | 6
008 | 7
009 | 8
010 | 9
011 | 10
... (truncated - too many lines)

Full output: too long to upload

[fleet bridge]

i just wanted to share it with you

[strange basin]

!e

@lambda _:_()
class print_:
    __getattr__ = lambda __,_:print(_.replace('_',' ').title()+'!')

print_.    hello_world
print_.    i_am_cursed
print_.    ᵃᵃᵃʳᵍʰ

[night quarry]

@strange basin :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | Hello World!
002 | I Am Cursed!
003 | Aaargh!

[strange basin]

!e

def echo():
    print('Echo ...')
    return echo

echo()() ()  ()   ()    ()     ()      ()       ()

[night quarry]

@strange basin :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | Echo ...
002 | Echo ...
003 | Echo ...
004 | Echo ...
005 | Echo ...
006 | Echo ...
007 | Echo ...
008 | Echo ...
009 | Echo ...

[strange basin]

!e

@lambda _:_()
def commented_out():
    print(f'Hello, cursed World!')

# commented_out()

[night quarry]

@strange basin :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, cursed World!

[strange basin]

!e

@lambda _:_()
def not_commented_out():
    print(f'Hello, cursed World!')

not_commented_out()

[night quarry]

@strange basin :x: Your 3.11 eval job has completed with return code 1.

001 | Hello, cursed World!
002 | Traceback (most recent call last):
003 |   File "/home/main.py", line 5, in <module>
004 |     not_commented_out()
005 | TypeError: 'NoneType' object is not callable

[dreamy pier]

!e

@lambda _:_()
def it_is_not_true():
    print(f'Hello, cursed World!')

[night quarry]

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

Hello, cursed World!

[unique heath]

ok but unicode characters and binary strings is just cheating

!e

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

[night quarry]

@unique heath :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello world!

[unique heath]

!e

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

[night quarry]

@unique heath :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello world!

[torn tulip]

weird doesn't work on my system

TypeError: BaseExceptionGroup.__new__() takes exactly 2 arguments (3 given)

[dreamy pier]

By me

TypeError: 'bool' object is not callable

[dreamy pier]

yeah, and the same error on 3.11

!e

(___ := print, _____ := lambda __: chr(__), eval('BaseExceptionGroup')(chr(__import__.__dir__().__len__().__add__(__builtins__.__dir__().__len__())), (object,), {chr(__name__.__dir__().__len__()): (lambda _: ___(_____(72).__add__(_____(101)).__add__(_____(108)).__add__(_____(108)).__add__(_____(111)).__add__(_____(32)).__add__(_____(119)).__add__(_____(111)).__add__(_____(114)).__add__(_____(108)).__add__(_____(100)).__add__(_____(33))))})().Q())

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     (___ := print, _____ := lambda __: chr(__), eval('BaseExceptionGroup')(chr(__import__.__dir__().__len__().__add__(__builtins__.__dir__().__len__())), (object,), {chr(__name__.__dir__().__len__()): (lambda _: ___(_____(72).__add__(_____(101)).__add__(_____(108)).__add__(_____(108)).__add__(_____(111)).__add__(_____(32)).__add__(_____(119)).__add__(_____(111)).__add__(_____(114)).__add__(_____(108)).__add__(_____(100)).__add__(_____(33))))})().Q())
004 |                                                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
... (truncated - too long)

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

[dreamy pier]

!e

eval('BaseExceptionGroup')("¼", (object,), {"Q": (lambda _: print("Hello world!"))})().Q()

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     eval('BaseExceptionGroup')("¼", (object,), {"Q": (lambda _: print("Hello world!"))})().Q()
004 |     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
005 | TypeError: BaseExceptionGroup.__new__() takes exactly 2 arguments (3 given)

[dreamy pier]

idk there should be type

but there is the baseexception

!e

type("¼", (object,), {"Q": (lambda _: print("Hello world!"))})().Q()

[night quarry]

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

Hello world!

[fleet bridge]

Do you know any algorithm with O(N^0.5) time complexity?

[earnest wing]

https://en.wikipedia.org/wiki/Grover's_algorithm

as well as things derived from it

[fleet bridge]

I expected classic algorithm, not quantum algorithm...

[earnest wing]

well, something simple then

def is_prime(n):
  for i in range(1, round(n**0.5)):
    if n % i == 0:
      return False
  return True

[fleet bridge]

Amother example: N is the number of cells in the square. Iterating over square side if O(N^0.5)

If you think about it, it is the same as your is_prime example

[vast wave]

!e ```py
@lambda :((),_)[1]
def what():
print("Hello chat")

what()

[night quarry]

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

001 | Hello chat
002 | Hello chat

[fleet bridge]

!e ```py
@lambda :()
def what():
print("Hello chat")
return _

what(what)(what)

[night quarry]

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

001 | Hello chat
002 | Hello chat
003 | Hello chat

[keen thicket]

!e

from collections import namedtuple
from itertools import takewhile
w_s,lhead_buf_sz=13,6
s_buf_sz,Token=w_s-lhead_buf_sz,namedtuple('Token',['offset','length','indicator'])
compress=lambda t,s_buf='':[((tok:=(lambda t,s_buf,l=0:Token(*[(0,0,t[0]),*([(len(s_buf)-i,(l:=fl),t[fl])for(i,ch)in enumerate(s_buf)if ch==t[0]and(fl:=(match_length:=lambda t,w,t_idx,w_idx:0 if t_idx+w_idx>=len(t)or not t or t[t_idx]!=w[w_idx]else 1+match_length(t,w+t[t_idx],t_idx+1,w_idx+1))(t,s_buf,0,i))>=l]if s_buf and len(t)>1 else[])][-1]))(t,s_buf)),(s_buf:=(s_buf+t[:tok.length+1])[-s_buf_sz:]),(t:=t[tok.length+1:]))[0]for _ in takewhile(lambda _:t,iter(int,1))]
decompress=lambda toks,o='':[([(o:=o+o[-tok.offset])for _ in range(tok.length)],(o:=o+tok.indicator))for tok in toks][-1][-1]
print(decompress(compress('"I will make no bargains with terrorist hardware."\n-- Peter da Silva')))

[night quarry]

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

001 | "I will make no bargains with terrorist hardware."
002 | -- Peter da Silva

[keen thicket]

lz77 in like 800 bytes

also does anybody know if there's a more elegant way to put a while loop in a list comprehension besides takewhile(lambda _: a and b or whatever,iter(int,1))

[versed eagle]

[*iter((lambda:your_boolean_expression_here),0)]

[keen thicket]

is the [*iter...] necessary?

[versed eagle]

you said you wanted it in a list

[keen thicket]

no haha, a list comprehension

as like the iterator for it

[something for _ in iter((lambda:your_boolean_expression_here),0)]

[versed eagle]

[_ for _ in iter((lambda:your_boolean_expression_here),0)]

[keen thicket]

but yeah I worded that confusingly

[versed eagle]

oh no i re-read it just now

i was misunderstanding

sorry

[versed eagle]

but yeah this is correct

[fringe oracle]

!e

something = [5064, 5111, 5148, 5175, 5193, 5239, 5273, 5324, 5349, 5355, 5360, 5364, 5401, 5405, 5410, 5459, 5519, 5569, 5577, 5677, 5684, 5696, 5720];WHAT = ''; exec("""for i in something: __import__("random").seed(i); WHAT += __import__("random").choice((__import__("string").ascii_lowercase + " ")); print(WHAT)""");

[night quarry]

@fringe oracle :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | n
002 | ne
003 | nev
004 | neve
005 | never
006 | never 
007 | never g
008 | never go
009 | never gon
010 | never gonn
011 | never gonna
... (truncated - too many lines)

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

[vast wave]

do we have numpy

!e print(import("numpy"))

[night quarry]

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

<module 'numpy' from '/snekbox/user_base/lib/python3.11/site-packages/numpy/__init__.py'>

[vast wave]

nice

!e ```py
import numpy as n
print(n.random.default_rng(0).integers(low=0, high=10, size=10))

[night quarry]

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

[8 6 5 2 3 0 0 0 1 8]

[vast wave]

!e ```py
import numpy as n
print(n.random.default_rng(0).random())

[night quarry]

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

0.6369616873214543

[vast wave]

i will eat a shoe

[night quarry]

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

lgbt eshek

[vast wave]

!e ```py
import numpy as n
n.random.seed(0);
print(n.random.rand())

[night quarry]

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

0.5488135039273248

[vast wave]

!e ```py
import numpy as n
n.random.seed(0);
print(n.random.choice("abc", 5))

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "mtrand.pyx", line 940, in numpy.random.mtrand.RandomState.choice
003 | TypeError: 'str' object cannot be interpreted as an integer
004 | 
005 | During handling of the above exception, another exception occurred:
006 | 
007 | Traceback (most recent call last):
008 |   File "/home/main.py", line 3, in <module>
009 |     print(n.random.choice("abc", 5))
010 |           ^^^^^^^^^^^^^^^^^^^^^^^^^
011 |   File "mtrand.pyx", line 942, in numpy.random.mtrand.RandomState.choice
... (truncated - too many lines)

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

[vast wave]

guh

!e ```py
import numpy as n
n.random.seed(0);
print(n.random.choice(list("abc"), 5))

[night quarry]

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

['a' 'b' 'a' 'b' 'b']

[vast wave]

!e ```py
import random
random.seed(56756)
print("".join(random.choices("abo olghwellrnopqhstud",k=11)))

[night quarry]

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

hello world

[gleaming linden]

wrong channel, and selfbots are against tos

!rule 5

[night quarry]

Rules

5. Do not provide or request help on projects that may violate terms of service, or that may be deemed inappropriate, malicious, or illegal.

[gleaming linden]

we can't help with it here

[grand urchin]

I posted this #software-architecture message in https://discord.com/channels/267624335836053506/782713858615017503 but I worry that this approach may be considered “esoteric” given how rare I've seen pumped-coroutine approaches in the wild. I am very curious to hear from people with experiece working with similar designs.

[grand urchin]

(For how rare the approach seems to be, it's quite pedestrian. It's also extremely effective in practice.)

[fleet bridge]

>>> class X(type(None)): ...
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: type 'NoneType' is not an acceptable base type
>>> class X(type(True)): ...
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: type 'bool' is not an acceptable base type
>>> class X(type(...)): ...
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: type 'ellipsis' is not an acceptable base type
>>> class X(type(print)): ...
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: type 'builtin_function_or_method' is not an acceptable base type
>>> class X(type(lambda:...)): ...
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: type 'function' is not an acceptable base type
>>> class X(type(object.__dict__['__class__'])): ...
...
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: type 'getset_descriptor' is not an acceptable base type

i guess there are more C-classes that you cannot subclass, than classes that you can subclass

almost all "normal" classes (list, tuple, dict, object, type, int) are subclassable (except bool and NoneType)
everything else is kinda implementation detail (their API is described, but not classes itself), so they are not subclassable

their API is described, but not classes itself
i mean print and lambda:... are "things you can call", exact classes of these things are impl detail (strictly speaking it is not...)
object.__dict__['__class__'] (=<attribute '__class__' of 'object' objects>) is a thing with __get__ and __set__, exact class is not documented (and it is an impl detail)

>>> class X(slice(slice)): ...
...
>>> X
slice('X', (slice(None, <class 'slice'>, None),), {'__module__': '__main__', '__qualname__': 'X'})
``` this will always be funny

>> class X(map(print,"abc")): ...
...
>>> X
<map object at 0x0000025950463280>
>>> [*X]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: 'str' object is not callable

>>> class X(map): ...
...
>>> X
<class '__main__.X'>
>>> X.__bases__
(<class 'map'>,)
>>> X(print, "abc")
<__main__.X object at 0x000002595092C460>
>>> [*X(print, "abc")]
a
b
c
[None, None, None]
``` wow!

you can subclass filter too
upd: and enumerate/zip/reversed

and classmethod/staticmethod/property

and super...

!e ```py
import builtins
for k, v in builtins.dict.items():
try:
class X(v):
pass
except:
pass
else:
if issubclass(v, BaseException):
continue
print(v.qualname, end=', ')

[night quarry]

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

BuiltinImporter, bytearray, bytes, classmethod, complex, dict, enumerate, filter, float, frozenset, property, int, list, map, object, reversed, set, staticmethod, str, super, tuple, type, zip, 

[fleet bridge]

# normal:
object
type

# kinda normal:
int
float
complex
dict
list
set
frozenset
tuple
str
bytes
bytearray

# cursed stuff:
classmethod
staticmethod
property

# ajsldjaslkdjaskd
super
enumerate
filter
map
reversed
zip

[arctic skiff]

how enumerate is ajsldjaslkdjaskd

[fleet bridge]

Why would you subclass enumerate?

[arctic skiff]

oh wait I didn't read full context its about subclass

[restive void]

To add support for step

[coarse void]

Quick question I was wondering about, but couldn't get it to work. Is there such a thing as "recursive decorators", where somehow a function or class decorates itself until some conditions happen or whatever...(?) It would be a hella esoteric piece of code

[dreamy pier]

!e

def foo(argument):
    print(argument)
    if argument < 5:
        @foo(argument + 1)
        def bar(function):
            return function
    else:
        def bar(function):
            return function
    return bar


@foo(0)
def test(): ...

[night quarry]

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

001 | 0
002 | 1
003 | 2
004 | 3
005 | 4
006 | 5

[kindred stone]

!e

foo = lambda a: (print(a), foo(a+1)(lambda f: f) if a < 5 else lambda f: f)[1]

@foo(0)
def test(): ...

[night quarry]

@kindred stone :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 0
002 | 1
003 | 2
004 | 3
005 | 4
006 | 5

[calm loom]

🛌

[golden finch]

What's the best way (in 3.11 bytecode) to increment the 3rd value on the stack?

[potent comet]

You'd want to use SWAP(3), LOAD_CONST(...), BINARY_OP(0), CACHE(0), SWAP(3) I think? Technically BINARY_OP_ADD_INT would be better, but that's a specialised opcode that probably wouldn't work if put directly in the code object. It should specialise to that quickly though. Also unsure if you need to add the cache entries yourself.

[golden finch]

Right, I got that far, minus CACHE 0. What does that do?

[golden finch]

Looking at the docs I can see it's some optimisation thing, and that certain opcodes expect to be followed by some number of caches. Is there a table of opcodes and the number of caches they need anywhere?

[potent comet]

I'm not sure if it's actually documented anywhere, which is a little problematic since it changes often.

opcode._cache_format defines the sizes, but that's not public...

[vast wave]

doesnt matter

you can access it regardless

https://github.com/0x3C50/pyasm if you dont want to manually write cache entries

[potent comet]

You can, but that might go away/change at any time, breaking your code and annoying the core devs...

[fleet bridge]

Weird behaviour of dicts (in #internals-and-peps):
#internals-and-peps message

[quartz wave]

yes there is

opcode._inline_cache_entries

or dis._inline_cache_entries if you prefer that

[golden finch]

Thanks!

[quartz wave]

that has to be accessed with the integer opcode because it's a list though

[golden finch]

cool

can't thank you enough

[quartz wave]

there's a problem with jumps here

it can't easily make a jump

[vast wave]

oh right i forgot to test jumps

[quartz wave]

import dis
from assembler import Assembler, Label

def func():
    a = Assembler()
    a.consts.extend([0] * 257 + [True]) # simulate EXTENDED_ARG
    a.insn("RESUME", 0)
    
    a.insn("LOAD_CONST", 257)
    a.insn("POP_JUMP_FORWARD_IF_FALSE", 2)
    
    a.insn("LOAD_CONST", 256)
    a.insn("COPY", 1)
    a.insn("POP_JUMP_BACKWARD_IF_TRUE", 5) # have to manually count jumps here
    
    a.insn("RETURN_VALUE", 0)
    
    return a.build()

dis.dis(func())

that's how i tested it

[fleet bridge]

https://leetcode.com/problems/partition-equal-subset-sum/
#algos-and-data-structs message

def canPartition(self, nums: List[int]) -> bool:
        sm = sum(nums)
        if sm & 1: return False
        bits = 1
        for x in nums: bits |= bits << x
        return bool(bits & (1<<sm//2))

[languid hare]

wait wtf

bits << x ???

[low lynx]

it's honestly a good idea

using an int for a dp table instead of a list

i'll borrow that

[simple panther]

how to make seive of erastothenes function using lambda?

i wish that i were exellent in lambda

[fleet bridge]

if you send here your "normal" code, i will try to turn it into lambda

[low lynx]

!e here's a quick one

test=lambda n:[[0,0,1,1][:n+1],(s:=[0,0]+[1]*~-n)and[s.__setitem__(k:=slice(p*p,n+1,p),[0]*len(s[k]))for p in range(2,int(n**.5)+1)if s[p]]and s][n>3]
print(test(100))

[night quarry]

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

[0, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0]

[night quarry]

@half escarp :white_check_mark: Your 3.11 eval job has completed with return code 0.

[2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97]

[night quarry]

@half escarp :white_check_mark: Your 3.11 eval job has completed with return code 0.

[2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97]

[fleet bridge]

That's cool

[dry cave]

20

[sonic birch]

Hi guys. I dynamicaly create a copy of function like new_f = FunctionType(f.__code__, f.__globals__, name=f.__name__ , argdef = inspect.getfullargspec(f).defaults, closure=f.__closure__). Is there any way for debugger to tell that the function is exactly the same so when I put a brackpoint into original f it will work?

[restive void]

How do you put a breakpoint in? If you insert breakpoint(), that will also apply to the copied function (because you also copy the call to breakpoint()).

[sonic birch]

In PyCharm

[sonic birch]

Oh, indeed, it works! It was some mess in the code. Thank you a lot

[cobalt mica]

scary

[languid hare]

mapping over discard is really neat

[night quarry]

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

[No output]

[sick hound]

!e

print([i for i in range(2, 100) if all(i % j != 0 for j in range(2, int(i ** 0.5) + 1))])

[night quarry]

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

[2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97]

[wheat river]

!e

sieve=lambda n:[*[p:={*range(2,n)},[p:=p-{*range(i*i,n,i)}for i in[*p]],p][2]]
print(sieve(100))

[night quarry]

@wheat river :white_check_mark: Your 3.11 eval job has completed with return code 0.

[2, 3, 67, 5, 7, 71, 73, 11, 13, 79, 17, 19, 83, 23, 89, 29, 31, 97, 37, 41, 43, 47, 53, 59, 61]

[quartz wave]

!e ```py
sieve=lambda n:[(p:={range(2,n)}).difference([range(ii,n,i)for i in p])]
print(sieve(100))

[night quarry]

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

[2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61, 67, 71, 73, 79, 83, 89, 97]

[quartz wave]

or ```py
sieve=lambda n:[*[p:={*range(2,n)},[p:=p-{range(ii,n,i)}for i in p],p][2]]

[wheat river]

oh yea, u can just iterate over p

[quartz wave]

now i wanted to see how it compares in ruby ```ruby
sieve=->(n){[t=2..n]-t.flat_map{[(_1*_1..n)%_1]}}

[karmic pumice]

!e

import sys;[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];
import random;[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];
[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[1];[1];
class App(object):#[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[
    @staticmethod #[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[1,]
    def main(args: list[str]) -> None:#[];[];[];[];[];[];[];[];[];[];[];[];[];[];[1,];
        if len(args)<=1:raise TypeError("Skill Issue");[];[];[];[];[];[];[];[];[];[1];
        buffer_hello:list[str]=[char for char in "hello"];[];[];[];[];[];[];[];[1];[];
        buffer_world:list[str]=[char for char in args[1]];[];[];[];[];[];[];[];[1];[];
        random.shuffle(buffer_hello);[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[1];
        random.shuffle(buffer_world);[];[];[];[];[];[];[];[];[];[];[];[];[];[];[];[1];
        while ''.join(buffer_hello)!="hello":random.shuffle(buffer_hello);[];[];[];[];
        while ''.join(buffer_world)!=args[1]:random.shuffle(buffer_world);[];[];[];[];
        sys.stdout.write(f"{''.join(buffer_hello)}, {''.join(buffer_world)}!");[];[1];
if __name__ == '__main__': App().main(['', 'shrek']);[];[];[];[];[];[];[];[];[];[];[];

[night quarry]

@karmic pumice :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello, shrek!

[jolly juniper]

i need an interpreter for vs code python

any help? pls

[versed eagle]

1. wrong channel
2. just sudo apt install python3 [or whatever the equivalent is if you're using windows]

[cursive portal]

Ok not sure what's the best channel but the question seems esoteric enough: "\U0001f40d", which is a nice snake unicode character 🐍, doesn't display in Windows terminal/Powershell. I only see the charcode. Yet it can show other characters such as ⚛️. Is that because the snake one is not supported ? Where can I get started to get a precise answer on such matters (I am a tad confused between unicode, utf8, "emojis", and their behaviour accross platforms...)

[vast wave]

cmon spamming an empty array isn't that based

[vast wave]

cmd and powershell are comically outdated when it comes to literally anything. try to literally print "\🐍"

works on gnome term

[cursive portal]

indeed I don't have the issue on Linux

I do print 🐍 directly, that's how I see it in the code editor

sad day for Python 😢

[cursive portal]

I am using the new shiny terminal too

but it runs powershell so I guess it's expected to have the same limitations

[versed eagle]

has nothing to do with python

it's an issue with windows stuff

[cursive portal]

well, it's about displaying a Python 🐍

more seriously I wondered if something could be off in the way I created/formatted the string but it doesn't seem so

[sonic birch]

Hi does anybody knows why ```
def foo():
locals()['f']=5
yield
print(locals())
yield

g = foo()
for s in g:
g.gi_frame.f_locals['f']=8
...

prints {'f': 8}

But ```
def foo():
  f=5
  yield
  print(locals())
  yield

g = foo()
for s in g:
  g.gi_frame.f_locals['f']=8
  ...

prints {'f': 5}

[fleet bridge]

Because locals()['f']=5 has no effect

Hmm

Im confused, i have no idea

[sonic birch]

BTW this is also strange

[faint vector]

~/NAME$ pip install hppx
Looking in indexes: https://package-proxy.replit.com/pypi/simple/
ERROR: Could not find a version that satisfies the requirement hppx (from versions: none)
ERROR: No matching distribution found for hppx
~/NAME$ 

[fleet bridge]

How is this related to this channel?

[faint vector]

idk what this channel is

idk where the help is for this

[fleet bridge]

#❓｜how-to-get-help

[mighty parcel]

!e

[night quarry]

Missing required argument

code

[sick hound]

What terminal is this?

[sick hound]

You probably meant httpx

[sick hound]

To get it to show the blue folder path I mean

[vast wave]

the name’s in the message

[sick hound]

Gnome terminal default looks like that?

[vast wave]

if you want to know the name of the prompt, powerlevel10k via zsh

but the terminal is gnome terminal, yes

[sick hound]

To get the blue folder thingy?

[vast wave]

yes

that’s pl10k

[sick hound]

Is that a zsh theme?

[vast wave]

yes

[forest meadow]

@cursive portal not an expert on this but each terminal in windows and linux implements its version of the emoji. So you can't rely on any terminal showing the same emoji or even supporting one at all.

This is similar to how a font will be shown for a unicode char. It takes the 16 bit (2 byte) number, and applies a font style and renders it. So it's not very portable in that terminals don't even copy paste an emoji correctly sometimes.

[magic wraith]

regarding the outdatedness of them on windows 10 you can resize windows width infinitely (try on discord), but in cmd, when i want to see a long mysql query output for example, it doesnt want to resize further than 1080px idk how it was related but felt like i gotta write it

[fleet bridge]

!e ```py
d = dict(a=1, b=2, c=3, d=4)
print(*map(list, zip(*d.items())))

[night quarry]

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

['a', 'b', 'c', 'd'] [1, 2, 3, 4]

[fleet bridge]

!e ```py
d = dict(a=1, b=2, c=3, d=4)
print(*zip(*d.items()))

[night quarry]

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

('a', 'b', 'c', 'd') (1, 2, 3, 4)

[night quarry]

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

True

[fleet bridge]

!e ```py
print(f'{any([]) = }')
print(f'{all([]) = }')

[night quarry]

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

001 | any([]) = False
002 | all([]) = True

[vast wave]

doesnt this break the contract

if all() is true, any() should as well

[low lynx]

not really

that's not really a rule

any is asking "is there anything here that is True?" so it defaults to a negative if there is nothing there
all is asking "is everything here True?" so it defaults to a positive if there is nothing there
if you negate all you have that it's (not (anything here that is False?)), so you have
any: (anything here that is True?)
all: (not (anything here that is False?))
if you have at least one thing, then you will have that at least one of those (anything here) conditions are true, so you can eventually deduce that all => any
but if you have nothing, then both of those will be False, so you can't deduce that

[serene garnet]

how do you think about this code?

g=len;o=range;z=list;k=map;v=int;y=zip
exec('''w=?u,f,l,a:f([[a([k[j:j+l]!k@u[i:i+l]])!j@o(0,g(u),l)]!i@o(0,g(u),l)])
e=?u,r:[[l !k@j !l@k]!i@o(1<<r)!j@y(*u[i])];n,m=$;u=[z($)!_@o(1<<n)]
f=[d:=?x:x[::-1],?x:[d(i)!i@x],?x:z(k(z,y(*d(x)))),?x:d(z(k(z,y(*x))))]
!_@o(m):a,b=$;r=n-b;l=g(u)>>r;u=e(w(u,z,l,f[a-1]),r)if a<5 else e(w(u,f[a-5],l,z),r)
!i@u:print(*i)'''.replace('?','lambda ').replace('$','k(v,input().split())').replace('!','for ').replace('@',' in '))

[fleet bridge]

I like that

[earnest wing]

you can also think of it this way: a contract with all() is that all(x) and all(y) is the same as all(x + y), but if x!=y==[] then this must mean all([])==True
this is different with any() because instead of and it uses or, you can probably figure out the logic there

[royal sun]

!e g=len;o=range;z=list;k=map;v=int;y=zip
exec('''w=?u,f,l,a:f([[a([k[j:j+l]!k@u[i:i+l]])!j@o(0,g(u),l)]!i@o(0,g(u),l)])
e=?u,r:[[l !k@j !l@k]!i@o(1<<r)!j@y(*u[i])];n,m=$;u=[z($)!@o(1<<n)]
f=[d:=?x:x[::-1],?x:[d(i)!i@x],?x:z(k(z,y(*d(x)))),?x:d(z(k(z,y(*x))))]
!@o(m):a,b=$;r=n-b;l=g(u)>>r;u=e(w(u,z,l,f[a-1]),r)if a<5 else e(w(u,f[a-5],l,z),r)
!i@u:print(*i)'''.replace('?','lambda ').replace('$','k(v,input().split())').replace('!','for ').replace('@',' in '))

[night quarry]

@royal sun :x: Your 3.11 eval job has completed with return code 1.

:warning: Note: input is not supported by the bot :warning:

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 2, in <module>
003 |     exec('''w=?u,f,l,a:f([[a([k[j:j+l]!k@u[i:i+l]])!j@o(0,g(u),l)]!i@o(0,g(u),l)])
004 |   File "<string>", line 2, in <module>
005 | EOFError: EOF when reading a line

[quartz wave]

making a nodevisitor rn

[quartz wave]

it works ```py

print(unparse(tree))
w = lambda u, f, l, a: f([[a([map[j:j + l] for k in u[i:i + l]]) for j in range(0, len(u), l)] for i in range(0, len(u), l)])
e = lambda u, r: [[l for k in j for l in k] for i in range(1 << r) for j in zip(*u[i])]
n, m = map(int, input().split())
u = [list(map(int, input().split())) for _ in range(1 << n)]
f = [(d := (lambda x: x[::-1])), lambda x: [d(i) for i in x], lambda x: list(map(list, zip(*d(x)))), lambda x: d(list(map(list, zip(*x))))]
for _ in range(m):
a, b = map(int, input().split())
r = n - b
l = len(u) >> r
u = e(w(u, list, l, f[a - 1]), r) if a < 5 else e(w(u, f[a - 5], l, list), r)
for i in u:
print(*i)

[sick hound]

Calculating the area of ​​figures

(lambda formuls: (lambda a: [print("Selected non-existent item"), exit()] if a not in range(len(formuls)) else (lambda formula, args: [print("Enter formula arguments:"), [args.update({arg["code_name"]: arg['type'](input(f"\t{arg['human_name']}(type {arg['type'].__name__}): "))}) for arg in formula["args"]], print("\n"), [print("Formula result:"), print(f"\t{formula['formula'](**args)}")]])(formuls[a], {}))((lambda: [print("Choose what you want to calculate:"), [print(f"\t{i}. {name}") for i, name in zip(range(1, len(formuls)+1), (formula['name'] for formula in formuls))], int(input("> ")) - 1, print("\n")])()[-2]))([{"name": "Square area", "formula": lambda x: x**2, "args": [{"code_name": "x", "human_name": "Side length", "type": float}]}, {"name": "Rectangle area", "formula": lambda a, b: a*b, "args": [{"code_name": "a", "human_name": "Side length 'a'", "type": float}, {"code_name": "b", "human_name": "Side length 'b'", "type": float}]}, {"name": "Circle area", "formula": lambda r: 3.14 * r**2,"args": [{"code_name": "r", "human_name": "Circle radius", "type": float}]}])

1087 symbols

[merry pasture]

to all ye maesteros

any way I could rewrite this in a single line ?

from requests.utils import default_headers
headers = default_headers()```

[static knoll]

use __import__

[quartz wave]

headers = __import__('requests.utils').utils.default_headers()

[merry pasture]

not familiar with _import_, so I ask - why do we use utils twice ?

like why is utils apperaing twice in the code ?

....

the repl shows once is enough, but twice also works

Thanks anyway !

[quartz wave]

__import__('requests.utils') returns the same thing as __import__('requests')

[fleet bridge]

But if you import requests it is not necessarily have .utils attr

Import system should import submodule and set attr on parent module, so it is more reliable to import requests.utils

[arctic skiff]

agree

[unique heath]

meh code tbh

!e cannot compare to the glory of

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

[night quarry]

@unique heath :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     (___:=eval(__builtins__.__dir__().__getitem__((420).__floordiv__(__import__.__dir__().__len__().__sub__(20)))),_____:=lambda __: eval(__builtins__.__dir__().__getitem__(14))(__),eval(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__getitem__(__import__.__dir__().__len__().__sub__(__name__.__doc__.__len__().__floordiv__(__builtins__.__doc__.__len__()))).__len__()).__sub__(__name__.__dir__().__len__().__sub__(__import__.__dir__().__len__()).__sub__(__package__.__dir__().__len__()).__sub__(__package__.__dir__().__len__())).__add__(4)))(chr(__import__.__dir__().__len__().__add__(__builtins__.__dir__().__len__())), (eval(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__getitem__(__import__.__dir__().__len__().__sub__(__name__.__doc__.__len__().__floordiv__(__builtins__.__doc__.__
... (truncated - too long)

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

[unique heath]

damn it doesnt work now

rewrite time

[unique heath]

!e

eval(__builtins__.__dir__().__getitem__(__name__.__dir__().__len__().__sub__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__()))))(__name__, (eval(__builtins__.__dir__().__getitem__(__name__.__dir__().__len__().__sub__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__sub__(__package__.__dir__().__len__().__add__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__floordiv__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__()))))), ), {chr(__name__.__dir__().__len__().__sub__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__sub__(__package__.__dir__().__len__().__add__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__floordiv__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())))): (lambda _: eval(__builtins__.__dir__().__getitem__(__package__.__dir__().__len__().__add__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__sub__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__add__(__package__.__dir__().__len__().__add__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__sub__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__()))).__sub__(__package__.__dir__().__len__().__add__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__floordiv__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())).__sub__(__name__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__().__sub__(__name__.__dir__().__len__().__floordiv__(__package__.__dir__().__len__())))))))("Hello, world!"))})().E()

[night quarry]

@unique heath :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello, world!

[arctic skiff]

good now explain its working

[unique heath]

its just an obfuscated version of hello world duh

[arctic skiff]

i know but what it does

[unique heath]

print hello world?

[arctic skiff]

https://haste.zneix.eu/talylarame.py

like mine does __import__('__hello__').main()

[dreamy pier]

unobfuscate it
: )

[restive void]

A few months ago, there was a challenge here about arbitrary code execution from eval() without dunders or builtins. Can someone send me a link to that, especially the/a solution? It's for a talk about eval being unsalvageable.

[dreamy pier]

type(
    __name__, (object,), {
        "E": (lambda _: print("Hello, world!"))
    })().E()

[fleet bridge]

#esoteric-python message

Scroll upwards from that

[restive void]

Thank you!

[kindred path]

hey vitnes

how does it work

@dreamy pier

[dreamy pier]

just creates a class with method E, which prints out "Hello, world!"

[kindred path]

what does this do again

what does this do

[kindred path]

oh ok thx

imma use it

[kindred path]

oh so it runs the os.system command?

@half escarp

does it work like a normal function

can i put parentheses at the end and run an os command

[nimble mirage]

Any concept/impl of comptime evalutation in python?

[kindred path]

code to concatenate two strings with a space inbetween

x = (string1, string2).__repr__().replace("(\'","(\"")
  y = x.replace("\', \'","\", \"")
  z = y.replace("\')","\")")
  (tempList:=z.split("(")[1].split(")")[0].split("\""))[1]+tempList[2][1]+tempList[3].replace("\\\\", "\\")

[ionic latch]

Hello python

Hi

[quartz wave]

!e ```py
string1 = "The World Wide Web (aka 'WWW')"
string2 = "('They' made) something"
x = (string1, string2).repr().replace("('","("")
y = x.replace("', '","", "")
z = y.replace("')","")")
(tempList:=z.split("(")[1].split(")")[0].split("""))[1]+tempList[2][1]+tempList[3].replace("\\", "\")

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 6, in <module>
003 |     (tempList:=z.split("(")[1].split(")")[0].split("\""))[1]+tempList[2][1]+tempList[3].replace("\\\\", "\\")
004 |                                                              ~~~~~~~~^^^
005 | IndexError: list index out of range

[kindred path]

you need to escape the single quotes

[quartz wave]

wdym by escape

[quartz wave]

!e ```py
string1 = "The World Wide Web (aka \'WWW\')"
string2 = "(\'They\' made) something"
x = (string1, string2).repr().replace("('","("")
y = x.replace("', '","", "")
z = y.replace("')","")")
(tempList:=z.split("(")[1].split(")")[0].split("""))[1]+tempList[2][1]+tempList[3].replace("\\", "\")

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 6, in <module>
003 |     (tempList:=z.split("(")[1].split(")")[0].split("\""))[1]+tempList[2][1]+tempList[3].replace("\\\\", "\\")
004 |                                                              ~~~~~~~~^^^
005 | IndexError: list index out of range

[quartz wave]

!e ```py
string1 = "The World Wide Web (aka 'WWW')"
string2 = "('They' made) something"
x = (string1, string2).repr().replace("('","("")
y = x.replace("', '","", "")
z = y.replace("')","")")
(tempList:=z.split("(")[1].split(")")[0].split("""))[1]+tempList[2][1]+tempList[3].replace("\\", "\")

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 6, in <module>
003 |     (tempList:=z.split("(")[1].split(")")[0].split("\""))[1]+tempList[2][1]+tempList[3].replace("\\\\", "\\")
004 |                                                              ~~~~~~~~^^^
005 | IndexError: list index out of range

[quartz wave]

what is "escape"

[kindred path]

\'

[quartz wave]

@kindred path ^

[kindred path]

Why

Doesn’t it work

Well I don’t rly care

It’s fine

U can try to fix it if you want

!e ```py
string1 = "The World Wide Web aka WWW"
string2 = "They made something"
x = (string1, string2).repr().replace("('","("")
y = x.replace("', '","", "")
z = y.replace("')","")")
(tempList:=z.split("(")[1].split(")")[0].split("""))[1]+tempList[2][1]+tempList[3].replace("\\", "\")

!e ```py
string1 = "The World Wide Web aka WWW"
string2 = "They made something"
x = (string1, string2).repr().replace("('","("")
y = x.replace("', '","", "")
z = y.replace("')","")")
print(tempList:=z.split("(")[1].split(")")[0].split("""))[1]+tempList[2][1]+tempList[3].replace("\\", "\")

[night quarry]

@kindred path :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[kindred path]

!e ```py
string1 = "The World Wide Web aka WWW"
string2 = "They made something"
x = (string1, string2).repr().replace("('","("")
y = x.replace("', '","", "")
z = y.replace("')","")")
print(tempList:=z.split("(")[1].split(")")[0].split("""))[1]+tempList[2][1]+tempList[3].replace("\\", "\")

[night quarry]

@kindred path :x: Your 3.11 eval job has completed with return code 1.

001 | ['', 'The World Wide Web aka WWW', ', ', 'They made something', '']
002 | Traceback (most recent call last):
003 |   File "/home/main.py", line 6, in <module>
004 |     print(tempList:=z.split("(")[1].split(")")[0].split("\""))[1]+tempList[2][1]+tempList[3].replace("\\\\", "\\")
005 |     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^
006 | TypeError: 'NoneType' object is not subscriptable

[versed eagle]

there's some const-folding in cpython

theres also that one almost-python thing that compiles into llvm stuff
which would have a lot more optimisations done to it

but no way to explicitly do anything at compile-time with that, its all up to the compiler

[versed eagle]

url to that if anyone cares https://github.com/exaloop/codon

[languid hare]

!e

import numpy as np
print([1, np.array([1,2])].count(1))

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 2, in <module>
003 |     print([1, np.array([1,2])].count(1))
004 |           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
005 | ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

[languid hare]

aw

[restive void]

In [88]: for a, b in zip(x, y, strict=True):
    ...:     if a != b:
    ...:         raise ValueError
    ...:

In [89]: x
Out[89]: 'ooooo'

In [90]: y
Out[90]: 'ooooo'

In [91]: x == y
Out[91]: False

What did I do? x and y really are strings, and I didn't mess with the str class (patching __eq__ or something like that).

[fleet bridge]

!e assert 'o' == 'о'

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     assert 'o' == 'о'
004 |            ^^^^^^^^^^
005 | AssertionError

[fleet bridge]

import sys

def hook(obj: object) -> None:
  if isinstance(obj, bool):
    print(not obj)
  else:
    print(repr(obj))

sys.displayhook = hook

###
x = 'ooooo'
y = 'ooooo'

for a, b in zip(x, y, strict=True):
  if a != b:
    raise ValueError

# then in REPL do this
>>> x
'ooooo'
>>> y
'ooooo'
>>> x == y
False
>>>

!e maybe you did this 🤔 ```py
print('''
In [88]: for a, b in zip(x, y, strict=True):
...: if a != b:
...: raise ValueError
...:

In [89]: x
Out[89]: 'ooooo'

In [90]: y
Out[90]: 'ooooo'

In [91]: x == y
Out[91]: False
'''.strip())

[night quarry]

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

001 | In [88]: for a, b in zip(x, y, strict=True):
002 |     ...:     if a != b:
003 |     ...:         raise ValueError
004 |     ...:
005 | 
006 | In [89]: x
007 | Out[89]: 'ooooo'
008 | 
009 | In [90]: y
010 | Out[90]: 'ooooo'
011 | 
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/U3AMCIJ4SKLS7R4XVLCCEWQZZE

[fleet bridge]

another possibility (you should replace a with something that looks like an o, for example russian o: о)

>>> zip = lambda *_, **__: ()
>>>
>>> x = 'ooooo'
>>> y = 'aaaaa'
>>>
>>> for a, b in zip(x, y, strict=True):
...   if a != b:
...     raise ValueError
...
>>>
>>> x
'ooooo'
>>> y
'aaaaa'
>>> x == y
False

did i guess?

i copy-pasted your symbols and they are in fact equal: ```py

'ooooo' == 'ooooo'
True

so i see two options:
1) `sys.excepthook` does nothing
2) `zip` is shadowed

it is also possible that you patched REPL itself too much (it ignores some predefined strings, or compiles them differently)

[restive void]

Good guesses!
No builtins have been patched at all, the REPL is normal.

These really are two strings of length 5, each of their elements is a normal ASCII o.

[fleet bridge]

>>> 'x' == 'x'
True
>>> 'y' == 'y'
True
``` 🤔

no weird name trickery too

[restive void]

Correct. No cheap tricks

[quartz wave]

oh

>>> x = 'ooooo'
>>> y = 'ooooo'
>>> for a, b in zip(x, y, strict=True):
...     if a != b:
...             raise ValueError
...
>>>

wait nvm why am i looking at that

[restive void]

Perhaps a hint:
Both x[1:] == y[1:] and x[:-1] == y[:-1] are true

[quartz wave]

i thought there was some cpython bug with zip() there but ig not

[fleet bridge]

you have some invisible unicode character in these strings?

hmm

no, that would trigger ValueError

[restive void]

No, as I said: exactly 5 characters, each of them an ASCII o

[fleet bridge]

you somehow forced x to be 2-byte kind, so _PyUnicode_Equal returns False

>>> x = 'oooooж'
>>> x = x[:-1]
>>> y = 'ooooo'
>>> x
'ooooo'
>>> y
'ooooo'
>>> x == y
True
``` that didnt work

[quartz wave]

cpython always adjusts a new string that's created

[fleet bridge]

:(

that makes sense

[quartz wave]

so that it always uses the least possible size

[restive void]

Correct!

[fleet bridge]

how did you do that? ctypes stuff or just normal python?

[restive void]

ctypes stuff

[fleet bridge]

ok, then it is not a bug 😄

[restive void]

Yeah, not a bug. But interesting behaviour :)
And you could make a z as well.

[quartz wave]

i thought this was gonna be easy ```pycon

import ctypes
p = ctypes.pythonapi.PyUnicode_FromKindAndData
p.argtypes = ctypes.c_int, ctypes.c_char_p, ctypes.c_ssize_t
p.restype = ctypes.py_object
x = p(2, b'o\0o\0o\0o\0o\0', 5)
y = 'ooooo'
for a, b in zip(x, y, strict=True):
... if a != b:
... raise ValueError
...
x
'ooooo'
y
'ooooo'
x == y
True

oh nvm it is ```pycon

p = ctypes.pythonapi.PyUnicode_New
p.argtypes = ctypes.c_ssize_t, ctypes.c_uint
p.restype = ctypes.py_object
x = p(5, 65535)
ctypes.memmove(id(x) + 72, b'o\0o\0o\0o\0o\0', 10)
2236930728024
x
'ooooo'
y
'ooooo'
for a, b in zip(x, y, strict=True):
... if a != b:
... raise ValueError
...
x
'ooooo'
y
'ooooo'
x == y
False
x[1:] == y[1:]
True
x[:-1] == y[:-1]
True

[restive void]

I just did x = 'ooſoo'; ctypes.c_byte.from_address(id(x) + 77).value = 0; ctypes.c_byte.from_address(id(x) + 76).value = 111.

[fleet bridge]

>>> import ctypes
>>> (ctypes.c_uint8 * 64).from_address()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: c_ubyte_Array_64.from_address() takes exactly one argument (0 given)
>>>
>>> x = 'ooooж'
>>> bytes((ctypes.c_uint8 * 64).from_address(id(x))).hex(' ', -8)
'0100000000000000 7094559ff97f0000 0500000000000000 0c65ac0d59249158 a8006e005c003300 f8d08d6556010000 0000000000000000 0000000000000000'
>>> hex(id(x))
'0x156658dd0b0'
>>> bytes.fromhex('f8d08d6556010000')
b'\xf8\xd0\x8deV\x01\x00\x00'
>>> bytes.fromhex('f8d08d6556010000')[::-1]
b'\x00\x00\x01Ve\x8d\xd0\xf8'
>>> bytes.fromhex('f8d08d6556010000')[::-1].hex()
'00000156658dd0f8'
>>> eval('0x'+bytes.fromhex('f8d08d6556010000')[::-1].hex())
1_470_582_608_120
>>> bytes((ctypes.c_uint8 * 64).from_address(1_470_582_608_120)).hex(' ', -8)
'6f006f006f006f00 3604000000006f00 6f00360427000000 0400000000000000 7094559ff97f0000 2800000000000000 88fa97a82d86ae21 e4006e005c003300'
>>> ord('o')
111
>>> hex(ord('o'))
'0x6f'
>>> bytes((ctypes.c_uint8 * 2).from_address(1_470_582_608_120 + 8)).hex(' ', -8)
'3604'
>>> (ctypes.c_uint8 * 2).from_address(1_470_582_608_120 + 8) = [0x6f, 0]
  File "<stdin>", line 1
    (ctypes.c_uint8 * 2).from_address(1_470_582_608_120 + 8) = [0x6f, 0]
    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
SyntaxError: cannot assign to function call here. Maybe you meant '==' instead of '='?
>>> (ctypes.c_uint8 * 2).from_address(1_470_582_608_120 + 8)[:] = [0x6f, 0]
>>>
>>> x
'ooooo'
``` i figured this out

[quartz wave]

i think you could do it like ctypes.c_ushort.from_address(id(x) + 76).value = 111

[fleet bridge]

x = 'жoooo'
index = 0
kind = 2
change_to = 'o'

import ctypes
ctypes.c_uint16.from_address(ctypes.c_void_p.from_address(id(x) + 5 * ctypes.sizeof(ctypes.c_void_p)).value + index * kind).value = ord(change_to)

for some reason there is a pointer to begginning of data in str object

so i used that

[languid night]

This is cool I will use in production thanks :))

[karmic pumice]

@quartz wave

get_string = efrem_utils.get_validated_input()

💀 0.0.7 patch going crazy

[quartz wave]

nice

[quartz wave]

now make msg a normal argument ```diff

• def get_validated_input(*, msg: str | None = None, validators: ...

• def get_validated_input(msg: str | None = None, *, validators: ...

[karmic pumice]

why 😢 kwargs gaming

ok imma add this because i just noticed i need to fix a typehint 💀:
hopefully the typehints make sense now and msg is now a normal argument

[versed eagle]

you can still pass it as a kwarg, its just not kwonly then

[mental vector]

Does anyone have any interesting articles they'd recommend on some obscure or bizarre python metaprogramming/black-magic techniques? Not looking to learn anything practical necessarily, I just enjoy reading about the crazy stuff people come up with and understanding the inner workings of python

[karmic pumice]

looks like .pyc to me

you can manually generate it with something like:

py -m compileall <file1.py> [<file2.py> <fileN.py>]

and its not really obfuscation, its.. compilation

[fickle stratus]

anyone got a cool idea i can try to work out in python?

[karmic pumice]

!e

(_while:=lambda cond, fn, args:((fn(args()if(callable(args))else(args))),_while(cond, fn, args)if(cond())else(None)),x:=10,_while(lambda: x>0, lambda _:(globals().__setitem__("x",x-1),print(x)),None))

[night quarry]

@karmic pumice :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 9
002 | 8
003 | 7
004 | 6
005 | 5
006 | 4
007 | 3
008 | 2
009 | 1
010 | 0

[karmic pumice]

experimenting with functional stuff

[vast wave]

!e (w:=lambda c,f:(f(),w(c,f)if c()else 0),x:=10,w(lambda:x>0,lambda:(x:=x-1,print(x))))

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     (w:=lambda c,f:(f(),w(c,f)if c()else 0),x:=10,w(lambda:x>0,lambda:(x:=x-1,print(x))))
004 |                                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
005 |   File "/home/main.py", line 1, in <lambda>
006 |     (w:=lambda c,f:(f(),w(c,f)if c()else 0),x:=10,w(lambda:x>0,lambda:(x:=x-1,print(x))))
007 |                     ^^^
008 |   File "/home/main.py", line 1, in <lambda>
009 |     (w:=lambda c,f:(f(),w(c,f)if c()else 0),x:=10,w(lambda:x>0,lambda:(x:=x-1,print(x))))
010 |                                                                           ^
011 | UnboundLocalError: cannot access local variable 'x' where it is not associated with a value
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/DRFT6ZOPLTNNMV7X6H3GVSCZBU

[vast wave]

ah thats why that was there

[night quarry]

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

001 | 9
002 | 8
003 | 7
004 | 6
005 | 5
006 | 4
007 | 3
008 | 2
009 | 1
010 | 0

[ashen adder]

the feeling when no x:-=1

[dreamy pier]

x:--

[rigid knoll]

!e False == False in [ False ]

!e print(False == False in [ False ])

[night quarry]

@rigid knoll :white_check_mark: Your 3.11 eval job has completed with return code 0.

True

[fleet bridge]

!e ```py
print(1 in [1] == True)

[night quarry]

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

False

[rigid knoll]

Yes Same pattern and Same Reason

Reason

[fleet bridge]

!e ```py
print(1 in [1]) # True
print(True == True) # True

print(1 in [1] == True) # False?

[night quarry]

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

001 | True
002 | True
003 | False

[rigid knoll]

!e

[night quarry]

Missing required argument

code

[rigid knoll]

!e
a=[],
try:
a[0]+=[42]
except Exception as e:
print(e)
print(a)

[night quarry]

@rigid knoll :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 'tuple' object does not support item assignment
002 | ([42],)

[languid hare]

neat

i suppose it makes sense, no matter how unexpected

[rigid knoll]

It's even extra documented: https://docs.python.org/3/faq/programming.html#why-does-a-tuple-i-item-raise-an-exception-when-the-addition-works

[rigid knoll]

!e [] = {}

[night quarry]

@rigid knoll :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[rigid knoll]

!e {} = []

[night quarry]

@rigid knoll :x: Your 3.11 eval job has completed with return code 1.

001 |   File "/home/main.py", line 1
002 |     {} = []
003 |     ^^
004 | SyntaxError: cannot assign to dict literal here. Maybe you meant '==' instead of '='?

[fleet bridge]

Imagine assigning to set literal

Variables are assigned to values in random order

{a, b} = 1, 2
print(a, b) # 1 2 or 2 1

[vast wave]

??

!e ```py
for _ in range(10):
{a, b} = 1, 2
print(a, b) # 1 2 or 2 1

[night quarry]

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

001 |   File "/home/main.py", line 2
002 |     {a, b} = 1, 2
003 |     ^^^^^^
004 | SyntaxError: cannot assign to set display here. Maybe you meant '==' instead of '='?

[fleet bridge]

I know, i was just dreaming about this

[rigid knoll]

The sad part here is that the variables within a dict have a defined order.
You can even use the dict as a stack with the popitem() function:

!e
print({1:1, 2:2}.popitem())

!e
() = [] = {}

[night quarry]

@rigid knoll :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[restive void]

Even the items in a set have a defined order, it's just not by insertion.

[ember flower]

return requests.get(...).json()['key']doesnt work?
It returns the same as:
return requests.get(...).json()

I need to do this:

return x['key']```

[restive void]

#❓｜how-to-get-help

[nova barn]

Is there a more pythonic way of combining two different sequences into like a tuple than tuple(itertools.chain(first, second))?

[rugged owl]

You are in the wrong channel for this question, either #python-discussion or #1035199133436354600 would be better

[digital mesa]

(*a, *b)

is maybe more readable

can do the same with [*a, *b] for lists

(where a and b are arbitrary iterables)

[nova barn]

Yeah I got this same thing on general and it blew my mind, when was this made possible?

[fleet bridge]

python2?

[nova barn]

damn, so it's just been something I never knew =D

[fleet bridge]

apparently not ```py

python.exe -c "x = (1,2); y = (3, 4); print (*x,*y)"
File "<string>", line 1
x = (1,2); y = (3, 4); print (*x,*y)
^
SyntaxError: invalid syntax

[nova barn]

https://peps.python.org/pep-0448/ 3.5

[fleet bridge]

yeah

2013

[nova barn]

10 years xD

It was like 3.5 or 3.6 when I started using pure 3 over hybrid 2 and 3.

And I internalized the {**sth, **other} but completely missed that well obviously (*sth, *other) should also work

I guess in retrospect this is the antithesis of the channel to ask this type of question in 😄

[fleet bridge]

!e ```py
chain = lambda *a: tuple((x:=[],[x.extend(_)for _ in a])[0])

print(chain())
print(chain([1,2],[3,4]))
print(chain('abcd','efg','h'))

[night quarry]

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

001 | ()
002 | (1, 2, 3, 4)
003 | ('a', 'b', 'c', 'd', 'e', 'f', 'g', 'h')

[fleet bridge]

!e ```py
chain = lambda *a: tuple((x:=[],[*map(x.extend,a)])[0])

print(chain())
print(chain([1,2],[3,4]))
print(chain('abcd','efg','h'))

[night quarry]

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

001 | ()
002 | (1, 2, 3, 4)
003 | ('a', 'b', 'c', 'd', 'e', 'f', 'g', 'h')

[dreamy pier]

!e

chain = lambda *a: (*sum(([*x] for x in a),[]),)

print(chain())
print(chain([1,2],[3,4]))
print(chain('abcd','efg','h'))

[night quarry]

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

001 | ()
002 | (1, 2, 3, 4)
003 | ('a', 'b', 'c', 'd', 'e', 'f', 'g', 'h')

[quartz wave]

!e ```py
f=lambda*a:sum(map(tuple,a),())
print(f())
print(f([1,2],[3,4]))
print(f('abcd','efg','h'))

[night quarry]

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

001 | ()
002 | (1, 2, 3, 4)
003 | ('a', 'b', 'c', 'd', 'e', 'f', 'g', 'h')

[wet fulcrum]

!e

print(isinstance(True, bool))```

[night quarry]

@wet fulcrum :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | True
002 | True

[restive void]

float is misleading; True is not a float. It is an int, though, yes.

[wet fulcrum]

interesting

[restive void]

!e Can be seen as a historical artifact. In early versions of Python bools didn't exist, and the ints 1 and 0 where used instead. When it was added, bool was made a subclass of int.

print(f"{issubclass(bool, int)=}")
print(f"{1 == True=}")
print(f"{0 == False=}")

[night quarry]

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

001 | issubclass(bool, int)=True
002 | 1 == True=True
003 | 0 == False=True

[fleet bridge]

>>> s = 'print(123)'
>>> n = int.from_bytes(s.encode()); print(f'eval(({n}).to_bytes({(n.bit_length()+7)//8}))')
eval((531015572043140546310953).to_bytes(10))

!e eval((531015572043140546310953).to_bytes(10))

[night quarry]

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

123

[fleet bridge]

it uses this character set: ().0123456789_abelostvy
it can execute arbitrary code

[lunar marsh]

whats the lowest amount of characters to execute arbitrary code anyone has managed in python3? i've got 8

[serene stratus]

wdym with arbitrary code

[lunar marsh]

any python program

[dreamy pier]

py x
where x is file
or what

[lunar marsh]

!e ```e
exec('x=%x'%('e'==''))==exec('e=%x'%(''==''))==exec('c=%x%%x'%e%e)==exec('c=%x%%x%%%%x%%%%%%%%x'%x%c%e%e)==exec('ee=%x%%x%%%%x'%e%x%c)==exec('ee=%x'%ee)==exec('ee=%x'%ee)==exec('xe=ee%ce%%ce'%ee%ee)==exec('xx=ee%cc'%xe)==exec('xc=xx%ce'%ee)==exec('ce=xc%ce'%ee)==exec('cx=ce%cc'%xe)==exec('cc=cx%ccx'%ee)==exec('eee=cc%cxc'%ee)==exec('eex=eee%cc'%xe)==exec('eec=eex%cxc'%xe)==exec('exe=eec%cee'%xe)==exec('exx=exe%ccc'%ee)==exec('exc=exx%ce'%ee)==exec('ece=exc%cc'%xe)==exec('ecx=ece%cc'%xe)==exec('ecc=ecx%ce'%xe)==exec('xee=ecc%cc'%xe)==exec('xex=xee%ce'%xe)==exec('xec=exx%cxex'%xe)==exec('xxe=xec%cecc'%ee)==exec('xxx=xxe%ce'%ee)==exec('xxc=xxx%cc'%xe)==exec('''ex=('%c%%c%%%%c%%%%%%%%c')''')==exec('''ec=('')''')==exec('ec%c=ex%%xxc%%xxe%%ecx%%exx'%ee)==exec('ec%c=ex%%eex%%xx%%cx%%ecc'%ee)==exec('ec%c=ex%%xee%%ece%%ece%%exc'%ee)==exec('ec%c=ex%%exe%%eee%%exc%%xxe'%ee)==exec('ec%c=ex%%ece%%xex%%cx%%xc'%ee)==exec(ec)

[night quarry]

@lunar marsh :white_check_mark: Your 3.11 eval job has completed with return code 0.

hello world

[lunar marsh]

like this

here using only exc()'%=

[dreamy pier]

ah
then idk never did that

[restive void]

You likely found the minimum. There's a slightly different solution here: https://codegolf.stackexchange.com/a/250500

[serene stratus]

impressive

[finite blaze]

Hey, do you know any place (discord sever for example) where I can talk about C# golfing?

Or just golfing in general

[fleet bridge]

There is esolangs server, maybe you also can talk about golfing there

I dont know if there is a server about golfing

[finite blaze]

Alright, thanks

[dreamy pier]

This site has discord server
https://code.golf/

There is not a lot of talking but something

[meager zinc]

im wondering if there's a better solution for https://leetcode.com/problems/merge-strings-alternately/
I have this:

from itertools import*
class Solution:mergeAlternately=lambda s,*w:''.join(chain(*zip_longest(*w,fillvalue='')))

(which ignoring the leetcode stuff becomes)

from itertools import*
lambda *w:''.join(chain(*zip_longest(*w,fillvalue='')))

[sick hound]

M,P,O,S,Z=3e4,0,'',[]
while 1:
 I=input();A='><+-.,[]'
 if I.lower()=='exit':print('Quitting... thanks for using!');break
 I=''.join(c for c in I if c in A)
 L=len(I)
 for i in range(L):
  c=I[i]
  if c in"><":P+=1if c=='>'else-1
  if c in"+-":Z[P]=(Z[P]+1)%256if c=='+'else(Z[P]-1)%256
  if c==".":O+=chr(Z[P])
  if c==",":Z[P]=ord(input("Enter a single character: ")[0])
  if c=="[":
   if Z[P]==0:C=1;R=0
   while C>0:
    i+=1
    C+=1if I[i]=='['else-1if I[i]==']'else 0
  if c=="]":i=S.pop()-1if Z[P]!=0else i
print(O)

[meager zinc]

uhh

Traceback (most recent call last):
  File "main.py", line 1, in <module>
    M,P,O,S,Z=3e4,0,'',[]
ValueError: not enough values to unpack (expected 5, got 4)

this might be similar to what you're trying to do (this isn't perfect however, and could likely be more optimized):

from sys import*
s=[i:=0]*30000
while 1:
 if(p:=input()).lower()=='exit':break
 else:
  i,c=0,''
  for o in p:c+=' '*i+[f'i+=1','i-=1','s[i]=-~s[i]%256','s[i]=~-s[i]%256','print(end=chr(s[i]))','s[i]=ord(stdin.read(1))','while s[i]:','s','']['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
  exec(c)

[arctic skiff]

instead of if(blah):break you can do while(blah):

that would save 2 lines

[meager zinc]

actually I don't need an else condition at all bc it breaks out of the loop

from sys import*
s=[i:=0]*30000
while 1:
 if(p:=input()).lower()=='exit':break
 i,c=0,''
 for o in p:c+=' '*i+[f'i+=1','i-=1','s[i]=-~s[i]%256','s[i]=~-s[i]%256','print(end=chr(s[i]))','s[i]=ord(stdin.read(1))','while s[i]:','s','']['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
 exec(c)

[arctic skiff]

you don't even need if

from sys import*
s=[i:=0]*30000
while(p:=input()).lower()!='exit':
 i,c=0,''
 for o in p:c+=' '*i+[f'i+=1','i-=1','s[i]=-~s[i]%256','s[i]=~-s[i]%256','print(end=chr(s[i]))','s[i]=ord(stdin.read(1))','while s[i]:','s','']['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
 exec(c)

[meager zinc]

it would be !=

[arctic skiff]

forgot about that

[meager zinc]

this is better

from sys import*
s=[0]*30000
while(p:=input()).lower()!='exit':
 i,c=0,''
 for o in p:c+=' '*i+'i+=1 i-=1 s[i]=-~s[i]%256 s[i]=~-s[i]%256 print(end=chr(s[i])) s[i]=ord(stdin.read(1)) while(s[i]): s '.split()['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
 exec(c)

[static knoll]

whats the sys import for

apart from stdin

[dreamy pier]

only it i guess

[dreamy pier]

f=open(0)
s=[0]*30000
while(p:=input()).lower()!='exit':
 i,c=0,''
 for o in p:c+=' '*i+'i+=1 i-=1 s[i]=-~s[i]%256 s[i]=~-s[i]%256 print(end=chr(s[i])) s[i]=ord(f.read(1)) while(s[i]): s '.split()['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
 exec(c)

even so

s=[0]*30000
while(p:=input()).lower()!='exit':
 i,c=0,''
 for o in p:c+=' '*i+'i+=1 i-=1 s[i]=-~s[i]%256 s[i]=~-s[i]%256 print(end=chr(s[i])) s[i]=ord(open(0).read(1)) while(s[i]): s '.split()['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
 exec(c)

[tough willow]

stdout, sys.platform, sys.modules, sys.setrecursionlimit

and more

and sys.exit

[static knoll]

i meant in the context of the golf

🥴

[tough willow]

oh

[rugged owl]

Pretty sure it was just used for stdin

[tough willow]

but cant you just open(0) instead of sys.stdin

no wiait it was 1

oridk

[proper vault]

open(0) should work, ye

[tough willow]

i was confused a bit because the REPL didnt work with both

[dreamy pier]

0 stdin
1 stdout
2 stderr

[quartz wave]

s=[0]*30000
while(p:=input()).lower()!='exit':
 i,c=0,''
 for o in p:c+=' '*i+'i+=1 i-=1 A=-~A%256 A=~-A%256 print(end=chrA) A=ord(open(0).read(1)) whileA: s '.split()['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
 exec(c.replace('A','(s[i])'))

[versed eagle]

brainfuck interpreter?

im pretty sure you have a bug
open(0).read(1) returns one character, not one byte
1 character can be multibyte sequences, meaning ordinals that cant fit in a single byte

[quartz wave]

just golfed here

maybe ask the source

[versed eagle]

ah

[quartz wave]

i also have a version while trying to golf that reads per byte

[versed eagle]

@final iron?

[quartz wave]

huh

[versed eagle]

oops wrong ping

skfnhfh sorry

@sick hound

there

[versed eagle]

good good

[quartz wave]

takes up +2b though

[sick hound]

Huh

[sick hound]

Ye

[versed eagle]

only 2 more bytes? how

,'rb' is 5

[quartz wave]

wait

i'm gonna have to do it again i lost it

[quartz wave]

it's actually +0b

s=[0]*30000
while(p:=next(f:=open(i:=0,'rb'))).lower()!=b'exit':
 c=''
 for o in p:c+=' '*i+'i+=1 i-=1 A=-~A%256 A=~-A%256 print(end=chrA) A=f.read(1) whileA: s '.split()[b'><+-.,['.find(o)]+'\n';i+=(92-o)*(o in b'][')
 exec(c.replace('A','(s[i])'))

[versed eagle]

oh

[quartz wave]

nvm

it should be +1b

since b'exit'!='exit'

[gleaming linden]

s=[0]*30000
while(p:=next(f:=open(i:=0,'rb'))).lower()!=b'exit':
 c=''
 for o in p:c+=' '*i+'i+=1 i-=1 A+=1 A-=1 print(end=chrA) A=f.read(1) whileA:# s '.split()[b'><+-.,['.find(o)]+';A%=256\n';i+=(92-o)*(o in b'][')
 exec(c.replace('A','(s[i])'))
``` does this work?

[quartz wave]

not sure

[dreamy pier]

Last two codes can't execute more than one line by me

    while(p:=next(f:=open(i:=0,'rb'))).lower()!=b'exit':
OSError: [Errno 9] Bad file descriptor

[arctic skiff]

huh that shouldn't happen

>>> open(0)
<_io.TextIOWrapper name=0 mode='r' encoding='utf-8'>
>>>  ```

[quartz wave]

works in my machine

[arctic skiff]

though what this code is?

[dreamy pier]

¯_(ツ)_/¯

++++++++++[>+++++++>++++++++++>+++>+<<<<-]>++.>+.+++++++..+++.>++.<<+++++++++++++++.>.+++.------.--------.

[arctic skiff]

your interpreter is having a stroke

[fleet bridge]

Oh, you guys created brainfuck REPL? 😄

[arctic skiff]

doesn't seems like it is a brainfuck repl

maybe it is

dunno

though why you are upside down? @fleet bridge

[fleet bridge]

Idk. Some guy said me to change my avatar, so i turned it upside down

[dreamy pier]

it is

[weak garden]

Anyone able to help with import hell?

root@1188a87ba730:/server# python data/processing/experiments/experiment_1/experiment.py
Traceback (most recent call last):
  File "/server/data/processing/experiments/experiment_1/experiment.py", line 1, in <module>
    from food_related_values import FoodRelatedValues
ModuleNotFoundError: No module named 'food_related_values'

Contents of file:

from food_related_values import FoodRelatedValues
from gpt_nlp_engine import GPTNLPEngine


def run_experiment(test_var):
    print('RUNNED', test_var)

[restive void]

I did a thing: https://gist.github.com/L3viathan/8bc92404f7084342175b95970a086b04
Usage:

from sumtypes import ST

class Result(ST):
    Ok[T]
    Err[E]


class Maybe(ST):
    Some[T]
    Not

res = Result.Err("hello")
nah = Maybe.Not

match res:
    case Result.Ok(value):
        print("got a result:", value)
    case Result.Err():
        print("got an error")

match nah:
    case Maybe.Not:
        print("not good")
    case Maybe.Some(value):
        print("some value:", value)

[floral spoke]

im guessing food_related_values is your own file youre trying to import?

[weak garden]

yeah, it's in the screenshot

[floral spoke]

yeah righto let me have a look

ok

[weak garden]

Tried all kinds of data.processing.translate and processing.translate

sorry food_related_values

instead of translate

[floral spoke]

yeah look my only idea is that it might be something to do with the file path and trying to reference python files multiple folders outside of the root program's folder? usually that isnt a problem but yknow hail maries and such

have you tried moving food_related_values into the root folder of experiment.py?

[weak garden]

that works

[floral spoke]

there ya go

[weak garden]

what if I don't want it there 👀

[floral spoke]

weird that that was the problem usually python can call to outside programs

erm usually when i reference outside files its in webdev and in that to escape the root folder you just go "../" before the filepath to go to the previous folder of the one that youre currently in

[weak garden]

When I try that I get attempted relative import with no known parent package

all those inits

for what

[restive void]

This is not a help channel, please continue discussion there, see #❓｜how-to-get-help

[floral spoke]

all goods

[weak garden]

ok sorry

[weak garden]

no ideas either? 👀

[zenith geode]

try from server.config.data.food_related_values import FoodRelatedValues

idk

oh sht its kinda necropost xd

[meager zinc]

iirc on windows each process is only allowed to have one file descriptor for each of stdout, stderr, and stdin but I might be misremembering

[fleet bridge]

import types
import fishhook

@fishhook.hook(types.FunctionType)
def __repr__(self: types.FunctionType) -> str:
    """
    def f():...
    lambda: ...
    """
    varnames = self.__code__.co_freevars
    varvalues = self.__closure__ or ()
    qualname = self.__qualname__

    if self.__module__ not in {'builtins', '__main__'}:
        qualname = f'{self.__module__}.{qualname}'

    if not varnames:
        return qualname

    return (
        '<'
        + ' '.join(
            [qualname] + [f'{n}={v.cell_contents}' for n, v in zip(varnames, varvalues)]
        )
        + '>'
    )

@fishhook.hook(types.BuiltinFunctionType)
@fishhook.hook(types.MethodWrapperType)
def __repr__(self: types.BuiltinFunctionType) -> str:
    """
    print # builtin function
    sys.exit # builtin function
    [].append # builtin method
    """
    obj = self.__self__
    if isinstance(obj, types.ModuleType):
        if obj.__name__ in {'builtins', '__main__'}:
            return self.__qualname__
        obj = obj.__name__
    else:
        obj = repr(obj)
    return f'{obj}.{self.__name__}'

@fishhook.hook(types.ModuleType)
def __repr__(self: types.ModuleType) -> str:
    """
    __main__ builtins
    sys
    """
    return self.__name__

@fishhook.hook(type)
def __repr__(self: type) -> str:
    """
    int enum.Enum X
    """
    if self.__module__ in {'builtins', '__main__'}:
        return self.__qualname__
    return f'{self.__module__}.{self.__qualname__}'

@fishhook.hook(types.MethodDescriptorType)
@fishhook.hook(types.WrapperDescriptorType)
def __repr__(self: types.MethodDescriptorType) -> str:
    """
    list.append # method descriptor
    object.__init__ # wrapper descriptor (slot wrapper)
    """
    mod = getattr(getattr(self, '__objclass__', None), '__module__', None)
    if mod in {None, 'builtins', '__main__'}:
        return self.__qualname__
    return f'{mod}.{self.__qualname__}'

``` UPD: added more overwrites

    varnames = self.__code__.co_freevars
    varvalues = self.__closure__ or ()

im not sure with this part
will this always be correct? i mean will these two sequences correspond to each other 1-to-1?

it seems to be working in my little tests, but i am not sure if this will work even in weird cases

[arctic skiff]

https://cdn.discordapp.com/emojis/852546249067593788.webp?size=48

[arctic skiff]

@narrow hound

[fleet bridge]

>>> {1:2,object:object(),(lambda:...):print}.keys.__call__
{1: 2, object: <object at 0x00199BCD250C0>, <lambda>: print}.keys.__call__
``` isn't this beautiful?

>>> class X(type): ...
...
>>> class X(X): ...
...
>>> class X(X, metaclass=X): ...
...
>>> X.__class__ = X
>>> assert X is type(X) is X.__class__
>>> X
Traceback (most recent call last):
  ...
RecursionError: maximum recursion depth exceeded while getting the str of an object
>>> 
``` 😭

>>> class X(type): ...
...
>>> X
X
>>> class X(X): ...
...
>>> X
X
>>> class X(X, metaclass=X): ...
...
>>> X
<X 'X'>
``` recursive repr :(

[narrow hound]

Damn

[ember vessel]

Code Golfed Snake

from pygame import*
from random import*
r=randrange
d=display.set_mode((z:=500,z))
m=Vector2
s=[m(0,0)]
v=1,0
f=m(1,1)
u=-1
j=50
while not event.get(256):
 h=s[u];d.fill(0)
 if(k:=key.get_pressed())[K_w]:v=0,u
 if k[K_s]:v=0,1
 if k[K_a]:v=u,0
 if k[K_d]:v=1,0
 for b in s+[f]:draw.circle(d,"red",b*10+(4,4),4)
 s+=[h+v];display.flip();time.wait(99)
 if h==f:f.xy=r(j),r(j)
 else:s.pop(0)

387 characters (including spaces and newlines)

(no u cant die)

(but i have a version that has that)

[restive void]

Anything you only use once is not worth re-assigning (e.g. randrange)

[quartz wave]

example used 2 times

[restive void]

Oh, right

[ember vessel]

?

[restive void]

Nevermind, I somehow read the first r as a different letter

[ember vessel]

lol

[quartz wave]

381 ```py
from pygame import*
from random import*
r=randrange
d=display.set_mode((500,500))
m=Vector2
s=[m()]
v=1,0
f=m(1,1)
u=-1
j=50
while not event.get(256):
h=s[u];d.fill(0);k=key.get_pressed()
if k[K_w]:v=0,u
if k[K_s]:v=0,1
if k[K_a]:v=u,0
if k[K_d]:v=1,0
for b in s+[f]:draw.circle(d,"red",b*10+(4,4),4)
s+=h+v,;display.flip();time.wait(99)
if h==f:f.xy=r(j),r(j)
else:_,*s=s

[ember vessel]

what did u change?

oh

the last line

huh

what does the do

it look invalid...but it works

[quartz wave]

the what?

[ember vessel]

else:_,*s=s

under scroe if for var assignment...

ohhhh

i get it

ur splitting s into two parts

why?

oh

[restive void]

It's just doing s.pop(0)

[ember vessel]

is that a way to pop the first element

[ember vessel]

yeah, got it

[quartz wave]

can it just be d.flip()?

[ember vessel]

wdym

[restive void]

!e

x = [1, 2, 3, 4]
_, *x = x
print(x)

[night quarry]

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

[2, 3, 4]

[quartz wave]

nvm

[ember vessel]

on the pygame discord server, we had a long line of versions

i wrote the above one

i think imma try to make one from scratch using the sketchiest techniques possible

[restive void]

You should be able to do something like

v={K_w:(0,u),K_s:(0,1),K_a:(u,0),K_d:(1,0)}[k.index(1)]

instead of all the ifs

[ember vessel]

from pygame import*
d=display
s=d.set_mode()
while not event.get(256):
    s.fill(0)
    d.flip()

Shortest start code possible (i think)

its even shorter than before

[restive void]

And depending on the actual value of K_*, using the int directly might be shorter

[ember vessel]

wait, give me an example with the keypresses

[restive void]

I'm pretty sure K_w etc. are just ints

[ember vessel]

oh, i get it now, i think

they are

but why k.index(1)

[restive void]

k is a sequence of bools, stating whether the key at that index is pressed. k.index(1) finds the first index where it's True.

[ember vessel]

ah

[quartz wave]

what if b is pressed

[restive void]

Undefined behavior? More critical would be the case where nothing is pressed :D

[ember vessel]

yeah

i just tested

i think i have a better idea

[quartz wave]

hmm true

[restive void]

Can add [1] to the list

[ember vessel]

ye, cuz the index isnt in the tuple

[quartz wave]

371 ```py
from pygame import*
from random import*
r=randrange
t=display
d=t.set_mode((500,500))
m=Vector2
s=m(),
v=1,0
f=m(1,1)
while not event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed()
if k[K_w]:v=0,-1
if k[K_s]:v=0,1
if k[K_a]:v=-1,0
if k[K_d]:v=1,0
for b ins,f:draw.circle(d,"red",b10+(4,4),4)
s+=h+v,;t.flip();time.wait(99)
if h==f:f.xy=r(50),r(50)
else:_,*s=s

[quartz wave]

this does not work for whatever reason

handling the cases where no key is pressed

[restive void]

You'd need to do ...[(k+[1]).index(1)], plus add another case in the dict, but then it won't save any bytes anymore

[quartz wave]

you see i found a way to still get -3b

i think

[ember vessel]

from pygame import*
d=display
s=d.set_mode()
u=-1
f=1,0
v=f
e=event.get
while not e(256):
    s.fill(0)
    k=e(768)
    if k:v=f if(k:=k[0].key)==K_d else(u,0)if k==K_a else(0,1)if k==K_w else(0,u)
    d.flip()
```?

it handles controls now

oh

forgot to code golf

there

[quartz wave]

even if that is done it still doesn't work

[ember vessel]

Did you guys get the controls script to work? @restive void @quartz wave

mine is way to long

nope

still to long :(

hows this real?

from pygame import*
from random import*
r=randrange
q=display
d=q.set_mode()
m=Vector2
s=[m()]
v=1,0
f=m(v)
u=-1
j=50
while not event.get(256):
 h=s[u];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,u
 if k[K_s]:v=0,1
 if k[K_a]:v=u,0
 if k[K_d]:v=1,0
 for b in s+[f]:draw.circle(d,"red",b*10+(4,4),4)
 s+=h+v,;q.flip();time.wait(99)
 if h==f:f.xy=r(j),r(j)
 else:_,*s=s

368 chars

[quartz wave]

360 ```py
from pygame import*
from random import*
r=randrange
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed()
if k[K_w]:v=0,-1
if k[K_s]:v=0,1
if k[K_a]:v=-1,0
if k[K_d]:v=1,0
for b ins,f:draw.circle(d,"red",b10+(4,4),4)
s+=h+v,;t.flip();time.wait(99)
if h==f:f.xy=r(50),r(50)
else:_,*s=s

[ember vessel]

why did u remove u=-1?

[quartz wave]

no benefit

[ember vessel]

yes

there is

the u= is two chars

[quartz wave]

u=-1;uuu
-1-1-1

no benefit

semicolon is supposed to represent a newline

[ember vessel]

hmmmm

ur right

did u add any other changes?

[quartz wave]

sort of merged the 371 with the 368

[ember vessel]

?

[quartz wave]

removed j ```
j=50;jj
5050

[ember vessel]

ah

[quartz wave]

these ```diff

• s=[m()]

• s=m(),

• ...for b in s+[f]...

• ...for b in*s,f:...

[ember vessel]

357 chars!

from pygame import*
from random import*
r=randrange
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
 h=s[-1];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,-1
 if k[K_s]:v=0,1
 if k[K_a]:v=-1,0
 if k[K_d]:v=1,0
 for b in*s,f:draw.circle(d,99,b*10+(4,4),4)
 s+=h+v,;t.flip();time.wait(99)
 if h==f:f.xy=r(50),r(50)
 else:_,*s=s

this is low contrast mode

higher contrast takes one more character

use this to count chars

https://www.easywordcounter.com/

and click "include spaces"

[finite blaze]

how did u color the text like that?

[quartz wave]

these ```diff
- s=[m()]

• s=m(),
  - ...for b in s+[f]...
• ...for b in*s,f:...
  ```

[finite blaze]

thanks 👍

[ember vessel]

@quartz wave , i added some int he og version to make it look like nicer

but i can remove those

351 chars

from pygame import*
from random import*
r=randrange
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
 h=s[-1];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,-1
 if k[K_s]:v=0,1
 if k[K_a]:v=-1,0
 if k[K_d]:v=1,0
 for b in*s,f:draw.circle(d,255,b*9,4)
 s+=h+v,;t.flip();time.wait(99)
 if h==f:f.xy=r(50),r(50)
 else:_,*s=s

and this is high contrast

i can do 350

[dreamy pier]

changing key to integer -1

[ember vessel]

?

[quartz wave]

ok get as low as you can

[ember vessel]

well

[dreamy pier]

from pygame import *
print(K_w, K_s, K_a, K_d)

119 115 97 100

[ember vessel]

low-contrast is pretty hard to see

[dreamy pier]

K_a can be 97

[ember vessel]

but ultra-low is impossible

[ember vessel]

so which one?

[ember vessel]

ahh

[quartz wave]

playable and as low as possible

[ember vessel]

can you do that?

[dreamy pier]

sure

[ember vessel]

OMG

i can do white

and its high contrast

[dreamy pier]

What are rules of moving? Keep them the same as they are here?

[ember vessel]

@dreamy pier , i got it

[ember vessel]

this is normal snake

it should still be snake

[dreamy pier]

Like i can make it shorter, but it will move only when pressed and can move diagonally 😄

[ember vessel]

from pygame import*
from random import*
r=randrange
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
 h=s[-1];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,-1
 if k[K_s]:v=0,1
 if k[97]:v=-1,0
 if k[K_d]:v=1,0
 for b in*s,f:draw.circle(d,-1,b*9,4)
 s+=h+v,;t.flip();time.wait(99)
 if h==f:f.xy=r(50),r(50)
 else:_,*s=s

349

my first draft was like 530

then on the pygame server, a bunch of ppl helped make it lower

[ember vessel]

nah, don't do it

thats not rlly snake

[dreamy pier]

actually diagonal moving looks very cool

but yeah

[ember vessel]

i can make the apple spawn in just a 9x9 grid for 2 less chars

but that's not fun

[dreamy pier]

choices -1

from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
 h=s[-1];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,-1
 if k[K_s]:v=0,1
 if k[97]:v=-1,0
 if k[K_d]:v=1,0
 for b in*s,f:draw.circle(d,-1,b*9,4)
 s+=h+v,;t.flip();time.wait(99)
 if h==f:f.xy=choices(range(50),k=2)
 else:_,*s=s

[ember vessel]

isn't that longer

oh

[dreamy pier]

348

[ember vessel]

i see

yesss

now that its in full screen

lets up 50 to 99

from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
 h=s[-1];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,-1
 if k[K_s]:v=0,1
 if k[97]:v=-1,0
 if k[K_d]:v=1,0
 for b in*s,f:draw.circle(d,-1,b*9,4)
 s+=h+v,;t.flip();time.wait(99)
 if h==f:f.xy=choices(range(99),k=2)
 else:_,*s=s

same chars

[dreamy pier]

from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
 h=s[-1];d.fill(0);k=key.get_pressed()
 if k[K_w]:v=0,-1
 if k[K_s]:v=0,1
 if k[97]:v=-1,0
 if k[K_d]:v=1,0
 for b in*s,f:draw.circle(d,-1,b*9,4)
 s+=h+v,;t.flip();time.wait(99)
 if h-f:_,*s=s
 else:f.xy=choices(range(99),k=2)

347

[ember vessel]

thats clever

the diff

[quartz wave]

331 ```py
from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed();exec('if k[%d]:v=%d-2,%d-2\n'4%(b'sw
a
d',))
for b ins,f:draw.circle(d,-1,b9,4)
s+=h+v,;t.flip();time.wait(99)
if h-f:_,*s=s
else:f.xy=choices(range(99),k=2)

[ember vessel]

WHAT

[quartz wave]

those question mark boxes are control characters

[ember vessel]

huh

[quartz wave]

technically valid as 1 byte since they're part of ASCII

[ember vessel]

explain what the exec is doing plz

and whered the keys go

[quartz wave]

just like this```py
if k[K_s]:v=0,1
if k[K_w]:v=0,-1
if k[97]:v=-1,0
if k[K_d]:v=1,0

[ember vessel]

but weird...

[quartz wave]

yeah

[ember vessel]

howd you do it

[dreamy pier]

isn't that 330?

[quartz wave]

[K_s, K_w, K_a, K_d] is technically [*b'swad']

[ember vessel]

the k?

[quartz wave]

did you do len("""...""")

because there's a newline escape in there

[dreamy pier]

ah, ye, 331

[ember vessel]

no use https://www.easywordcounter.com/

if you want

cuz its rlly easy

[quartz wave]

the keys correspond to their ascii equivalents

[ember vessel]

oh, ok...

[quartz wave]

and a quick n easy way to get those is by doing b'swad'

if that is iterated over, it'll return the value of those keys

[ember vessel]

did u write that by hand?

[quartz wave]

yeah

well i had to do print('\2\2\1\3') n stuff but other than that yeah

[ember vessel]

k, thats rlly cool

(im posting all the codes that are being sent in the pygame discord, if you don't mind)

[dreamy pier]

191 then

exec(bytes('牦浯瀠杹浡⁥浩潰瑲ਪ牦浯爠湡潤浩潰瑲ਪ㵴楤灳慬੹㵤⹴敳彴潭敤⤨洊嘽捥潴㉲猊洽⤨ਬ㵶ⰱਰ㵦⡭⥶眊楨敬渠瑯攠敶瑮朮瑥㈨㘵㨩 㵨孳ㄭ㭝⹤楦汬〨㬩㵫敫⹹敧彴牰獥敳⡤㬩硥捥✨晩欠╛嵤瘺┽ⵤⰲ搥㈭湜⨧┴⨨❢ɳ眃Ăš搂ȃⰧ⤩ 潦⁲⁢湩猪昬携慲⹷楣捲敬搨⴬ⰱ⩢ⰹ⤴ ⭳栽瘫㬬⹴汦灩⤨琻浩⹥慷瑩㤨⤹ 晩栠昭强⨬㵳ੳ攠獬㩥⹦祸挽潨捩獥爨湡敧㤨⤹欬㈽
','u16')[2:])

[quartz wave]

191c

[ember vessel]

HUH

[quartz wave]

519b

[dreamy pier]

depends on what we do

[ember vessel]

you definetly generated that

[dreamy pier]

sure

https://clemg.github.io/pythongolfer/

[ember vessel]

ok, whoa

@quartz wave , is that where you leanred it?

[quartz wave]

learned what?

[ember vessel]

this technique

[quartz wave]

from where?

[ember vessel]

here

[quartz wave]

no

the chinese character thingy is a common tool

[ember vessel]

i don't feel like these two should be counted though

cuz its not even python

[quartz wave]

used to generate the shortest character solutions

but they're very expensive in terms of bytes

[ember vessel]

its binary

[quartz wave]

so generating that results in a 519b, 191c solution

[ember vessel]

ok

[quartz wave]

341b, 341c using escapes only ```py
from pygame import*
from random import*
r=randrange
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed();exec('if k[%d]:v=%d-2,%d-2\n'4%(b's\2\3w\2\1a\1\2d\3\2',))
for b ins,f:draw.circle(d,-1,b9,4)
s+=h+v,;t.flip();time.wait(99)
if h==f:f.xy=r(50),r(50)
else:_,*s=s

[ember vessel]

still, its not even python

[quartz wave]

it is python

[ember vessel]

fine

but this is way to hacky

even for my standards

[dreamy pier]

xd

[quartz wave]

that's code golfing for you

[ember vessel]

lol, yeah i just started

so all those code golfing jams, does everyone just use the shortest-char thing

[quartz wave]

no there are people who consider that cheating

so usually bytes is counted

[ember vessel]

ok, yeah, same

[quartz wave]

turns out i didn't use the latest golfed version before the exec()
here's the 339b, 339c from the 347 ```py
from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while not event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed();exec('if k[%d]:v=%d-2,%d-2\n'4%(b's\2\3w\2\1a\1\2d\3\2',))
for b ins,f:draw.circle(d,-1,b9,4)
s+=h+v,;t.flip();time.wait(99)
if h-f:_,*s=s
else:f.xy=choices(range(99),k=2)

338b, 338c py from pygame import* from random import* t=display d=t.set_mode() m=Vector2 s=m(), v=1,0 f=m(v) while[1]>event.get(256): h=s[-1];d.fill(0);k=key.get_pressed();exec('if k[%d]:v=%d-2,%d-2\n'*4%(*b's\2\3w\2\1a\1\2d\3\2',)) for b in*s,f:draw.circle(d,-1,b*9,4) s+=h+v,;t.flip();time.wait(99) if h-f:_,*s=s else:f.xy=choices(range(99),k=2) 330b, 330c ```py
from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while[1]>event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed();exec('if k[%d]:v=%d-2,%d-2\n'4%(b'sw
a
d',))
for b ins,f:draw.circle(d,-1,b9,4)
s+=h+v,;t.flip();time.wait(99)
if h-f:_,*s=s
else:f.xy=choices(range(99),k=2)

- while not event.get(256):
+ while[1]>event.get(256):

[ember vessel]

i still kinda don't like this thing. Your right that it is python, but still

[quartz wave]

picked up from the 347 n made it a 346 ```py
from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while[1]>event.get(256):
h=s[-1];d.fill(0);k=key.get_pressed()
if k[K_w]:v=0,-1
if k[K_s]:v=0,1
if k[97]:v=-1,0
if k[K_d]:v=1,0
for b ins,f:draw.circle(d,-1,b9,4)
s+=h+v,;t.flip();time.wait(99)
if h-f:_,*s=s
else:f.xy=choices(range(99),k=2)

if that's what you like

[ember vessel]

yes, good

i think that is where the challenge is

oh, thats good

[proper vault]

~~```py
from pygame import*
from random import*
t=display
d=t.set_mode()
m=Vector2
s=m(),
v=1,0
f=m(v)
while[1]>event.get(256):
h=s[-1];d.fill(0)
for u in key.get_pressed().values():v=dict(w=(0,-1),s=(0,1),a=(-1,0),d=(1,0)).get(u,v)
for b ins,f:draw.circle(d,-1,b9,4)
s+=h+v,;t.flip();time.wait(99)
if h-f:_,*s=s
else:f.xy=choices(range(99),k=2)

ah nvm, doesn't work

[unreal echo]

overload decorator
https://paste.pythondiscord.com/TTBA

how do i make @night quarry execute a code that is too big to send here, when i paste it it just comes up at message.txt

whoops recursion error

didnt know cache could do that

[fleet bridge]

!e ```py
data = '789c2b28cacc2bd150cf48cdc9c95728cf2fca4951d704004ee30729'
import zlib
exec(zlib.decompress(bytes.fromhex(data)))

[night quarry]

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

hello world

[unreal echo]

no i will not golf it

someone else can try that, but not me

[fleet bridge]

i am not talking about golfing

just compress your code

[unreal echo]

hm

maybe

hm still too big

[fleet bridge]

lmao

[unreal echo]

what

[fleet bridge]

you know this does literally nothing?

global x has no runtime effect

so it cannot raise exception

try:
    pass
except WtfError:
    ...
``` this is your code

[unreal echo]

whoops

its supposed to be try:__overloads__except:

[fleet bridge]

still does nothing, because __overloads__ is always defined

[unreal echo]

its supposed to be a failsafe in case a user tries to delete the overloads

so that it doesnt completely break

just removes previous overloads

[fleet bridge]

you shouldnt worry about that case

[unreal echo]

https://paste.pythondiscord.com/TTBA

[unreal echo]

i did anyway

[unreal echo]

to be fair this could possibly run into a MemoryError if a gamma particle hits a bit in my memory at a specific time

[fleet bridge]

KeyboardInterrupt is far more frequent than gamma particles

[calm elk]

https://blog.adarshd.dev/posts/everything-is-an-object/

[dreamy stone]

from functools import partial
from typing import Iterable, Callable
class Seq:
    _seq = []
    chained = []
    def __init__(self, i: Iterable) -> None:
        self._seq = i
    def filter(self, func: Callable):
        self.chained.append(partial(filter, func))
        return self
    def map(self, func: Callable):
        self.chained.append(partial(map, func))
        return self
    def to_list(self):
        for func in self.chained:
            self._seq = func(self._seq)
        return list(self._seq)
def seq(i: Iterable):
    return Seq(i)
l = [1, 2, 3, "4", "a", "b", 7, "8"]
r = seq(l).map(lambda x: str(x)).filter(lambda x: x.isdigit()).to_list()
print(r)

I wish Python had something like this built in, or at least a way to chain generator functions without hurting your eyes.

[dreamy stone]

mutable defaults not even once

[fleet bridge]

lambda x: str(x) is just str
lambda x: x.is_digit() is just str.is_digit

[dreamy stone]

d'oh

[wide crow]

What about this?

l = [1, 2, 3, "4", "a", "b", 7, "8"]
r = [*filter(str.isdigit, map(str, l))]
print(r)

[fleet bridge]

Why do you have seq function that just instantiates Seq class? Just remove fjnction and rename class to seq

[wide crow]

Or even this?

l = [1, 2, 3, "4", "a", "b", 7, "8"]
r = [s for x in l if (s := str(x)).isdigit()]
print(r)

[low lynx]

!d iters

oh wait

[low lynx]

!pypi iters

[night quarry]

iters v0.13.1

Composable external iteration.

[low lynx]

there we go

[wide crow]

This fails on the "built-in" requirement.

[low lynx]

yeah but I thought they might be interested regardless

[wide crow]

Perhaps this is the kind of thing that you were trying to do.

class compose(list):
    def map(self, func):
        return compose(map(func, self))
    def filter(self, func):
        return compose(filter(func, self))

data = [1, 2, 3, '4', 'a', 'b', 7, '8']
f = lambda x: str(x)
g = lambda x: x.isdigit()
r = compose(data).map(f).filter(g)
print(r)

And then you can easily extend the class to include other features.

[dreamy stone]

that's a tight solution

the str functions are not important. they're just what I used for an example. yeah I could have just named the class lowercase, I was being specific about style guidelines.

[dreamy stone]

Thanks!

[gleaming linden]

I'm pretty sure there's something like using snake_case if the intended usecase is similar to a function in pep8

[restive void]

Yes:

However, know when to be inconsistent – sometimes style guide recommendations just aren’t applicable. When in doubt, use your best judgment. Look at other examples and decide what looks best.

[rugged sparrow]

!e ```py
from fishhook import hook

@hook.cls(list)
class list_hooks:
def map(self, func):
return list(map(func, self))

def filter(self, func):
return list(filter(func, self))

data = [1, 2, 3, '4', 'a', 'b', 7, '8']
f = lambda x: str(x)
g = lambda x: x.isdigit()
r = data.map(f).filter(g)
print(r)```

[night quarry]

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

['1', '2', '3', '4', '7', '8']

[pearl socket]

!e

from fishhook import hook

@hook.cls(list)
class list_hooks:
  def map(self, func):
    return list(map(func, self))

  def filter(self, func):
    return list(filter(func, self))

data = [1, 2, 3, '4', 'a', 'b', 7, '8']
print(data.map(str).filter(str.isdigit))

[night quarry]

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

['1', '2', '3', '4', '7', '8']

[sonic birch]

I see more and more fishhook used in this chat. The previous superhero here was einspect. Also somebody using ctypes directly to hook some system behaviour. Which way is better?

[languid hare]

the previous previous superhero was fishhook :p

i never learnt how to use einspect fully (sorry ionite)

[fleet bridge]

When type annotations and py.typed marker in fishhook?

[restive void]

from fishhook import hook

@hook
def map(self: list, func):
    ...

like that?

[proper vault]

The real og is forbidden fruit.

[rugged sparrow]

i think the order was forbidden fruit, using ctypes directly, then fishhook, then einspect

but fishhook and einspect are slightly different in how they work

(at least in regards to hooking, einspect has a ton of introspection features that i didnt add to fishhook)

and fishhook has fishhook.asm for hooking C functions that einspect doesnt (i think?) have

!e fishhook.asm lets you do stuff like this to leak internal values and other way more in-depth control flow modification (this code grabs a reference to the interned strings dict)```py
from fishhook.asm import hook
from ctypes import *

@hook(pythonapi.PyDict_SetDefault, restype=py_object, argtypes=[py_object]*3)
def setdefault(self, key, value):
if key == 'MAGICVAL':
return self
return pythonapi.PyDict_SetDefault(self, key, value)

pythonapi.PyUnicode_InternFromString.restype = py_object
interned = pythonapi.PyUnicode_InternFromString(b'MAGICVAL')
setdefault.unhook()
print(interned)

[night quarry]

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

{'type': 'type', 'AttributeError': 'AttributeError', '__qualname__': '__qualname__', 'obj': 'obj', 'update': 'update', '__dict__': '__dict__', 'getattr': 'getattr', 'setattr': 'setattr', 'hasattr': 'hasattr', '__doc__': '__doc__', '__name__': '__name__', '__module__': '__module__', 'replace': 'replace', 'old': 'old', 'new': 'new', 'sys': 'sys', 'name': 'name', '_DeadlockError': '_DeadlockError', 'waiters': 'waiters', 'count': 'count', 'owner': 'owner', 'wakeup': 'wakeup', 'lock': 'lock', 'allocate_lock': 'allocate_lock', '_thread': '_thread', 'self': 'self', 'add': 'add', 'get': 'get', '_blocking_on': '_blocking_on', 'set': 'set', 'get_ident': 'get_ident', 'seen': 'seen', 'tid': 'tid', 'me': 'me', 'release': 'release', 'acquire': 'acquire', 'has_deadlock': 'has_deadlock', 'RuntimeError': 'RuntimeError', 'id': 'id', 'format': 'format', '__repr__': '__repr__', '__init__': '__init__', '_ModuleLock': '_ModuleLock', '_DummyModuleLock': '_DummyModuleLock', '_lock': '_lock', '_name': '_name',
... (truncated - too long)

Full output: https://paste.pythondiscord.com/T375LWEFNTIRUGVFC5BPRPNYIM

[rugged sparrow]

!e ```py
from fishhook.asm import hook
from ctypes import *

@hook(pythonapi.PyDict_SetDefault, restype=py_object, argtypes=[py_object]*3)
def setdefault(self, key, value):
if key == 'MAGICVAL':
return self
return pythonapi.PyDict_SetDefault(self, key, value)

pythonapi.PyUnicode_InternFromString.restype = py_object
interned = pythonapi.PyUnicode_InternFromString(b'MAGICVAL')
setdefault.unhook()
interned['chilaxan'] = 'foo'
print(eval("'chilaxan'"))```

[night quarry]

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

foo

[rugged sparrow]

interned dict lets you muck around with some odd stuff

[fleet bridge]

I am talking sbout type annotations in fishhook itself

[rugged sparrow]

when i have time to revamp it for 3.12 probably

but i also want to rewrite fishhook.asm to be able to give access to different registers and do stuff like instruction injection point matching

like ```py
@hook(pythonapi.Some_C_Function, before='pop rax;')
def injected(rax: py_object):
...
return Before('pop rax;', rax=rax) # this would be the default to make hooks happen inline, but having it dynamic like this would allow for really cool tricks

[sonic birch]

What about this?
!e

import sys
import builtins
from functools import partial as part
from collections import ChainMap

class arg:

  def __init__(self, data):
    self.data = data

  def __getattr__(self, key):
    f = ChainMap(builtins.__dict__, sys._getframe(1).f_globals)[key]
    return lambda *args, **kwargs: arg(part(f, *args, **kwargs)(self.data))

(
  arg([1,2,'3','4','b',5])
  .map(str)
  .filter(str.isdigit)
  .map(int)
  .list()
  .data
)

cc: @dreamy stone

[sonic birch]

Why it is not running?

[astral rover]

you need to put !e first

[night quarry]

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

[1, 2, 3, 4, 5]

[dreamy stone]

too readable

[tough willow]

Thats actually pretty useful

[burnt gust]

look at what i made 8 years ago when i was still sorta new to python

stats = '**-Stats for {} members-**\n'.format(message.server.member_count)+'\n'.join(['`{:.<6}{:.>17}`'.format(len([m for m in message.server.members if r in m.roles]), r.name) for r in sorted(message.server.roles, key=lambda role: role.position, reverse=True) if r != message.server.default_role])

[wide crow]

And after a little tidying up it becomes:

!e

from functools import partial

class arg(list):
    def __getattr__(self, key):
        f = vars(__builtins__)[key]
        return lambda *args: arg(partial(f, *args)(self))

a = [1,2,'3','4','b',5]
r = (
    arg(a)
    .map(str)
    .filter(str.isdigit)
    .map(int)
    )
print(r)

[night quarry]

@wide crow :white_check_mark: Your 3.11 eval job has completed with return code 0.

[1, 2, 3, 4, 5]

[sonic birch]

Looks elegant! But it will not work with custom functions defined in the global scope. Also it will not be able to work with data other than sequences. But, yeah - it depends on use-case

UDP: I meant this will not be possible with your approach

def add_one(x): return x+1

arg('5.76').float().round().add_one().data

[golden finch]

I've been trying to teach myself a little bit of "functional" (or at least, as functions-as-objects as Python allows) stuff here and there, explicitly for the purposes of obfuscation, and I've hit what seems to me to be something vaguely interesting. It's certainly given me one hell of a headache - and I hope whoever tries to deobfuscate it gets one of a similar magnitude.

So I had the following:

def a(x):
    return lambda y:x

Straightforward enough...

So I was toying around, and I thought that the following would behave the same:

@lambda _b:lambda x:lambda y:_b(y)
def b(y):
    return x

That throws a NameError, for reasons I don't understand.
I then looked at the lambda y:_b(y) part, and realised that's redundant (it's just _b, obviously), so I rewrote it to be a bit more palatable, like so:

@lambda _b:lambda x:_b
def b(y):
    return x

However, that still throws a NameError (x is not defined). By my understanding, when the final version of b(n) is called, x=n, and in the scope of the "original" version, x should be defined. Evidently not, though.

Where's the mistake in my logic here, and any ideas on how I can fix it?

(the goal is to produce a decorator for the function b that makes it behave just like a)

[fleet bridge]

x in lambda args and x in finction are two independent variables.
So first x is unused, and second is used but never assigned, so it raises NameError

[golden finch]

tysm mate

[mental vector]

Does anyone have any interesting resources or articles/blog posts on writing polyglots for other languages in Python that you could share? The other language I'm targeting is C++, but any language would be interesting to read about. I'm just looking for interesting tricks you can use to twist the python syntax to work with other stuff. While anything would be useful, I want to try something other than the usual c/c++ polyglot method of just spamming #defines everywhere to make C++ syntax Pythonic, because I think it's a lot more challenging to make Python C++-like.

[finite blaze]

Hey, how would you return value in a custom language without using "return"?

I'm creating a small interpreted language

and i thought u might have some cool ideas

[mental vector]

Rust has it so that if a function ends in an expression instead of a statement (so, no semicolon), the expression is returned

[finite blaze]

thats boring ;p

[fleet bridge]

A lot of other languages do the same