esoteric-python

[low lynx]

i mean the golf of max(arr,key=len) is not too far from a serious answer, but yeah

[bitter granite]

Yes, I was looking for a serious answer as I'm thinking on how to complete a piece of work of mine 😄

[bitter granite]

Sorry to ask, butwhat is the golf in this case?

[low lynx]

oh i just meant that if you try to code golf it (shortest number of characters/bytes), you'd end up with max(arr,key=len), which is what a serious answer would probably give you as well

plus or minus a space

[bitter granite]

so, in essence, a full iteration is still needed, either by the time the array is built, or at tergo

there's no built in approach in place

[meager zinc]

it depends how your list is structured.

1. if unsorted, you must go through all elements at some point (you could use a tuple for a 5% theoretical speedup in indexing/looping)
2. if sorted, you could use a binary search which would be much faster

[bitter granite]

My list is made of dictionary like this:

max(union, key=len)
{'inspectionResultLink': 'https://search.googl...source=api', 'indexStatusResult.verdict': 'PASS', 'indexStatusResult.co...erageState': 'Submitted and indexed', 'indexStatusResult.ro...tsTxtState': 'ALLOWED', 'indexStatusResult.in...exingState': 'INDEXING_ALLOWED', 'indexStatusResult.la...tCrawlTime': '2023-11-05T21:24:26Z', 'indexStatusResult.pa...FetchState': 'SUCCESSFUL', 'indexStatusResult.go...eCanonical': 'https://www.url.com', 'indexStatusResult.us...rCanonical': 'https://www.url.net/', 'indexStatusResult.re...ringUrls.0': 'https://www.url...nomia-seo/', 'indexStatusResult.re...ringUrls.1': 'https://www.url...essionale/', 'indexStatusResult.crawledAs': 'MOBILE', 'mobileUsabilityResult.verdict': 'PASS', 'richResultsResult.verdict': 'PASS', ...}

And it's not sorted. I don't see how I can sort for something that has not an "index" per se.

A max(list, key=len) works fine for few elements ... but will that be the case for 25K rows?

At end, what I need to build is a namedTuple collection for fast querying the above items.
And I need to find the dictionary with the longest key set at first.
Although, while writing I can alreay smell the subsequent problem ... those dictionaries where the keys was not in place will thrown an error for the missing key

[bitter granite]

Or perhaps I should consider a SimpleNamespace instead of complicating my life with the namedTuple

In the end the result will be the same, a on object that can be queried by properties

I appreciate the fact that namedTuple will be immutable and guarantee a structure, but my class goal is to facilitate the reading of data not the saving

[serene stratus]

is it possible to get the code object of a module that's already imported?

like the main code object

not from functions or anything

[proper vault]

Ye, should be the f_code on the appropriate frame

[rugged sparrow]

!e ```py
def ReadAfterFree(size, do):
b = bytearray(size)
class T:
def iter(self):
b.clear()
self.v = do()
yield b''
yield b''

c = b.join(t:=T())
return memoryview(c).cast('P'), t.v

leak, obj = ReadAfterFree(bytearray.basicsize, lambda: bytearray(8))
print('bytearray:', obj)
print('leaked memory of buffer:', leak.tolist())``` read after free in stringlib join

[night quarry]

@rugged sparrow :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | bytearray: bytearray(b'\x00\x00\x00\x00\x00\x00\x00\x00')
002 | leaked memory of buffer: [1, 140216309412032, 8, 9, 140216299095120, 140216299095120, 0]

[serene stratus]

Yeah but how do I access the frame?

[proper vault]

Ah, if you don't have it on the call stack, that would be a bit difficult. I would expect importlib to have it handy

If not, you can call compile on its source code

[long fulcrum]

how can i golf this more?

from itertools import*
E=print
D=str
A=sorted
C=lambda n:A(set(B*C for(B,C)in combinations(range(10**(n//2-1),10**(n//2)),2)if B%10+C%10and A(D(B*C))==A(D(B)+D(C))))
for B in C(4):E(B)
for B in C(6):E(B)```

[serene stratus]

I'll check importlib but if that doesn't work I'm stuck cuz there's no other way to access the code through files etc so I'd need to find a different way

[gleaming linden]

from itertools import*
D=str
A=sorted
for n in 4,6:*map(print,A(set(B*C for(B,C)in combinations(range(10**(n//2-1),10**(n//2)),2)if B%10+C%10and A(D(B*C))==A(D(B)+D(C)))),
```does this work?

from itertools import*
D=str
A=sorted
for n in 1,2:E=10**n;*map(print,A({B*C for(B,C)in combinations(range(E,E*10),2)if B%10+C%10if A(D(B*C))==A(D(B)+D(C))})),

[long fulcrum]

yeah that works

howd you get rid of the //2

[gleaming linden]

for n in 4,6 -> for n in 1,2

[long fulcrum]

can this be written any shorter for i in range(50,0,-1):

[vast wave]

depends on the code inside of the for

[orchid nymph]

!e

from itertools import*
D=str
A=sorted
for n in 1,2:E=10**n;*map(print,A({B*C for(B,C)in combinations(range(E,E*10),2)if B%10+C%10if A(D(B*C))==A(D(B)+D(C))})),

[night quarry]

@orchid nymph :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | /home/main.py:4: SyntaxWarning: invalid decimal literal
002 |   for n in 1,2:E=10**n;*map(print,A({B*C for(B,C)in combinations(range(E,E*10),2)if B%10+C%10if A(D(B*C))==A(D(B)+D(C))})),
003 | 1260
004 | 1395
005 | 1435
006 | 1530
007 | 1827
008 | 2187
009 | 6880
010 | 102510
011 | 104260
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/MN4TQLX5R224ENYDRSQ5QL3PGU

[orchid nymph]

n=50
while n:n-=1

[orchid nymph]

from itertools import*
A=sorted
for n in 10,100:*map(print,A({B*C for(B,C)in combinations(range(n,n*10),2)if B%10+C%10if A(f"{B*C}")==A(f"{B}{C}")})),

well its a code.golf hole so I don't think I can help more than that

https://code.golf/vampire-numbers#python

[bitter granite]

Hello there,
Is there any way to cast the value contained in dictionary's value to the corresponding enum value description?
Doing in a way that performance are not affected given I have several of them in my dictionary included in a list ?

Imagine something like the following

{'verdict': 'PASS', 'coverageState': 'Submitted and indexed', 'robotsTxtState': 'ALLOWED', 'indexingState': 'INDEXING_ALLOWED', 'lastCrawlTime': '2023-12-02T18:07:42Z', 'pageFetchState': 'SUCCESSFUL', 'googleCanonical': 'https://www.url.com', 'userCanonical': 'https://www.url.com/', 'crawledAs': 'MOBILE'}

I need to replace the indexingState, the pageFetchState, the robotsTxtState.

Ideally I should do this when I'm building my object.

[versed eagle]

can you not just use the enum value directly

also, this is probably the wrong channel

[bitter granite]

What could be the right channel? I'm still struggling to understand the 'esotheric' ssense, sorry

[versed eagle]

#❓｜how-to-get-help

[modest bough]

I just learned that [[]] * n yields shared references to the same list. I tried using list() or [][:] but couldn't find a way to use the ... * n to have n different empty lists

[earnest wing]

[*map(list,[()]*n)]

[earnest wing]

[[]for()in[()]*n]

for fun

[modest bough]

ok, nothing cuter than [[] for _ in range(n)] then

[thorn vale]

for _ in range(int(input())):
n = int(input()) - 1
arr = [i for i in list(input())]
index_list = []
if arr[:-1].count("A") > 0:
count = 0
while n > 0:
if (arr[n] == "B") and (arr[n - 1] == "A") and (n not in index_list):
index_list.append(n)
arr[n], arr[n - 1] = arr[n - 1], arr[n]
n = int(len(arr))-1
count += 1
else:
n -= 1
print(count)
else:
print("0")

Can anyone help me with this please?

Im a beginner. When I submit the code it says Time limit exceeded

The problem is that every time the action is performed, It goes back to the end of the array and starts checking again which is utterly inefficient

But I dont know how to make it faster

[modest bough]

i can't read your assignment, but if you focus on finding a criterion that can avoid scanning the array fully each time you win

[thorn vale]

I sort of did and it was +1 hour of thinking why doesnt it work this time

so I thought, enough for today ahahah

[modest bough]

sleep based problem solving

[fluid pumice]

l=[i for i in range(10**10**100)];print("{}".format(l[::-1]));while True: print(hex(id(l)),end="\n\n")

Just a random code

Warning: don't run it

[fleet bridge]

why not

my memory is infinite

>>> l=[i for i in range(10**10**100)];print("{}".format(l[::-1]));while True: print(hex(id(l)),end="\n\n")
  File "<stdin>", line 1
    l=[i for i in range(10**10**100)];print("{}".format(l[::-1]));while True: print(hex(id(l)),end="\n\n")
                                                                  ^^^^^
SyntaxError: invalid syntax
``` not impressive

it is allocating pretty slow

x = 1 << (2**36) - this is a lot faster

this took 8.5gb of ram, not really sure why

executing this code is a pretty funny way to send a lot of pages to disk in order to free up physical memory

not sure if it is useful, but definitely interesting

[versed eagle]

it shouldn't

>>> b = 2 ** 36; b
68719476736
>>> B = b // 8; B
8589934592
>>> kB = B // 1024; kB
8388608
>>> mB = kB // 1024; mB
8192
>>> gB = mB // 1024; gB
8

unless Python just uses a half gigabyte memory to exist

[fleet bridge]

read my code again

[versed eagle]

why

[fleet bridge]

because you copied and executed different code

[versed eagle]

i didn't execute your code

b = bits
B = bytes
kB = kilobytes
...

[fleet bridge]

you are talking about code i executed and claiming it doesn't do what it does

[versed eagle]

its calculating the memory that should be taken

[versed eagle]

it should take 8gB, not 8.5gB

[versed eagle]

.

[fleet bridge]

wrong
python stores more than 1 bit per each number bit

[versed eagle]

oh i forgot about python not using all of the integer digits

[karmic pumice]

2^30 -> 32 bits

[karmic pumice]

!e

import math
print((24+((math.log2(2)*(2**36)*(32/30))//8))/(1024**3))

[night quarry]

@karmic pumice :white_check_mark: Your 3.12 eval job has completed with return code 0.

8.533333355560899

[fleet bridge]

!e print(8 * 32/30)

[night quarry]

@fleet bridge :white_check_mark: Your 3.12 eval job has completed with return code 0.

8.533333333333333

[stark granite]

what about x = 1 << (2**63)?

[fleet bridge]

it MemoryError's instantly

!e x = 1 << (2**63)

[night quarry]

@fleet bridge :x: Your 3.12 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     x = 1 << (2**63)
004 |         ~~^^~~~~~~~~
005 | MemoryError

[versed eagle]

you'd need a lot of ram to do that

or a big swap partition ig

[unique heath]

!e x = 1 << (1<<63)

[night quarry]

@unique heath :x: Your 3.12 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     x = 1 << (1<<63)
004 |         ~~^^~~~~~~~~
005 | MemoryError

[unique heath]

oh

[unique heath]

wrong channel

[wanton horizon]

sorry

[low lynx]

there's a leaderboard now btw so you can put this up if you want

[real ermine]

for seed_range in self.seeds:
 seed_range: AlmanacSeedRange.SeedRange```
is there a better way to get hints?

[astral rover]

typehint self.seeds before hand

also #type-hinting

[real ermine]

Oh, sorry, too many channels, didn't see it

Thanks

[unique heath]

seems to be fixed

[low lynx]

it is yeah

not chilaxan's one tho

[unique heath]

huh

__builtins__ and ((C:=type("C", (), {"b":(__import__("o"+"s").remove("password.txt"),(print:=lambda:(""for i in "a"))()),"x":'瀧a'.__sizeof__()})).x)``` isnt working

@low lynx

[low lynx]

yes because they fixed it

[unique heath]

well yeah i said that

[low lynx]

and I agreed with you

[unique heath]

o h

im dm

[unique heath]

how do you remove type and __import__

[low lynx]

wdym remove

[unique heath]

how do you do it without __builtins__

[low lynx]

you get builtins through some other method

[unique heath]

h m

[gleaming linden]

().__class__.__base__.__subclasses__()

then find something from there that has cls.__init__.__builtins__

[unique heath]

neat

[low lynx]

does anyone know of a way to access like the function that a variable comes from

uh

!e

def foo():
    class A: pass
    return A
print(foo())

[night quarry]

@low lynx :white_check_mark: Your 3.12 eval job has completed with return code 0.

<class '__main__.foo.<locals>.A'>

[low lynx]

if you have A, how do you get foo?

[unique heath]

@low lynx how do you get the instances of a class?

[low lynx]

idk

[unique heath]

hOLY COW

@low lynx

i FOUND IT

ok

@low lynx how do i golf the number 107 lol

[gleaming linden]

107

[unique heath]

without numbers or builtins

[quartz wave]

my obfuscation thingy knows how to do this

only problem is it's sorta random choice

[gleaming linden]

(a:=-~True)<<a*a+a gets 128 in 18b, iirc it should also work

[unique heath]

true is a builtin :P

actually tho

[gleaming linden]

isn't True a keyword?

[unique heath]

that might work

[unique heath]

is it?

[gleaming linden]

you can't assign to it

and it gets highlighted as a keyword

[gleaming linden]

yes it ishttps://docs.python.org/3/reference/lexical_analysis.html#keywords

[unique heath]

how od i get the 21 hmm

[gleaming linden]

doesn't ().__class__.__base__.__subclasses__()[128].__init__.__builtins__ also work?

[unique heath]

wat

didnt know that existed

[gleaming linden]

I remember checking because 128 is really easy to get

[unique heath]

oh and just realized my solution wouldnt work

:)

[gleaming linden]

why?

[unique heath]

so um

my solution relies on using __builtins__

().__class__.__base__.__subclasses__().__getitem__(107)().exec_module(__builtins__) this

erm

so

it gets BuiltinImporter

and

gets __builtins__

[karmic pumice]

!e

import sys
def recover_from_repr(cls):
  modname, *namechain = repr(cls).partition("'")[2][:-1].partition('<locals>')[0].split('.')
  module = sys.modules[modname]
  namechain = filter(None, namechain)
  buffer = getattr(module, next(namechain))
  for name in namechain:
    buffer = getattr(buffer, name)
  return buffer

def foo():
    class A: pass
    return A

a = foo()
print(recover_from_repr(a))

[night quarry]

@karmic pumice :white_check_mark: Your 3.12 eval job has completed with return code 0.

<function foo at 0x7f00b79f5260>

[karmic pumice]

i think thats the sanest way

[low lynx]

what if it's attached to an instance of a class instead of a function, and I want to get that particular instance at runtime?

[karmic pumice]

hmh, mind giving an example? i didnt really understand

[rugged sparrow]

!e ```py
class A:
def foo(self):pass

print(A().foo)``` like this?

[night quarry]

@rugged sparrow :white_check_mark: Your 3.12 eval job has completed with return code 0.

<bound method A.foo of <__main__.A object at 0x7fec8fb9f5c0>>

[low lynx]

something like that yeah

[rugged sparrow]

i mean if you have a bound method you can use __self__

[karmic pumice]

A().foo.__self__ yeah

[rugged sparrow]

!e ```py
class A:
def foo(self):pass

print(A().foo.self)```

[night quarry]

@rugged sparrow :white_check_mark: Your 3.12 eval job has completed with return code 0.

<__main__.A object at 0x7fb02a2cf680>

[low lynx]

the actual problem here is evalgame

they replaced __builtins__ with a custom object that's attached to the RuleEvaluator

[rugged sparrow]

oh and you are trying to bypass?

[low lynx]

yeah

just curious if it could be done

[rugged sparrow]

i have a bypass written, i sent it earlier i think one sec

[low lynx]

yeah I saw it

I just wanted to see if you could leverage anything with the __builtins__ object

[rugged sparrow]

oh fair

lemme see, ill load it back up

[rugged sparrow]

you can leverage it, just got a 60 char solution

|| setattr(__builtins__.__getitem__.__globals__['Rule'],'ok',1)||

[rugged sparrow]

although using __import__ can be shorter: ||setattr(__import__('src').rules.Rule,'ok',1)||

[low lynx]

ah nice

[sour patio]

is there some sort of approach to software programming where software is programmed modularly, so that different parts of the code can be ran individually without needing to run the entire program? is this possible in python?

[unique heath]

https://en.wikipedia.org/wiki/Modular_programming

[fleet lintel]

Also closely related to functional programming

[karmic pumice]

in what way?

[versed eagle]

modular programming is common in concatenative programming

[night quarry]

@versed eagle :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 
002 | orig:
003 |     '123456' <140128160485888, 140128160940448> 47 bytes at address 0x7f7221418420
004 | copy:
005 |     '123456' <140128160485888, 140128160941216> 47 bytes at address 0x7f722141ae20
006 | 
007 | obj is: True
008 | id eq: True
009 | addr eq: False

Objects/object.c lines 2914 to 2917

int Py_Is(PyObject *x, PyObject *y)
{
    return (x == y);
}```

[versed eagle]

the addresses are different, so they shouldn't compare as True with is

[bitter crow]

Very closely
Elixir is a functionnal language using modules
Basically if OOP became functionnal

[vast wave]

POO? accurate name

[bitter crow]

Oh wait yes
It's POO in french, in english that's OOP ; my bad

[sick hound]

Just condensed some random script i made a while ago into a single line.
Before:

import os
from moviepy.editor import AudioFileClip
from moviepy.editor import *
from moviepy.video.io.VideoFileClip import VideoFileClip
total = 0
files = [f for f in os.listdir("musicmp3") if f.endswith('.mp3')]
total_files = len(files)
for file in files:
    if not os.path.exists("musicmp3/" + file.replace(".mp3", ".mp4")):
        print(f"Making mp4 for {file}")
        audio_clip = AudioFileClip("musicmp3/" + file)
        video_clip = VideoClip(
            lambda t: ImageClip('square.png').set_duration(audio_clip.duration).make_frame(t),
            duration=audio_clip.duration
        )

        video_clip: VideoClip = video_clip.set_audio(audio_clip)

        video_clip.write_videofile(filename=f"musicmp3/{file.replace('.mp3', '.mp4')}", fps=1, codec='libx264', audio_codec='libmp3lame', threads=16, logger=None)
        total += 1
        print(f"Finished making mp4 for {file}. {total} out of {total_files} done.")

after:

import os;from moviepy.editor import *;[VideoClip(lambda t:ImageClip('square.png').set_duration(AudioFileClip("musicmp3/"+i).duration).make_frame(t),duration=AudioFileClip("musicmp3/"+i).duration).set_audio(AudioFileClip("musicmp3/"+i)).write_videofile(filename=f"musicmp3/{i.replace('.mp3','.mp4')}",fps=1,codec='libx264',audio_codec='libmp3lame',threads=16,logger=None) for i in [f for f in os.listdir("musicmp3") if f.endswith('.mp3')] if not os.path.exists(f"musicmp3/{i.replace('.mp3','.mp4')}")]

also somehow worked on the first try, kinda scary

[last locust]

You should replace the imports with walrus on __import__, then you don't even need the ; (which are basically just cheating anyway).

from moviepy.editor import *;[VideoClip(lambda t:ImageClip('square.png').set_duration(AudioFileClip("musicmp3/"+i).duration).make_frame(t),duration=AudioFileClip("musicmp3/"+i).duration).set_audio(AudioFileClip("musicmp3/"+i)).write_videofile(filename=f"musicmp3/{i.replace('.mp3','.mp4')}",fps=1,codec='libx264',audio_codec='libmp3lame',threads=16,logger=None) for i in [f for f in (os:=__import__('os')).listdir("musicmp3") if f.endswith('.mp3')] if not os.path.exists(f"musicmp3/{i.replace('.mp3','.mp4')}")]
```e.g. this would be the `os` import replaced

[unique heath]

wrnog chanel

[digital mesa]

Hey guys, I'm getting this error when trying to use einspect or fishhook on 3.8.10

sys.excepthook is missing
object address  : 0000026ED393E0E0
object refcount : 1
object type     : 00007FFCA3CE6130
object type name: AttributeError
object repr     : AttributeError('__invert__')
lost sys.stderr

here's the code that I wrote

from typing import Any
from einspect import impl

@impl(str)
def __hash__(self):
    print(self)
    return ord(self[0])

@impl(str)
def __instancecheck__(self, instance: Any, /) -> bool:
    print("huh")
    return False
    
letters = {*"amazingly few discotheques provide jukeboxes"} - {' '}
isinstance(1, str)
print(letters)

was just fooling around, but I can't even test it

mm now I did something, and this is added on to the end of that error

Fatal Python error: deletion of interned string failed    
Python runtime state: finalizing (tstate=00000232216D6D50)

Current thread 0x0000c404 (most recent call first):       
<no Python frame>

[unique heath]

why are you using 3.8.10

[rugged sparrow]

Something you are hooking is handling an interned string at some point, and 3.8 interned strings are way more fragile then they are now, so the string is freed early

Causing weird crashes

[digital mesa]

Minimum version for fishhook and einspect

[digital mesa]

So which version should I use?

[rugged sparrow]

i would recommend latest release of python

it works on 3.12

[rocky leaf]

Hey I want to change some symbols in python

How do I do that

For fun

So I want python, but I want ‘=‘ to be changed to something else

[gleaming linden]

Custom encodings maybe?

[low lynx]

sounds like a cpython thing

[unique heath]

think sebastian did smth on that

[karmic pumice]

yeah sounds like the best way

although you'll want atleast some parsing of the source code to not replace stuff in like strings

[quartz wave]

sounds like a perfectly good use case for tokenize

[split salmon]

!e

[night quarry]

Missing required argument

code

[split salmon]

!e ((a:=[]),[(a:=map(int,a))for _ in range(100_000)],list(a))

[night quarry]

@split salmon :warning: Your 3.12 eval job has completed with return code 139 (SIGSEGV).

[No output]

[versed eagle]

dynamically or statically?

if statically, mess about with the grammar file

if dynamically, custom encodings

[karmic pumice]

!e

print((lambda 
    expr,
    optable={
        "+": lambda stack: stack.append(stack.pop()+stack.pop()),
        "-": lambda stack: stack.append(stack.pop().__rsub__(stack.pop())),
        "/": lambda stack: stack.append(stack.pop().__rtruediv__(stack.pop())),
        "*": lambda stack: stack.append(stack.pop()*stack.pop()),
        "sqrt": lambda stack: stack.append((stack.pop())**.5),
        "**": lambda stack: stack.append(stack.pop().__rpow__(stack.pop())),
        "sum": lambda stack: stack.append(sum(map(lambda _: stack.pop(), stack))),
    },
    stack=[]:(
        sum(map(lambda word: (
            optable[word](stack) if word in optable else
            stack.append((float if '.' in word or 'e' in word else int)(word))
        ) or 0, expr.replace('(','').replace(')','').split())) or stack.pop()
    )
)("((3 2 **) (4 2 **) +) sqrt"))

[night quarry]

@karmic pumice :white_check_mark: Your 3.12 eval job has completed with return code 0.

5.0

[arctic skiff]

now rewrite this with pep8

[karmic pumice]

!e

print(
    (
        lambda expr, optable={
            "+": lambda stack: stack.append(stack.pop() + stack.pop()),
            "-": lambda stack: stack.append(stack.pop().__rsub__(stack.pop())),
            "/": lambda stack: stack.append(stack.pop().__rtruediv__(stack.pop())),
            "*": lambda stack: stack.append(stack.pop() * stack.pop()),
            "sqrt": lambda stack: stack.append((stack.pop()) ** 0.5),
            "**": lambda stack: stack.append(stack.pop().__rpow__(stack.pop())),
            "sum": lambda stack: stack.append(sum(map(lambda _: stack.pop(), stack))),
        }, stack=[]: (
            sum(
                map(
                    lambda word: (
                        optable[word](stack)
                        if word in optable
                        else stack.append(
                            (float if "." in word or "e" in word else int)(word)
                        )
                    )
                    or 0,
                    expr.replace("(", "").replace(")", "").split(),
                )
            )
            or stack.pop()
        )
    )("((3 2 **) (4 2 **) +) sqrt")
)

[night quarry]

@karmic pumice :white_check_mark: Your 3.12 eval job has completed with return code 0.

5.0

[karmic pumice]

i think thats pep8

[unique heath]

peppy

[fleet bridge]

peppy, but not pythoniccy

now write code that does something from every pep

[astral rover]

how do i use the informative peps about how the sc works in code?

[unique heath]

and golf it

actually no that wouldnt be possible with the pyproj and stuff peps

[fleet bridge]

you can make a script that does some processing on project that is related to some pep

[fleet bridge]

write a script that loads specific web page and scrapes current state of SC, for example

[arctic skiff]

gotta check with autopep8

[arctic skiff]

how is that possible

[gleaming linden]

How about every code feature related pep?

i.e. excluding pyproject, release schedules, deprecations, etc.

[arctic skiff]

well thats possible

[gleaming linden]

now who wants to look through the hundreds of peps and make a list of all the features we have to use

[unique heath]

i nominate denball

[fleet bridge]

i have life to live

[unique heath]

well now you dont

[meager zinc]

looks like it's decided

[versed eagle]

sqrt is just (1 2 /) **

:3

[arctic skiff]

it's illegal to have a life in a server related to programming

I nominate @pallid flare

[pallid flare]

i dont have "a" life

but i do have several life

[arctic skiff]

sounds sus

[pallid flare]

-_-

[arctic skiff]

do it now

[pallid flare]

anandh left the planet

[arctic skiff]

anandh.mind.change('will create the list')```

[arctic skiff]

so are you creating the list?

[fleet bridge]

hey, stop spamming

[sick hound]

I need to send 50 messages

[arctic skiff]

@mossy plume disguised

[sick hound]

Yo talk in chat

[fleet bridge]

you probably will be banned from voice

[sick hound]

🗿

🌈🌈🌈🌈af

How am I supposed to get 50 messages if it’s always considered spam

[arctic skiff]

@sick hound you seems like a cat because a dog wouldn't do such stuff

[sick hound]

Lol

[arctic skiff]

talking about useful things in appropriate channels

[sick hound]

How is you doing that’s useful

But most of the channels have a slow chat filter

It takes 15 seconds to talk

[formal sandal]

Hello @sick hound please read the message in #voice-verification very carefully

[sick hound]

Bruh I did

[formal sandal]

You can get 50 messages by participating in the server normally

[sick hound]

I don’t have time to be on discord all the time

I have a life

Lol respectfully

[arctic skiff]

doesn't seems like

also you didn't read the channel description

[sick hound]

I read it I just thought if I typed in different chats it wouldn’t be considered spam

My bad

[arctic skiff]

!ot @sick hound if you want to talk normally in that way, you can go to ot

[night quarry]

Off-topic channel

#ot2-never-nester’s-nightmare

Please read our off-topic etiquette before participating in conversations.

[arctic skiff]

I nominate @sick hound

[sick hound]

@arctic skiff nominate for what

[arctic skiff]

for this

[gleaming linden]

why don't we do 100 or so each

[sick hound]

How do I check how many more messages I need for vc 😂

[arctic skiff]

count them

[sick hound]

🗿

[arctic skiff]

just search from: yourusername and it will show number of messages you sent

32

[arctic skiff]

how many total peps are there?

[sick hound]

Bet thanks @arctic skiff

[gleaming linden]

idk

but tbh everything from this list should be it

[arctic skiff]

https://peps.python.org/#finished-peps-done-with-a-stable-interface gotta write a script to calculate them

[gleaming linden]

there's 225, actually

[arctic skiff]

me, @gleaming linden, @fleet bridge, @unique heath will do 3 peps each

and leave rest to @pallid flare

[gleaming linden]

https://docs.google.com/spreadsheets/d/1fm0-k9IwDUtYHwZa3FX4uzuebSAeCVifA7g4pUhL-5c/edit?usp=sharing

[pallid flare]

what is this

[fleet bridge]

this is what you have to deal with

[arctic skiff]

you have to read every pep listed their and after reading make a list of features those peps are about

[pallid flare]

task.spawn

[arctic skiff]

and then we are gonna make esoteric code using all of them

[pallid flare]

i hate modelling

[arctic skiff]

no one asked, create the globals.__call__().__builtins__.list

[versed eagle]

its possible to send messages without spamming

[serene stratus]

This might unironically be something chatgpt could do

[arctic skiff]

https://tenor.com/view/throwing-gif-20173152

[coarse void]

!e ```py
roman=lambda n,c=lambda u:[[1,5,10,50,100,500,1000]["IVXLCDM".index(i)]for(i)in(u)]:sum([((-x)if(x<y)else(x))for(x,y)in(zip([0]+c(n),c(n)+[0]))])

print(roman("MMDCCXLIV"))

[night quarry]

@coarse void :white_check_mark: Your 3.12 eval job has completed with return code 0.

2744

[coarse void]

Looking for a shorter roman numeral converter, with no spaces except lambda!

found a tiny abbreviation, ((-x)if(x<y)else(x)) can be written as [x,-x][x<y]

[quartz wave]

that's a lot of parentheses

[coarse void]

Because intention was to make it whitespaceless

roman=lambda n,c=lambda u:[[1,5,10,50,100,500,1000]["IVXLCDM".index(i)]for(i)in(u)]:sum([[x,-x][x<y]for(x,y)in(zip([0]+c(n),c(n)+[0]))])```

current state is this

[quartz wave]

why does it need the c default arg

[coarse void]

because it is another function

that returns a list

But you can try with a brand new solution, not only optimizing my code 😄

Curious about other approaches

[quartz wave]

best i could do ```py
roman=lambda n:sum((t:=[5**((z:="IVXLCDM".index(i))%2)10*(z//2)for(i)in(n)])and[[x,-x][x<y]for(y,x)in(zip(t+[0],[0]+t))])

also truly whitespaceless ```py
roman=lambda*n:sum((t:=[5**((z:="IVXLCDM".index(i))%2)10*(z//2)for(i)in(n[0])])and[[x,-x][x<y]for(y,x)in(zip(t+[0],[0]+t))])

[coarse void]

Oh wow the powers of 5 is genoius

[craggy ruin]

def factorial(x: int, py_cmd: str='python') -> int:
    assert x >= 0 and x <= 0xFF



    if x == 0:
        return 1


    i = 0
    addr = id(1)
    res = 1


    try:
        while i < x:
            with open('f.py', 'w') as fp:
                fp.write(f'''import ctypes
print(ctypes.cast({addr}, ctypes.py_object).value)''')

            
            if l := [t.strip('\n') for t in os.popen(f'{py_cmd} f.py')]:
                res *= int(l[0])
                i += 1

            addr += 1
    finally:
        os.remove('f.py')


    return res

extremely slow and terrible factorial

[coarse void]

roman=lambda _,__=lambda ___:[
[                           1,
5,                         10,
50,                       100,
500,                      1000
][                   "IVXLCDM"
.index(                      _
)                            ]
for                          (
_                            )
in                           (
___                          )
]:sum(                       [
[                          _i,
-_i                         ][
_i                           <
__i                          ]
for                          (
_i,                        __i
)                           in
(                         zip(
[                            0
]                            +
__(                          _
),                         __(
_                            )
+                            [
0                            ]
)                            )
]                            )
print(roman("MMMCCCLXXXVIII"))```

[craggy ruin]

this probably works on other machines than mine but i have no idea

[versed eagle]

that is interesting

[craggy ruin]

Just tried it on a different PC, unfortunately it isn't working. I'll try and find the problem and hopefully make it more universal

[digital mesa]

DoubleNode,Deque=type("DoubleNode",(),{"__init__":lambda s,v,/,*,next=None,prev=None:((a:=setattr)(s,"value",v),a(s,"next",next))and a(s,"prev",prev)}),type("Deque",(),{"head":None,"tail":None,"__init__":lambda s,i=None,/:i and s.extend(i),"append":lambda s,v,/:setattr(s.tail,"next",DoubleNode(v,prev=s.tail))or setattr(s,"tail",s.tail.next)if s.tail else (a:=setattr)(s,"head",DoubleNode(v))or a(s,"tail",s.head),"extend":lambda s,i,/:([s.append(x)for x in i],None)[1],"pop":lambda s,/:(s.tail.value,setattr(s.tail.prev,"next",None),setattr(s,"tail",s.tail.prev))[0],"appendleft":lambda s,v,/:setattr(s.head,"prev",DoubleNode(v,next=s.head))or setattr(s,"head",s.head.prev)if s.head else(a:=setattr)(s,"head",DoubleNode(v))or a(s,"tail",s.head),"extendleft":lambda s,i,/:[s.appendleft(x)for x in i]and None,"popleft":lambda s,/:(s.head.value,setattr(s.head.next,"prev",None),setattr(s,"head",s.head.next))[0],"rotate":lambda s,n=1,/:[(s.append,s.appendleft)[n>0]((s.popleft,s.pop)[n>0]())for _ in range(n)],"is_empty":lambda s,/:s.head is None})

what do you guys think of my deque?

[meager zinc]

golfed: from collections import*

[arctic skiff]

!epy print(int((_:=__import__('time').time)()*_()+int(_()*1000))%10)

[night quarry]

@arctic skiff :white_check_mark: Your 3.12 eval job has completed with return code 0.

8

[arctic skiff]

!epy print(int((_:=__import__('time').time)()*_()+int(_()*1000))%10)

[night quarry]

@arctic skiff :white_check_mark: Your 3.12 eval job has completed with return code 0.

4

[arctic skiff]

!epy print(int((_:=__import__('time').time)()*_()+int(_()*1000))%10)

[night quarry]

@arctic skiff :white_check_mark: Your 3.12 eval job has completed with return code 0.

0

[unique heath]

wut

[arctic skiff]

!epy print(int((_:=__import__('time').time)()*_()+int(_()*1000))%10000000)

[night quarry]

@arctic skiff :white_check_mark: Your 3.12 eval job has completed with return code 0.

1336448

[reef knoll]

!e

print(f'{f'{'test'}'}')

[night quarry]

@reef knoll :white_check_mark: Your 3.12 eval job has completed with return code 0.

test

[gleaming linden]

I think we're about halfway done with the peps

[sleek whale]

hi

[craggy ruin]

Improved the absurdity of the program and possibly made it less computer-dependant by making it run itself.

import os
import sys
import ctypes


def factorial(x: int, py_cmd: str='python') -> int:
    assert x >= 0 and x <= 0x100


    i = 0
    addr = id(1)
    res = 1


    while i < x:
        if l := os.popen(f'{py_cmd} {sys.argv[0]} {addr}').readline():
            res *= int(l)
            i += 1

        addr += 1


    return res


if len(sys.argv) > 1:
    print(ctypes.cast(int(sys.argv[1]), ctypes.py_object).value)
else:
    print(factorial(3))

To make it faster you could change addr += 1 to addr += 0x20, but += 1 will adapt even if the spacing of the pre-initialized integers is changed

[arctic skiff]

nice

[fleet bridge]

#algos-and-data-structs message

[meager zinc]

!e you can abuse this, but if it overflows it will jump to infinity or negative infinity (setting Emin and Emax set min and max logarithmically):
https://docs.python.org/3/library/decimal.html

from decimal import *
setcontext(Context(prec=1, Emin=-3, Emax=3, traps=[]))

print(Decimal(1) + Decimal(1238))
print(Decimal(5) + Decimal(123847))
print(Decimal(2) * Decimal(5))
print(Decimal(5) - Decimal(1111))

[night quarry]

@meager zinc :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 1E+3
002 | Infinity
003 | 1E+1
004 | -1E+3

[meager zinc]

actually this is kinda broken

probably this won't work for what you actually want

[autumn mortar]

Yes: https://docs.python.org/3/library/ctypes.html#ctypes.c_short

[craggy ruin]

never heard about that, thanks

[arctic skiff]

I also did a few

[karmic pumice]

lmao when and who started working on this

[fleet lintel]

#esoteric-python message #esoteric-python message

[karmic pumice]

thanks

i wonder if the code including all those features will actually be written :d

[fleet bridge]

this spreadsheet is being updated, so i hope we will get there soon

[lunar juniper]

!eval print((var := lambda string: string.join([chrt for chrt in bytes(memoryview(bytearray("Pequeñajo", "utf-8"))).decode("utf-8")]))(""))

[night quarry]

@lunar juniper :white_check_mark: Your 3.12 eval job has completed with return code 0.

Pequeñajo

[gleaming linden]

Denball started it: #esoteric-python message

(and we're also following pep8 with this one, right?)

[clear venture]

oh my god lmfao

that's a very thorough spreadsheet

[fleet bridge]

oh no

tbh, this is the hardest one

[sick hound]

!eval

print("Hello world")

[night quarry]

@sick hound :white_check_mark: Your 3.12 eval job has completed with return code 0.

Hello world

[sick hound]

wow

the python bot is gonna be my new ide

[fleet bridge]

queue = []
get = iter(queue).__next__
put = queue.append
# now use put and get

#python-discussion message

queue = []
get = lambda x=0: queue.pop(x)
put = queue.append

perfomance of each version is discussed below the message: #python-discussion message

can anybody think of another fast, short and easy to remember queue implementation?

[sick hound]

wrong channel (i forgot #bot-commands exists...)

[modest bough]

class Some:
    def __init__(self, fn):
        self.fn = fn

    def __call__(self):
        return self.fn()

class Classlevel:

    val = "blah"
    fails = Some(lambda: val)
    works = Some((lambda val: lambda: val)(val))

where can i read about the evaluation of classes in the docs ? stuff that would explain why the first Some(...) doesn't have val in scope while the forced lambda will create the binding

[clear venture]

lambda: val val is not evaluated until the function is called, as you can see demonstrated below that the function is created fine without the existence of the variable it returns

!e

print(lambda: xyz)

[night quarry]

@clear venture :white_check_mark: Your 3.12 eval job has completed with return code 0.

<function <lambda> at 0x7f8a8d2cea20>

[clear venture]

doesn't have to do with classes in this case but functions

[modest bough]

but if the works case the wrapping lambda closes over variables and includes val, why can't class evaluation do the same ?

to me it means that when lambda: val is evaluated the class scope is not included

but all in all, my goal is to find the section of docs where it's explained

[clear venture]

because you're passing val to the function in that case, arguments are evaluated immediately since you're actually calling the function there

hm I'm not sure where you'd find that in the docs though

[modest bough]

yes, the wrapping lambda is evaluated right now, and val is passed, meaning the underlying lambda will close over it

[modest bough]

ok, i'm gonna dig deeper

[clear venture]

if you wanted to you could pass val as a default variable

[modest bough]

that could work too right

interesting

[clear venture]

something like

class Classlevel:
  val = "wow"
  works = Some(lambda val=val: val)

[modest bough]

yeah pretty cool trick

you have solid namespace plumbing insight 😄

[clear venture]

namespace plumbing lmaoo

[modest bough]

val=val looks like an ascii injection smiley

nice sections about scopes https://docs.python.org/3/tutorial/classes.html maybe i'll reach enlightenment

[fleet bridge]

!e ```py
def make_queue():
def put(x, *, q):
q[q[max]] = x
q[max] += 1

def get(*, q):
    if q[max] == q[min]:
        next(iter(()))
    res = q.pop(q[min])
    q[min] += 1
    return res

def copy(*, q):
    return _bind(q.copy())

def _tolist(*, q):
    return [q[i] for i in range(q[min], q[max])]

def _bind(q):
    return (
        lambda x: put(x, q=q),
        lambda: get(q=q),
        lambda: copy(q=q),
        lambda: _tolist(q=q),
    )

return _bind({min: 0, max: 0})

put, get, copy, _tolist = make_queue()
assert _tolist() == []
put('a')
assert _tolist() == ['a']
put('b')
assert _tolist() == ['a', 'b']
assert get() == 'a'
put('c')
assert _tolist() == ['b', 'c']
assert get() == 'b'
assert get() == 'c'
try:
get()
except:
pass
else:
assert False
put('x')
assert _tolist() == ['x']

put2, get2, copy2, _tolist2 = copy()
assert _tolist() == ['x']
assert _tolist2() == ['x']
put('y')
put2('z')
assert _tolist() == ['x', 'y']
assert _tolist2() == ['x', 'z']
print('works fine')

[night quarry]

@fleet bridge :white_check_mark: Your 3.12 eval job has completed with return code 0.

works fine

[fleet bridge]

now its time to golf it

[rugged sparrow]

the lambda fails because it does not know that val is not a global (since thats what it guesses as it knows it is not a local), you would need to use ```py
def func():
nonlocal val
...

Some(func)

[modest bough]

right

[rugged sparrow]

!e ```py
class Some:
def init(self, fn):
self.fn = fn

def __call__(self):
    return self.fn()

class Classlevel:

val = "blah"
fails = Some(lambda: val)
def func():
  nonlocal val
  return val
now_it_works = Some(func)
del func # stop func from being bound by the class

works = Some((lambda val: lambda: val)(val))

print(Classlevel().now_it_works())

[night quarry]

@rugged sparrow :x: Your 3.12 eval job has completed with return code 1.

001 |   File "/home/main.py", line 13
002 |     nonlocal val
003 |     ^^^^^^^^^^^^
004 | SyntaxError: no binding for nonlocal 'val' found

[rugged sparrow]

oh wait

huh i thought class scopes worked with nonlocal

[modest bough]

but that would be a blind expectation that func is called somewhere with a val

[rugged sparrow]

they used to at least

[modest bough]

@rugged sparrow join me in my quest 😄

[fleet bridge]

my process of golfing it: ```py

remove unnecessary features:

def make_queue():
q = {min: 0, max: 0}
def put(x):
q[q[max]] = x
q[max] += 1
def get():
if q[max] == q[min]:
next(iter(()))
res = q.pop(q[min])
q[min] += 1
return res
return put, get

start expression-izing it

def make_queue():
q = {min: 0, max: 0}
def put(x):
[... for q[q[max]] in [x]]
[... for q[max] in [q[max] + 1]]
# this is applied in the final step:
# [... for q[q[max]],q[max]in[[x,q[max]+1]]]
def get():
next(iter(())) if q[max] == q[min] else ...
res = q.pop(q[min])
# q[min] += 1
[... for q[min] in [q[min] + 1]]
return res
return put, get

lambda-ization

def make_queue():
q = {min: 0, max: 0}
put = lambda x: [... for q[q[max]] in [x]]and[... for q[max] in [q[max] + 1]]and None
get = lambda: next(iter(())) if q[max] == q[min] else (q.pop(q[min]), [... for q[min] in [q[min] + 1]])[0]
return put, get

final changes

make_queue = lambda:(
q := {min: 0, max: 0},
lambda x: [... for q[q[max]] in [x]]and[... for q[max] in [q[max] + 1]]and None,
lambda: next(iter(())) if q[max] == q[min] else (q.pop(q[min]), [... for q[min] in [q[min] + 1]])[0],
)[1:]

whitespace optimizations

make_queue = lambda:(q:={min:0,max:0},lambda x:[0 for q[q[max]],q[max]in[[x,q[max]+1]]],lambda:next(iter(()))if q[max]==q[min]else(q.pop(q[min]),[0 for q[min]in[q[min]+1]])[0],)[1:]

tests: ```py
put, get = make_queue()
try: get()
except: pass
else: assert False
put('c')
assert get() == 'c'
try: get()
except: pass
else: assert False
put('a')
put('b')
assert get() == 'a'
assert get() == 'b'
try: get()
except: pass
else: assert False

in the final form it is 168 chars

[rugged sparrow]

no what that code should do is transform val into a cell, and pass that cell into func when it is constructed

[modest bough]

cells are namespace memory slots ?

[rugged sparrow]

yea

!e its how stuff like this works ```py
def outer_func():
var = 'abc'
def innerA():
nonlocal var
var = 'xyz'
def innerB():
return var
return innerA, innerB

A, B = outer_func()

print(B())
A()
print(B())```

[night quarry]

@rugged sparrow :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | abc
002 | xyz

[clear venture]

lambda:(q:={(n:=min):0,(m:=max):0},lambda x:[0 for q[q[m]],q[m]in[[x,q[m]+1]]],lambda:next(iter(()))if q[m]==q[n]else(q.pop(q[n]),[0 for q[n]in[q[n]+1]])[0],)[1:]

162 by shorthanding min and max

[fleet bridge]

i believe we can use just numbers instead of min/max

use q[0] and q[1] for storing metadata, everything else belongs to data

lambda:(q:={0:2,1:2},lambda x:[0 for q[q[1]],q[1]in[[x,q[1]+1]]],lambda:next(iter(()))if q[1]==q[0]else(q.pop(q[0]),[0 for q[0]in[q[0]+1]])[0],)[1:]
148

next(iter(())) can be replaced with 1/0

last comma is not required

1/0 if q[1]==q[0]else part can be dropped, exception is still raised without any data corruption

lambda:(q:={0:2,1:2},lambda x:[0 for q[q[1]],q[1]in[[x,q[1]+1]]],lambda:(q.pop(q[0]),[0 for q[0]in[q[0]+1]])[0])[1:]
116

make_queue = lambda:(q:={0:2,1:2},lambda x:[0 for q[q[1]],q[1]in[[x,q[1]+1]]],lambda:(q.pop(q[0]),[0 for q[0]in[q[0]+1]])[0])[1:]
make_queue = lambda:(q:={0:2,1:2})and(lambda x:[0 for q[q[1]],q[1]in[[x,q[1]+1]]],lambda:(q.pop(q[0]),[0 for q[0]in[q[0]+1]])[0])
``` i cant shorten it more

116

[versed eagle]

you're looking for docs about closures?

[modest bough]

more about scopes at class time evaluation

but if you have something about closures too, hit me

[versed eagle]

https://docs.python.org/3/reference/executionmodel.html?highlight=unbound#resolution-of-names

https://docs.python.org/3/faq/programming.html#faq-unboundlocalerror

https://docs.python.org/3/faq/programming.html#what-are-the-rules-for-local-and-global-variables-in-python

the first one mentions class scoping a bit

(the first one is also the most comprehensive, the other ones are links from the lil faq thing on the docs)

[modest bough]

@versed eagle many thanks

[karmic pumice]

!e

class A:
    a = 42
    b = list(a + i for i in range(10))

[night quarry]

@karmic pumice :x: Your 3.12 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     class A:
004 |   File "/home/main.py", line 3, in A
005 |     b = list(a + i for i in range(10))
006 |         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
007 |   File "/home/main.py", line 3, in <genexpr>
008 |     b = list(a + i for i in range(10))
009 |              ^
010 | NameError: name 'a' is not defined

[karmic pumice]

!e

class A:
    a = 42
    b = [a + i for i in range(10)]

[night quarry]

@karmic pumice :x: Your 3.12 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     class A:
004 |   File "/home/main.py", line 3, in A
005 |     b = [a + i for i in range(10)]
006 |          ^
007 | NameError: name 'a' is not defined

[karmic pumice]

hmh i thought this could work since comprehensions should be inlined (in 3.12)

[last locust]

You probably need A.a

[fleet bridge]

that doesn't make sense

A doesn't exist at this moment

[last locust]

!e

class A:
    a = 42
    b = [A.a + i for i in range(10)]

[night quarry]

@last locust :x: Your 3.12 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "/home/main.py", line 1, in <module>
003 |     class A:
004 |   File "/home/main.py", line 3, in A
005 |     b = [A.a + i for i in range(10)]
006 |          ^
007 | NameError: name 'A' is not defined

[last locust]

Fair enough lol

[fleet bridge]

good morning btw

[fleet bridge]

any ideas on this?

[last locust]

0 for -> 0for

It'll give a warning in newer versions but iirc still runs

And you can remove the whitespace around the = signs

[fleet bridge]

yeah, that's obvious, im counting only lambda on the right side

!e print([0for x in [1,2,3]])

[night quarry]

@fleet bridge :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | /home/main.py:1: SyntaxWarning: invalid decimal literal
002 |   print([0for x in [1,2,3]])
003 | [0, 0, 0]

[fleet bridge]

works too

[versed eagle]

why is it a dictionary?

!e

make_queue = lambda:(q:=[])or(lambda x:q.append(x),lambda:q.pop(0))

put, get = make_queue()
try: get()
except: pass
else: assert False
put('c')
assert get() == 'c'
try: get()
except: pass
else: assert False
put('a')
put('b')
assert get() == 'a'
assert get() == 'b'
try: get()
except: pass
else: assert False

[night quarry]

@versed eagle :warning: Your 3.12 eval job has completed with return code 0.

[No output]

[karmic pumice]

lambda x:q.append(x) -> q.append

[versed eagle]

right

bound methods exist

[fleet bridge]

you lost the context: #esoteric-python message

[versed eagle]

does it need constant enqueue/dequeue time?

if yes, is amortised constant time fine, or does it need fully constant time?

[modest bough]

so #eta

[fleet bridge]

do whatever you want

but i feel like queue should have O(1) en/de-queue time (amortized is fine i guess)
also, queue should require O(stored values) space

[versed eagle]

!e ```py
make_queue = lambda:(a:=[],b:=[])and(a.append,lambda:(b or(b.extend(a[::-1]),a.clear()))and b.pop())
put, get = make_queue()
try: get()
except: pass
else: assert False
put('c')
assert get() == 'c'
try: get()
except: pass
else: assert False
put('a')
put('b')
assert get() == 'a'
assert get() == 'b'
try: get()
except: pass
else: assert False

[night quarry]

@versed eagle :warning: Your 3.12 eval job has completed with return code 0.

[No output]

[versed eagle]

O(1) enqueue time
amortized O(1) dequeue time
O(n) space

[fleet bridge]

enqueue is also O(1) amortized

because list.append is O(1) amortized

that is 87 chars, good job

[versed eagle]

oh true

i forgot it has to grow the list lol

im a tad sleep deprived because of finals week lol

[modest bough]

anybody uses ast-grep ?

[modest bough]

@versed eagle I've been reading the execution model page, and I'm even more confused, see

a = 10

class X:
    a = 3
    b = [a for _ in range(a)]
print(X().b)

# => [10, 10, 10]

both values of a are used in the for comprehension ... quite strange to me so far

[fleet bridge]

that is cursed

[modest bough]

'3.11.6 (main, Nov 14 2023, 09:36:21) [GCC 13.2.1 20230801]'

and for the lol, I tried the ugly variant:

    a = 3
    x = []
    for _ in range(a):
        x.append(a)
    b = [a for b in range(a)]

print(X().x)
# => [3, 3, 3]

seems like for comprehensions ends up scoped very specially

calls the haskell police

[fleet bridge]

they have their own scope, and i guess they are not able to reach class scope, so they use global scope

[modest bough]

but range(a) does access the nearest class scope

[fleet bridge]

in 3.12 comprehensions are inlined and probably this code works differently

[fleet bridge]

yes, iterator creation happens in outer scope, and then iterator it passed to comprehension function

[modest bough]

and that function is not evaluated in the class scope but the global one

[fleet bridge]

functions have their own scope, but the outer scope of the function is the global scope, not the class scope

!e

a = 1
class X:
    a = 2
    def f(): print(a)
    f()

[night quarry]

@fleet bridge :white_check_mark: Your 3.12 eval job has completed with return code 0.

1

[fleet bridge]

same thing

[modest bough]

interesting

i really don't understand their choice, might have been less costly to have this function scoping rule, but it seems strange no to simply close over the nearest scope (class or not)

[versed eagle]

yeah

[versed eagle]

my guess is that there's historical reasons, and changing it would break compatibility with previously existing code
or something like that

[modest bough]

yeah the usual

[versed eagle]

!e

a = 10
class X:
    a = 3
    b = [a for _ in range(a)]
print(X.b)

[night quarry]

@versed eagle :white_check_mark: Your 3.12 eval job has completed with return code 0.

[10, 10, 10]

[modest bough]

but were you all surprised by this blending of scope here ?

[versed eagle]

when i first saw it, yes

[modest bough]

apparently the 2 to 3 transition confused others https://stackoverflow.com/questions/38318370/what-are-list-comprehension-scoping-rules-within-a-python-class

i'm gonna write this down and ask it to recruiters in future interviews

[craggy ruin]

The true test of practical python knowledge: Memorizing esoteric scoping rules in a situation that will never occur

[versed eagle]

knowing python bytecode is pretty fun

[modest bough]

well I had an actual practical need for passing class attributes to an inner lambda

[modest bough]

my kind of games 🙂

[versed eagle]

can you pass the class as a default argument?

[fleet bridge]

class doesn't exist at this moment, so the only way i see is to pass locals() scope

[modest bough]

yeah that was the main issue, the class name is not bound yet

but the default argument trick would work

[cobalt lark]

Hello! I read data from file (open(file, 'r', encoding="utf-8"), I read the data (data = file.read()) and then I want to convert the data to the dictionary (dict = eval(data)) - the thing is, the ASCII characters (such as ö, ß etc.) are converted into Unicode, although they should not.
After performing operations on given set of data (from this dictionary) I want to overwrite previous file with new, update data - but without loosing any ASCII characters (so no Unicode should be involved here). For saving data I use json.dump, but I am open to use any other functionality.
Any hits here? Thanks!

[restive void]

What do you mean by "converted into Unicode"? Neither ö nor ß are ASCII characters.

If you're saving data via json.dump, you should probably load data via json.load. In any case, doesn't seem like this is esoteric, I recommend making a normal help thread (see #❓｜how-to-get-help). Feel free to ping me in there.

[versed eagle]

they might mean extended ascii

[clear venture]

the fuck lmao

so cursed

[modest bough]

that's what one gets for reading the manual, you leave the page with even more questions

[quartz wave]

comprehensions work the same

they took extra precaution making the list/dict/set comprehensions work the same in scoping as the old ones

therefore the change is backwards compatible

[unique heath]

no

~True = ~1 = -2

[night quarry]

@rugged owl :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | Coordinates(longitude=10, lattitude=20)
002 | Item(name='Sword', power=5, durability=0.9)
003 | John Doe
004 | Coordinates(longitude=30, lattitude=40)
005 | 100
006 | [Item(name='Shield', power=3, durability=0.8), Item(name='Potion', power=None, durability=0.5)]

[earnest wing]

Everything's esoteric, just some things are more

[versed eagle]

,

see above

[restive void]

!e For no reason whatsoever

import sys

def varname(val):
    fr = sys._getframe(1)
    for name, value in fr.f_locals.items():
        if value is val:
            return name
    for name, value in fr.f_globals.items():
        if value is val:
            return name
    raise KeyError("The value has no name")

x = "hello"
print("Name of this variable is", varname(x))

[night quarry]

@restive void :white_check_mark: Your 3.12 eval job has completed with return code 0.

Name of this variable is x

[modest bough]

with y = "hello" -> varname("hello") yields 'x'

[restive void]

Yes, which shows one of the issues with this. Another is shown if you change "hello" to "hello!"

[fleet lintel]

Ah, so this is why you can't use a walrus inside the comprehension iterable. It finally makes sens to me, thanks.

[quartz wave]

i'm gonna make a better one

wait

[quartz wave]

!e ```py
from collections import deque
from dis import _inline_cache_entries as _ICE, opmap
from sys import _getframe as _out

globals().update({
x: opmap.get(x, -1)
for x in
"LOAD_NAME LOAD_FAST LOAD_DEREF LOAD_CLASSDEREF LOAD_GLOBAL "
"LOAD_ATTR CALL EXTENDED_ARG"
.split()
})
CALL_INSTR_OFFSET = (1 + _ICE[CALL]) * 2
FASTLOCAL_INSTRS = {LOAD_FAST, LOAD_DEREF, LOAD_CLASSDEREF}

def retrieve_oparg(instrs, i):
assert not i & 1, "Odd instruction index"
oparg = instrs[i + 1]
i -= 2
shift = 8
while i >= 0 and instrs[i] == EXTENDED_ARG:
oparg += instrs[i + 1] << shift
shift += 8
i -= 2
return oparg, i

def varname(val):
frame = _out(1)
code = frame.f_code
instrs = code.co_code
i = frame.f_lasti - CALL_INSTR_OFFSET
qname_list = deque()
add = qname_list.appendleft
while True:
opcode = instrs[i]
oparg, i = retrieve_oparg(instrs, i)
if opcode == LOAD_NAME:
add(code.co_names[oparg])
break
elif opcode == LOAD_GLOBAL:
add(code.co_names[oparg >> 1])
break
elif opcode in FASTLOCAL_INSTRS:
if oparg < code.co_nlocals:
add(code.co_varnames[oparg])
break
oparg -= code.co_nlocals
n_deref = len(code.co_cellvars)
if oparg < n_deref:
add(code.co_cellvars[oparg])
break
oparg -= n_deref
# if it isn't a local var or a cell var,
# it's a free var
add(code.co_freevars[oparg])
break
elif opcode == LOAD_ATTR:
add('.' + code.co_names[oparg >> 1])
if not qname_list:
raise ValueError(f"No name found for value: {val!r}")
return "".join(qname_list)

x = "hello"
y = "hello"
print(f"{varname(x)=}, {varname(y)=}")

[night quarry]

@quartz wave :white_check_mark: Your 3.12 eval job has completed with return code 0.

varname(x)='x', varname(y)='y'

[quartz wave]

this should fix the problem

[quartz wave]

also added support for 3.12+ super() ```pycon

class A:
... a=lambda _:None
...
class B(A):
... def a(self):
... print(varname(self.a))
... print(varname(super().a))
... print(varname(super(B, self).a))
...
B().a()
self.a
super().a
super(B, self).a

[restive void]

Nice!

[modest bough]

pretty cool

[quartz wave]

i like this ```pycon

varname(varname.code.co_code.len.text_signature.add)
'varname.code.co_code.len.text_signature.add'

[craggy ruin]

you know you've got some good code when vscode thinks there's undeclared variables but there isnt

[unique heath]

how does it-

time comlpexity: O(n^999999)

lower bound

[candid light]

w h a t

[karmic pumice]

!e

open("_.txt","w").write(str(((-~int().__add__(-~int())).__pow__(int()).__add__(((-~int().__add__(-~int())).__pow__(-~int().__add__(-~int())))).__add__(((-~int().__add__(-~int())).__pow__(-~int().__add__(-~int()).__add__(-~int()).__add__(-~int()).__add__(-~int()).__add__(-~int())))))))

[night quarry]

@karmic pumice :warning: Your 3.12 eval job has completed with return code 0.

[No output]

[karmic pumice]

hmh when are the files printed

!e

f=open("_.txt","w")
f.write(str(((-~int().__add__(-~int())).__pow__(int()).__add__(((-~int().__add__(-~int())).__pow__(-~int().__add__(-~int())))).__add__(((-~int().__add__(-~int())).__pow__(-~int().__add__(-~int()).__add__(-~int()).__add__(-~int()).__add__(-~int()).__add__(-~int())))))))
f.close()

[night quarry]

@karmic pumice :warning: Your 3.12 eval job has completed with return code 0.

[No output]

[karmic pumice]

[fleet bridge]

!e open('test.py','wt').write('1/0')

[night quarry]

@fleet bridge :white_check_mark: Your 3.12 eval job has completed with return code 0.

Files with disallowed extensions can't be uploaded: .py

[fleet bridge]

!e open('test.txt','wt').write('1/0')

[night quarry]

@fleet bridge :white_check_mark: Your 3.12 eval job has completed with return code 0.

test.txt

1/0

[fleet bridge]

[covert rivet]

Wait the bot can handle file operations?

!e

with open("test.txt", 'w') as f:
    f.write("Wow...")

[night quarry]

@covert rivet :white_check_mark: Your 3.12 eval job has completed with return code 0.

test.txt

Wow...

[covert rivet]

damn

[night quarry]

Bot's GitHub Repository

Repository

Go to GitHub

[long fulcrum]

there has to be a shorter way of doing this

string.encode().hex(" ",-2)```

[dusky ice]

!d bytes.hex

[night quarry]

bytes.hex

hex([sep[, bytes_per_sep]])```
Return a string object containing two hexadecimal digits for each byte in the instance.

```py
>>> b'\xf0\xf1\xf2'.hex()
'f0f1f2'
```  If you want to make the hex string easier to read, you can specify a single character separator *sep* parameter to include in the output. By default, this separator will be included between each byte. A second optional *bytes\_per\_sep* parameter controls the spacing. Positive values calculate the separator position from the right, negative values from the left...

[dusky ice]

!e
print(b"123".hex(' ',-2))

[night quarry]

@dusky ice :white_check_mark: Your 3.12 eval job has completed with return code 0.

3132 33

[long fulcrum]

thats what i do already

[dusky ice]

its uncommon, usually you split it among each byte

why 2 bytes?

[long fulcrum]

its meant to be like a hex viewer

like xxd

[fleet bridge]

it is always not very convenient
splitting by every 2/4/8 byte creates visual separation and you can calculate stuff quicker

[versed eagle]

its also useful for when you're working with 2/4/8 byte sized things

[bright valve]

how would you guys go for the shortest code to count letters and digits in a str?

[clear venture]

>>> sum(map(str.isalnum, 'apples: 5\nbananas: 10'))
16

[bright valve]

I need 2 vars

or a list with 2 vals

representing numbers and letters

[quartz wave]

A=B=0
for c in s:A+='/'<c<':';B+=c.isalpha()
l=[A,B]

[bright valve]

huh

let me process that

[karmic pumice]

f=lambda s:(lambda g:(g(str.isalpha),g(str.isdigit)))(lambda h:sum(map(h,s)))

Id do something like that idk

Can't actually test on phone lmao

[bright valve]

A+='/'<c<':' can you explain this?

[quartz wave]

'0'<=c<='9' but using the character before '0' ('/') and the character after '9' (':') to make a shorter comparison

[bright valve]

so it is comparing with the ascii number?

[quartz wave]

yea i guess

[bright valve]

yo this is smart

thanks cereal

[karmic pumice]

Not an expression 😭

[bright valve]

it was for a clash of code challenge

I just did a basic if else

[karmic pumice]

Can improve by moving g to default params

[bright valve]

but wanted to know better ways

[karmic pumice]

Typing on phone sucks

[quartz wave]

[*map(sum,zip(*[('/'<c<':',c.isalpha())for c in s]))]

happy?

[karmic pumice]

Yes

[quartz wave]

okie :3

[karmic pumice]

f=lambda s,g=lambda h,s:sum(map(h,s)):(g(str.isalpha,s),g(str.isdigit,s))

Does that work

[clear venture]

you can use !e here

[quartz wave]

mhm

[karmic pumice]

The word lambda is so long

[clear venture]

petition to add keyword reassignment to 3.13

l = lambda

[karmic pumice]

λ=lambda

[quartz wave]

depending on the constraints this can be shortened further

[karmic pumice]

Would be really neat if the outer s argument leaked into g

f=lambda s,g=lambda h:sum(map(h,s)):(g(str.isalpha),g(str.isdigit))

Like that

[quartz wave]

found u >:3

[bright valve]

haha

im grinding leaderboard

rn im pos 696

would be cool a coc event in here

[karmic pumice]

Imagine

f=λs,g=λh:Σ(map(h,s)):(g(str.isalpha),g(str.isdigit))

That would be crazy

[bright valve]

lol

that would be cool

[karmic pumice]

Imagine str.(isalpha,isdigit)

As like x.(y,z) == (x.y,x.z)

If I ever to get to making a language it will absolutely have all of the crazy stuff I can think of

[quartz wave]

i implemented that in python once...

[karmic pumice]

Codec/ast manipulation?

Or forked interpreter

[quartz wave]

forked interp

YISJ

i beat the higher ranksroaweijhfrij >:D

m the fastest in fastest mode

[bright valve]

want to play a shortest clash of code @quartz wave?

[quartz wave]

mayb

after this on

[bright valve]

👌

fuck I missed an opportunity of using while else

[bright valve]

print(sum(map(lambda x:int(x,16),input())))```
any way to shorten this?

idk if theres a way to remove the lambda

[quartz wave]

somewhat late buut ```py
print(sum(int(x,16)for x in input()))

[bright valve]

dw i wasnt submiting it

looks clean ngl

[quartz wave]

btww what rank r u?

[bright valve]

g78

678

[magic dust]

wouldn't this work as well?

print(sum(int(x, 16) for x in input().split()))

now i think about it not really a shorter way, i need to go to bed 🥱

[quartz wave]

nop

works differently

[quartz wave]

nic

[bright valve]

that split is indiferent

because the string is just letters

or numbers

1a56d5c f.e.

[magic dust]

figured so.

[wheat river]

It would put it in a list and now when you convert the hex, you would get the whole hex string as an int, instead of each char of the hex string as int

[quartz wave]

effectively just print(int(input(),16))

[wheat river]

Yea

[bright valve]

but

that would print the hex value

not the sum of all singular values

!e

print(int("aa", 16))
print(sum(int(x, 16) for x in "aa"))

[night quarry]

@bright valve :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 170
002 | 20

[quartz wave]

yis

we're talking about that

[quartz wave]

.

[quartz wave]

ohmygo i figurd out the distance between letters thing

[bright valve]

wait

which challenge

I think I got the same one

AB = C first example?

[quartz wave]

mhm

i was on the same game as u

:3

[bright valve]

wait

what's your name

you followed me?

yo I got sub 600

[quartz wave]

yupp

[quartz wave]

congrats!!

[bright valve]

@quartz wave are you here?

[versed eagle]

multiargument lambdas 😔

[karmic pumice]

just @curry smh

[versed eagle]

?

that's not what i was talking about lol

[long fulcrum]

how can i make this shorter

try:
  if a[0]<2:
    s="1💎 1🥇 ";s+=f"{a[1]}🥈 "
    if a[1]<2:s+=f"{a[2]}🥉"
  else:s=f"{a[0]}🥇 "
  if a[0]==2:s+=f"{a[1]}🥉"
except:s
print(s)```

[proper vault]

some inline ifs perhaps

[long fulcrum]

i cant figure out how to move anything onto another line

[magic wraith]

def some_func():
    ...
my_code = some_func.__code__``` Suppose I have `my_code` and not `some_func`, and that `my_code`'s name and filename are useless, and co_argcount is zero. How can I know if `my_code` comes from a function or a module (like a pyc), other than looking at the bytecode and seeing if there are STORE_FASTs or similar?

[long fulcrum]

is it possible to golf this

n=a=0
v=[]```

contents of v dont matter

just has to be a list

[magic wraith]

I doubt it, the only thing i can think of is this:py n=a=0;v=[] n,=a,=v=[0]but it is one byte longer

[fleet bridge]

look also at type of a code object
module level code object cannot be a generator, for example

[magic wraith]

in my case I only should be able to differentiate between function code and normal code (that you get by calling compile for example)
by type of code object do you mean type() ? isn't it always types.CodeType ?

[fleet bridge]

also look at returns
if non-None is returned - it isn't a module code
and i guess in module codes return should happen only once at the end, but im not sure (optimizer and compiler can do something crazy)

[fleet bridge]

no, not type()
some flag inside of code object that says that it is not a regular function

!d types.CodeType

https://docs.python.org/3/reference/datamodel.html#index-58

[magic wraith]

bruh I wish i knew this existed before

[fleet bridge]

look at co_flags and https://docs.python.org/3/library/inspect.html#code-objects-bit-flags

[fleet bridge]

"Data model" page is the most important and helpful one across entire documentation

[magic wraith]

co_flags are insanely helpful, especially the CO_NESTED and CO_NEWLOCALS which seem to be always set for inner functoins and functions

tyvm

[quartz wave]

i don't think so, no ```py
n=a=0;v=[]
*v,n=a,=0,
a,*v,n=0,0

[gleaming linden]

if cond:s+=... -> s+=...*cond

[clear venture]

any ideas on ways to golf a one-line try/(empty) except using context manager?

!e

with type('',(),{'__enter__': lambda self: None, '__exit__': lambda self, *args: True})(): print('1') or 1/0

[night quarry]

@clear venture :white_check_mark: Your 3.12 eval job has completed with return code 0.

1

[rugged sparrow]

!e I wrote this a while back ```py
_try = lambda t,*a,f=lambda a:a,e=Exception,**k:(r:={}).pop('r',type('',(import('contextlib').ContextDecorator,),{'enter':int,'exit':lambda s,*a:isinstance(a[1],e)and[r.update(r=f(a))]})()(t)(*a,**k))

print(_try(lambda:1/0))``` which lets you pass arguments to your try block via *args and **kwargs, and lets you specify an except block via f. the return value is either the return value of the try block if it succeeds, or the return value of the except block. It also allows for scoping the catch via e which can be a single Exception subclass or a tuple of subclasses

[night quarry]

@rugged sparrow :white_check_mark: Your 3.12 eval job has completed with return code 0.

(<class 'ZeroDivisionError'>, ZeroDivisionError('division by zero'), <traceback object at 0x7f4b2b65c800>)

[rugged sparrow]

i haven't looked back into it to see if it could be golfed further since then

[surreal wasp]

this is what happens after i type py into the command terminal

[karmic pumice]

this is exactly whats supposed to happen

and this is not what #esoteric-python is for :d, its for cursed code

this is whats called a python repl - read eval print loop, you can enter python stuff in it and it will get evaluated

if you want to run a python file, then you can do py name_of_file.py (obviously replace the name of the file with one that you have)

[surreal wasp]

ok it seems python has chamged dramaticlly from a couple yerars ago. thank you

[karmic pumice]

wdym?

thats what was happening since like python 1

maybe you're confusing it with a python editor? like pycharm

[surreal wasp]

it used to bring up a virtual enviroment with diffrent tools but this isnt it

[karmic pumice]

[surreal wasp]

oh you know what pycharm may be it.

[versed eagle]

no it hasnt

[surreal wasp]

yep it was pycharm. thanks for eduring my awful memory lol

[modest bough]

cute

[north wraith]

At one point I was in 38th place, just 1 year ago

[bright valve]

Got sub 500 and just stopped

[candid grove]

yall think my code is obfuscated enough?

class _:
    _ = " + "
    class __:
        _ = "1"
        class ___:
            _ = ")"
            class ____:
                _ = "("
                class _____:
                    _ = "print"
                    class ______:
                        _ = " = "
                        class _______:
                            _ = "'"
                            class ________:
                                _ = "str"
                                class _________:
                                    _ = "int"
                                    class __________:
                                        _ = "eval"
                                        class ___________:
                                            _ = "f"
                                            class ____________:
                                                _ = "{"
                                                class _____________:
                                                    _ = "}"


exec(f"{_.__.___.____._____._}{_.__.___.____._}{_.__.___.____._____.______._______._}{_.__._}{_._}{_.__._}{_.__.___.____._____.______._}"+eval(f'{_.__.___.____._____.______._______.________._________.__________._}{_.__.___.____._}{_.__.___.____._____.______._______.________._________.__________.___________._}"{_.__.___.____._____.______._______.________._________.__________._}{_.__.___.____._}{_.__.___.____._____.______._______.________._________.__________.___________._}{_.__.___.____._____.______._______._}{_.__.___.____._____.______._______.________._}{_.__.___.____._}{_.__.___.____._____.______._______.________._________._}{_.__.___.____._}_.__._{_.__.___._}{_._}{_.__.___.____._____.______._______.________._________._}{_.__.___.____._}_.__._{_.__.___._}{_.__.___._}{_.__.___.____._____.______._______._}{_.__.___._}"{_.__.___._}')+f"{_.__.___.____._____.______._______._}{_.__.___._}")

it outputs 1 + 1 = 2

[candid grove]

my brain hurts D:

btw if you're wondering exactly how I made it, I wrote the script, and then kept expanding it by stacked classes, until I thought it was complicated enough

well actually I don't really know how to make it more complicated but yk

[gleaming linden]

looks pretty simple to deobfuscate

[unique heath]

yeah

this isnt dunder spam but its under spam lool

[fleet bridge]

i kinda like that

woah, i have a crazy idea

[night quarry]

Nope.

Sorry, you can only do that in #bot-commands!

[fleet bridge]

I'll do it later

[quartz wave]

after running my renaming script ```py
class A:
A = ' + '

class B:
    A = '1'

    class C:
        A = ')'

        class D:
            A = '('

            class E:
                A = 'print'

                class F:
                    A = ' = '

                    class G:
                        A = "'"

                        class H:
                            A = 'str'

                            class I:
                                A = 'int'

                                class J:
                                    A = 'eval'

                                    class K:
                                        A = 'f'

                                        class L:
                                            A = '{'

                                            class M:
                                                A = '}'

exec(f'{A.B.C.D.E.A}{A.B.C.D.A}{A.B.C.D.E.F.G.A}{A.B.A}{A.A}{A.B.A}{A.B.C.D.E.F.A}' + eval(f'{A.B.C.D.E.F.G.H.I.J.A}{A.B.C.D.A}{A.B.C.D.E.F.G.H.I.J.K.A}"{A.B.C.D.E.F.G.H.I.J.A}{A.B.C.D.A}{A.B.C.D.E.F.G.H.I.J.K.A}{A.B.C.D.E.F.G.A}{A.B.C.D.E.F.G.H.A}{A.B.C.D.A}{A.B.C.D.E.F.G.H.I.A}{A.B.C.D.A}.__.{A.B.C.A}{A.A}{A.B.C.D.E.F.G.H.I.A}{A.B.C.D.A}.__.{A.B.C.A}{A.B.C.A}{A.B.C.D.E.F.G.A}{A.B.C.A}"{A.B.C.A}') + f'{A.B.C.D.E.F.G.A}{A.B.C.A}')

bottom part becomes ```py
exec("print('1 + 1 = " + eval('eval(f"eval(f'str(int(.__.) + int(.__.))')")') + "')")

nice attempt though

[tulip harbor]

Why isn't integer division faster than float division?

%timeit 101 / 2
4.91 ns ± 0.0026 ns per loop (mean ± std. dev. of 7 runs, 100000000 loops each)

 %timeit 101 // 2
4.91 ns ± 0.00175 ns per loop (mean ± std. dev. of 7 runs, 100000000 loops each)

[clear venture]

it's not integer division, it's floor division. Which I assume has more steps

x // y

floored quotient of x and y

Also referred to as integer division. For operands of type int, the result has type int. For operands of type float, the result has type float. In general, the result is a whole integer, though the result’s type is not necessarily int. The result is always rounded towards minus infinity: 1//2 is 0, (-1)//2 is -1, 1//(-2) is -1, and (-1)//(-2) is 0.

[unique heath]

although i'd assume a smart compiler would be able to optimize that more

[karmic pumice]

a smart compiler would precompute everything thats not IO

[modest bough]

it's so pretty

[earnest wing]

nice interpreter you got there

[unique heath]

python does that though?

>>> dis.dis("30 / 5")
  0           0 RESUME                   0

  1           2 RETURN_CONST             0 (6.0)```

(well kinda)

[earnest wing]

it doesn't precompute "everything that's not IO"

[quartz wave]

that's precomputed

if you want a non-precomputed one you could store them in variables and time only the division of those variables

but i doubt floor division will be faster

[modest skiff]

I sent this to non-coding server, they were confused and it has hurt their eyes
||```py
exec(import('zlib').decompress(bytes.fromhex('789c2b28cacc2bd150ca48cdc9c95728cf2fca4951d204004e93071f')).decode())

[unique heath]

get me to that server immediately i need to show off

[quartz wave]

fun fact .decode() is not needed

[modest skiff]

we want it to be more confusing lol

[unique heath]

!e anyway more dunder spam lolpy __builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))((_:=__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()),__:=_.__add__(_).__lshift__(_.__add__(_).__add__(_).__lshift__(_)),"".__getattribute__("join")([__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__lshift__(_.__lshift__(_.__add__(_))).__sub__(_.__add__(_))))(i) for i in [__.__sub__(_.__add__(_).__lshift__(_.__add__(_).__add__(_)).__add__(_.__add__(_).__lshift__(_.__add__(_)))), 101, 108, 108, 111, _.__add__(_).__add__(_).__add__(_).__lshift__(_.__add__(_).__add__(_)), 119, 111, 114, 108, 100]])).__getitem__(_.__add__(_)))

[night quarry]

@unique heath :white_check_mark: Your 3.12 eval job has completed with return code 0.

hello world

[stark granite]

is it possible

to write such program or class

[unique heath]

to

[stark granite]

which returns a random instance which just exists (even if not in scope via any measures)

eacb time

you request its type

[unique heath]

class InstanceGetter:
    def get():
        return None```

[stark granite]

no randomly selected at runtime

not random programmer's choice...

[unique heath]

https://xkcd.com/221/

xkcd: Random Number

RFC 1149.5 specifies 4 as the standard IEEE-vetted random number.

[stark granite]

ye thats what i was referring to

[quartz wave]

here's a real implementation btw ```py
import gc, random
class InstanceGetter:
def get():
return random.choice(gc.get_objects())

it runs in about 88 microseconds

which is pretty fast

[rugged sparrow]

!e ```py
import gc, random
class Instance:
def new(cls):
return random.choice(gc.get_objects())

for i in range(4):print(Instance())```

[night quarry]

@rugged sparrow :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | <member 'st_ctime_ns' of 'os.stat_result' objects>
002 | {'__new__': <built-in method __new__ of type object at 0x559c290c3200>, '__repr__': <slot wrapper '__repr__' of '_thread._ExceptHookArgs' objects>, '__reduce__': <method '__reduce__' of '_thread._ExceptHookArgs' objects>, 'exc_type': <member 'exc_type' of '_thread._ExceptHookArgs' objects>, 'exc_value': <member 'exc_value' of '_thread._ExceptHookArgs' objects>, 'exc_traceback': <member 'exc_traceback' of '_thread._ExceptHookArgs' objects>, 'thread': <member 'thread' of '_thread._ExceptHookArgs' objects>, '__doc__': 'ExceptHookArgs\n\nType used to pass arguments to threading.excepthook.', '__module__': '_thread', 'n_sequence_fields': 4, 'n_fields': 4, 'n_unnamed_fields': 0, '__match_args__': ('exc_type', 'exc_value', 'exc_traceback', 'thread')}
003 | <class '_io.BytesIO'>
004 | <member 'releaselevel' of 'sys.version_info' objects>

[quartz wave]

there's also a 1 in i-don't-know chance where it returns the class itself

[fleet bridge]

iirc, not all objects are stored in gc.get_objects()
some non-gc-tracked objects will never be returned, such as ints, strings, some tuples, ...

[fleet bridge]

1 in several thousands

on 3.12 repl with fancy stuff: ```py

len(gc.get_objects())
24_934

on clean 3.7 repl: ```py
>>> len(gc.get_objects())
4518

clean 3.7, no repl:

>py -3.7 -c "print(len(__import__('gc').get_objects()))"
4499

[quartz wave]

well it'd be hard to retrieve them all anyway

searching every tracked object for untracked objects and adding them all to a list that varies every time it's called is pretty costly

[static knoll]

!e

[night quarry]

Missing required argument

code

[static knoll]

oh

[sick hound]

u=[10]+[0]*19
m=0
def _():print(chr(u[m]),end="")
while u[m]:  
    m+=1       
    u[m]+=7
    m+=1       
    u[m]+=10
    m+=1       
    u[m]+=3
    m+=1
    u[m]+=1
    m-=4
    u[m]-=1
m+=1;m-=1;m+=1;u[m]+=2;(_:=_,_());m+=1;m-=1;m+=1;u[m]+=1;(_:=_,_())
u[m]+=4;u[m]+=2;u[m]+=1;(_:=_,_());(_:=_,_());u[m]+=2;u[m]+=1
_();m+=1;m-=1;m+=1;u[m]+=2;(_:=_,_());m-=2;u[m]+=15
_();m+=1;m-=1;m+=1;(_:=_,_());u[m]+=3;(_:=_,_());u[m]-=6;(_:=_,_())
u[m]-=8;(_:=_,_());m+=1;m-=1;m+=1;u[m]+=1;(_:=_,_());m+=1;m-=1;m+=1;(_:=_,_())

[stark granite]

interesting

[unique heath]

is there a short way to obfuscate a for loop?

[low lynx]

turn it into an ast node

it'll largely depend on what is in the for loop

[unique heath]

like

x(i) for i in ...

[low lynx]

just map?

[unique heath]

oh shit im dumb

[low lynx]

idk what u mean by obfuscate tho

[unique heath]

!e

__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))((_:=__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()),__:=_.__add__(_).__lshift__(_.__add__(_).__add__(_).__lshift__(_)),___:=_.__add__(_).__add__(_),_____:=__.__sub__(___.__add__(_).__add__(_).__lshift__(_.__add__(_))),_______:=__.__sub__(_.__add__(_).__lshift__(___).__sub__(_.__add__(_))),__name__.__getattribute__(__name__.__dir__().__getitem__(___.__lshift__(___)))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__lshift__(___.__add__(_)).__add__(___.__add__(_))))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__lshift__(_.__lshift__(_.__add__(_))).__sub__(_.__add__(_)))),[__.__sub__(_.__add__(_).__lshift__(___).__add__(_.__add__(_).__lshift__(_.__add__(_)))), _______.__sub__(_.__add__(_).__lshift__(___)).__add__(___), _____, _____,_______.__sub__(___),_.__add__(___).__lshift__(___),__.__sub__(_.__lshift__(___)).__sub__(_), _______.__sub__(_).__sub__(_).__sub__(_),_______,_______.__sub__(_.__add__(_).__add__(_).__lshift__(_)), __.__sub__(_.__add__(_).__add__(_).__add__(_).__mul__(_.__add__(_).__add__(_).__add__(_).__add__(_).__add__(_).__add__(_)))]))).__getitem__(_.__sub__(_).__sub__(_)))```

[night quarry]

@unique heath :white_check_mark: Your 3.12 eval job has completed with return code 0.

h__main__e__main__l__main__l__main__o__main__ __main__w__main__o__main__r__main__l__main__d

[unique heath]

what

the cow

!e py __builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))((_:=__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()),__:=_.__add__(_).__lshift__(_.__add__(_).__add__(_).__lshift__(_)),____:=_.__add__(_),___:=____.__add__(_),_____:=__.__sub__(___.__add__(_).__add__(_).__lshift__(_.__add__(_))),_______:=__.__sub__(_.__add__(_).__lshift__(___).__sub__(_.__add__(_))),__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__add__(_).__lshift__(___.__add__(_)).__sub__(___.__add__(_).__add__(_))))().__getattribute__(__name__.__dir__().__getitem__(___.__lshift__(___)))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__lshift__(___.__add__(_)).__add__(___.__add__(_))))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__lshift__(_.__lshift__(_.__add__(_))).__sub__(____))),[__.__sub__(____.__lshift__(___).__add__(____.__lshift__(____))), _______.__sub__(_.__add__(_).__lshift__(___)).__add__(___), _____, _____,_______.__sub__(___),_.__add__(___).__lshift__(___),__.__sub__(_.__lshift__(___)).__sub__(_), _______.__sub__(___),_______,_______.__sub__(___.__lshift__(_)), __.__sub__(___.__add__(_).__mul__(___.__add__(___).__add__(_)))]))).__getitem__(_.__sub__(____)))

[night quarry]

@unique heath :white_check_mark: Your 3.12 eval job has completed with return code 0.

hello world

[unique heath]

sorry for dunder spam

[unique heath]

.bm

[quartz wave]

i should make a deobfuscator for this

[unique heath]

plz dint

[quartz wave]

b-but...

[unique heath]

okay fine

[quartz wave]

i can't read it :(

yaiyyyy :D

[unique heath]

yay

[karmic pumice]

def deobfuscate_pathlib_shill(text: str) -> str:
  return 'print("hello world")'

[unique heath]

damn

genius

[karmic pumice]

😋

its correct most of the time

[unique heath]

__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))((_:=__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()),__:=_.__add__(_).__lshift__(_.__add__(_).__add__(_).__lshift__(_)),___:=_.__add__(_).__add__(_),_____:=__.__sub__(___.__add__(_).__add__(_).__lshift__(_.__add__(_))),_______:=__.__sub__(_.__add__(_).__lshift__(___).__sub__(_.__add__(_))),__name__.__getattribute__(__name__.__dir__().__getitem__(___.__lshift__(___)))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__lshift__(___.__add__(_)).__add__(___.__add__(_))))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__lshift__(_.__lshift__(_.__add__(_))).__sub__(_.__add__(_)))),[__.__sub__(_.__add__(_).__lshift__(___).__add__(_.__add__(_).__lshift__(_.__add__(_)))), _______.__sub__(_.__add__(_).__lshift__(___)).__add__(___), _____, _____,_______.__sub__(___),_.__add__(___).__lshift__(___),__.__sub__(_.__lshift__(___)).__sub__(_), _______.__sub__(_).__sub__(_).__sub__(_),_______,_______.__sub__(_.__add__(_).__add__(_).__lshift__(_)), __.__sub__(_.__add__(_).__add__(_).__add__(_).__mul__(_.__add__(_).__add__(_).__add__(_).__add__(_).__add__(_).__add__(_)))]))).__getitem__(_.__sub__(_).__sub__(_)))
``` deobfuscation algorithm fail

also @quartz wave

[quartz wave]

hm?

[unique heath]

can you add the imsosorry C thing to the letsbuilda org?

[quartz wave]

i downt kno how

[unique heath]

just send

i have write perms

to letsbuilda

[quartz wave]

owh

okie

[unique heath]

although i dont think doggo would like it

@mossy plume DOGGO IMSOSORRY BUT C???????

and we wait

[quartz wave]

c file
and
needed header

[modest skiff]

hope the bot doesn't detect it as spam lol

[ruby raft]

hi if I have two files
mod.py

import inspect
print(inspect.stack[-1].lineno)```
`test.py`
```py
#Some random file that imports the module
import mod

print("a")```
I can get the line number of the `import mod` (2) right, but can I also get the column number? (python 3.9)

I know that the code_context property exists but if I can prevent searching through strings that would be great

[proper vault]

3.9 I don't believe tracked that. Short of writing a custom parser (maybe just finding it via .index on the right line) and running that on getsource, it is probably impossible

[ruby raft]

cool

well I guess I should state my original problem in case there are other solutions:
I would like to preprocess the __main__ file from mod.py and then let it execute
Problem is that if there is code before the import mod statement it obviously should not be executed again. What makes it so painful is that I cant just say: "Give me everything after the import statement, I will execute the rest" because it could be nested in some class or function

Do you have any idea?

I would do the preprocessing too for other files imported but for those its easier because I can set up import hooks

My alternative would be somehow using the site module to my advantage but Im not sure if this is a great solution

Or I could use a custom encoding combined with import hooks so all I need to do is set the encoding to the custom one in the __main__ file

which is much more explicit than going to the site module and preprocessing any python file that is run, even if it has nothing to do with my project

[mossy plume]

works for me

[fluid pumice]

!e

listA:list=[0.0,0.5];b:int=round((listA[0]+0.6)+(listA[1]+1));print("{:^20}".format(b))

[night quarry]

@fluid pumice :white_check_mark: Your 3.12 eval job has completed with return code 0.

         2          

[long fulcrum]

is this the shortest way to do this?

import sys
v=sys.argv[1:]```

[unique heath]

ok doggo heres the code #esoteric-python message

[fluid pumice]

Is this considered esoteric?

[karmic pumice]

from sys import*;v=argv[1:]

probally

import sys;v=sys.argv[1:]
from sys import*;v=argv[1:]

yeah yours one is shorter

[quartz wave]

import sys;v=sys.argv[1:]
import sys;_,*v=sys.argv

[long fulcrum]

oh fuck

thats smart

ty

[karmic pumice]

that doesnt work does it

[fluid pumice]

idk

[karmic pumice]

__import__ is not the same as import, it wont bring the name into the scope, it returns the module object

[fluid pumice]

yeah ik

I just tried smth

[karmic pumice]

v=__import__('sys').argv[1:] 🤷

[fluid pumice]

oh yeah

ty

[karmic pumice]

thats sadly not shorter

the unders and quote marks take quite some space

[fluid pumice]

hmmm

it's not shorter, but still looks cool tho

LMAO

v=(lambda:__import__('sys').argv[1:])

not so short

and useless

[long fulcrum]

it does

[fluid pumice]

nono, he said that to me

I deleted the message

[long fulcrum]

ah

[karmic pumice]

it doesnt even do the same thing, its a function now, you didnt call it

[fluid pumice]

__import__('sys').argv[1:]

[fluid pumice]

I know what lambda does lol, I mentioned it's useless

[karmic pumice]

v=(lambda f:f(*map(__import__,f.__code__.co_varnames)))(lambda sys:sys.argv[1:])

[fluid pumice]

fu#k you ||(no offense please, this is a joke)||

[unique heath]

__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))((_:=__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()),__:=_.__add__(_).__lshift__(_.__add__(_).__add__(_).__lshift__(_)),____:=_.__add__(_),___:=____.__add__(_),_____:=__.__sub__(___.__add__(_).__add__(_).__lshift__(_.__add__(_))),_______:=__.__sub__(_.__add__(_).__lshift__(___).__sub__(_.__add__(_))),__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__add__(_).__lshift__(___.__add__(_)).__sub__(___.__add__(_).__add__(_))))().__getattribute__(__name__.__dir__().__getitem__(___.__lshift__(___)))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__lshift__(___.__add__(_)).__add__(___.__add__(_))))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__lshift__(_.__lshift__(_.__add__(_))).__sub__(____))),[__.__sub__(____.__lshift__(___).__add__(____.__lshift__(____))), _______.__sub__(_.__add__(_).__lshift__(___)).__add__(___), _____, _____,_______.__sub__(___),_.__add__(___).__lshift__(___),__.__sub__(_.__lshift__(___)).__sub__(_), _______.__sub__(___),_______,_______.__sub__(___.__lshift__(_)), __.__sub__(___.__add__(_).__mul__(___.__add__(___).__add__(_)))]))).__getitem__(_.__sub__(____)))```

i had to

[quartz wave]

v=(lambda f:f(*map(next(j for f in object.__subclasses__() for j in [getattr(getattr(f, '__init__', object), '__globals__', 0)] if j != 0).__getitem__,f.__code__.co_varnames)))(lambda sys:sys.argv[1:])

[unique heath]

!e

v=(lambda f:f(*map(next(j for f in object.__subclasses__() for j in [getattr(getattr(f, '__init__', object), '__globals__', 0)] if j != 0).__getitem__,f.__code__.co_varnames)))(lambda sys:sys.argv[1:])()

print("hi")```

[night quarry]

@unique heath :x: Your 3.12 eval job has completed with return code 1.

001 |   File "/home/main.py", line 1
002 |     v=(lambda f:f(*map(next(j for f in object.__subclasses__() for j in [getattr(getattr(f, '__init__', object), '__globals__', 0)] if j != 0]).__getitem__,f.__code__.co_varnames)))(lambda sys:sys.argv[1:])()
003 |                                                                                                                                              ^
004 | SyntaxError: closing parenthesis ']' does not match opening parenthesis '('

[quartz wave]

1. i edited it
2. it's already called

[unique heath]

!e

v=(lambda f:f(*map(next(j for f in object.__subclasses__() for j in [getattr(getattr(f, '__init__', object), '__globals__', 0)] if j != 0).__getitem__,f.__code__.co_varnames)))(lambda sys:sys.argv[1:])

print('ono')````

[night quarry]

@unique heath :white_check_mark: Your 3.12 eval job has completed with return code 0.

ono

[quartz wave]

!e ```py
v=(lambda f:f(*map(next(j for f in object.subclasses() for j in [getattr(getattr(f, 'init', object), 'globals', 0)] if j != 0).getitem,f.code.co_varnames)))(lambda sys:sys.argv)
print(v == import('sys').argv)

[night quarry]

@quartz wave :white_check_mark: Your 3.12 eval job has completed with return code 0.

True

[quartz wave]

ok

[unique heath]

lmao

[fluid pumice]

!e

a:list=[float(chr(int(b'\xff\xfe4\x009\x00'.decode('utf-16')))),float(chr(int(b'\xff\xfe4\x009\x00'.decode('utf-16'))))];d:float=(++a[0]+--a[1]);print("{:<10}".format("{}".format(int(d))))

[night quarry]

@fluid pumice :white_check_mark: Your 3.12 eval job has completed with return code 0.

2         

[quartz wave]

deobfuscated a bit ```py

py dundeobfuscator.py
builtins.print(((A := 1), (B := 128), (C := 2), (D := 3), (E := 108), (F := 114), ''.join(builtins.map(builtins.chr, [104, 101, 108, 108, 111, 32, 119, 111, 114, 108, 100])))[-1])

[unique heath]

:(

[quartz wave]

this too ```py

py dundeobfuscator.py
builtins.print(((A := 1), (B := 128), (C := 3), (D := 108), (E := 114), 'main'.join(builtins.map(builtins.chr, [104, 101, 108, 108, 111, 32, 119, 111, 114, 108, 100])))[-1])

[unique heath]

:(

[restive void]

!e

def foo():
    yield 42
    print("finishing")
bar = foo()
print(next(bar))
print("before")
[] = bar
print("after")

[night quarry]

@restive void :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 42
002 | before
003 | finishing
004 | after

[stark granite]

huh?

why is it called

in [] = bar

[karmic pumice]

Iterating for unpacking like in [x]=[1] ig

[rapid elk]

hi

[ruby raft]

why does the codecs IncrementalDecoder seem to work perfectly fine when opening file with a custom encoding but struggles when importing a file with a # coding: some encoding

it always ends in some recursive loop

[quartz wave]

something with the code?

[ruby raft]

You mean the code that is imported?

Just some simple print statements

Funnily enough in termux on my phone the IncrementalDecoder isnt even called when importing a file lol

[long fulcrum]

for i in range(100):print(i%3//2*'Fizz'+i%5//4*'Buzz'or-~i)```

what?

!e

for i in range(100):print(i%3//2*'Fizz'+i%5//4*'Buzz'or-~i)```

[night quarry]

@long fulcrum :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 1
002 | 2
003 | Fizz
004 | 4
005 | Buzz
006 | Fizz
007 | 7
008 | 8
009 | Fizz
010 | Buzz
011 | 11
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/ANMLEQB36XDODHYQDIHFGTAGLU

[hard spoke]

!e there's a nice book, "ten essays on fizzbuzz". a favorite of mine from there, which technically fullfills your condition, is

import random
random.seed(23_977_775)
for i in range(1, 16):
    print(random.choice(["FizzBuzz", "Fizz", str(i), "Buzz"]))

[night quarry]

@hard spoke :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 1
002 | 2
003 | Fizz
004 | 4
005 | Buzz
006 | Fizz
007 | 7
008 | 8
009 | Fizz
010 | Buzz
011 | 11
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/HVZTR6CK4MG75GUDM2KW4QB4JY

[karmic pumice]

https://gist.github.com/gyu-don/3777bfa02d2e86cfa9fb64084f3c42dc

[long fulcrum]

does this work for any i?

[hard spoke]

nah, but one could modify it to just reuse the info on the first 15 elements for the rest of the sequence.

[long fulcrum]

aw

[fleet bough]

I've got a decorator for classes that adds logging

is there some functools wraps equivalent?

or maybe it should be a mixin?

it mostly extends the init method

[restive void]

Why doesn't functools.wraps cut it?

[fleet bough]

the class module sets set to zen_logging.loggify too, which is wrong

https://github.com/desultory/zenlib/blob/main/src/zenlib/logging/loggify.py

this is what i was doing before and it worked alright

[clear venture]

Lmao I love this, finding seeded rng for sequences of numbers or characters is also a cute idea for an obfuscation technique, makes the code literally unreadable without executing it.

[unique heath]

this is amazing

[clear venture]

I wonder if the seeding is version specific

!e

import random

a = [-3, 1, -5, 3, 5, -8, -6, -11, -2, 7]

random.seed(69_69_420)
b = ''.join(map(chr,[random.randint(0, 19) - a[i] + 100 for i in range(len(a))]))
print(b[:5],b[5:])

[night quarry]

@clear venture :white_check_mark: Your 3.12 eval job has completed with return code 0.

hello world

[clear venture]

Guess not

[unique heath]

!d random

[night quarry]

random

Source code: Lib/random.py

This module implements pseudo-random number generators for various distributions.

For integers, there is uniform selection from a range. For sequences, there is uniform selection of a random element, a function to generate a random permutation of a list in-place, and a function for random sampling without replacement.

On the real line, there are functions to compute uniform, normal (Gaussian), lognormal, negative exponential, gamma, and beta distributions. For generating distributions of angles, the von Mises distribution is available.

Similar names: django.random

[craggy ruin]

exec('''
for i in range(16):
    qv i % 3 == 0 and i % 5 == 0:
        print(\'fizzbuzz\')
    ykqv i % 3 == 0:
        print(\'fizz\')
    ykqv i % 5 == 0:
        print(\'buzz\')
    ykmy:
        print(i)
'''.translate(str.maketrans('kyvqm', 'lefis')))

[unique heath]

lol

that isnt really a "master" thing

its pretty trivial

for i in range(100):print(["fizzbuzz", "buzz", "fizz", i][bool(i % 3) + (bool(i % 5) * 2)])``` pretty neat solution

!e

for i in range(10):print(["fizzbuzz", "buzz", "fizz", i][bool(i % 3) + (bool(i % 5) * 2)])```

[night quarry]

@unique heath :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | fizzbuzz
002 | 1
003 | 2
004 | fizz
005 | 4
006 | buzz
007 | fizz
008 | 7
009 | 8
010 | fizz

[proper vault]

!e my favourite has been

for i in range(1, 17):
    q=[str(i),'']
    j=0
    if i % 3 == 0:
        q[j] = 'Fizz'
        j+=1
    if i % 5 == 0:
        q[j] = 'Buzz'
        j+=1
    print(''.join(q))

[night quarry]

@proper vault :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 1
002 | 2
003 | Fizz
004 | 4
005 | Buzz
006 | Fizz
007 | 7
008 | 8
009 | Fizz
010 | Buzz
011 | 11
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/A3WJ2Z77JDHOMLIU2DMPNWGA2I

[unique heath]

smart

[proper vault]

!e Not sure if this one was mentioned yet, but this is also a fun one

for i in range(1,33):
    print(['FizzBuzz', i, i, 'Fizz', i, 'Buzz', 'Fizz', i, i, 'Fizz', 'Buzz', i, 'Fizz', i, i, 'Buzz'][i%16])

[night quarry]

@proper vault :white_check_mark: Your 3.12 eval job has completed with return code 0.

001 | 1
002 | 2
003 | Fizz
004 | 4
005 | Buzz
006 | Fizz
007 | 7
008 | 8
009 | Fizz
010 | Buzz
011 | 11
... (truncated - too many lines)

Full output: https://paste.pythondiscord.com/VFWMTMGINH22IZ4MIQ6SN2FCAE

[unique heath]

its not

[unique heath]

so basically mine but worse /j

[sick hound]

@unique heath how did u learn py?

[arctic elm]

oh, fizz buzz

I have a very old C thing of mine

i;main(){101-++i&&printf("%d\n\0FizzBuzz\n\0Fizz\n"+(7*!(i%3)+4*!(i%5))%9*2,i)&main();}

a tad more readable form

i;
main() {
  101-++i
  && printf(
    "%d\n\0FizzBuzz\n\0Fizz\n"
    + (7*!(i%3)+4*!(i%5))%9*2,
    i
  ) & main();
}

[unique heath]

if the first one is '' (empty), it goes to the or
else, it doesnt

[unique heath]

here, and writing a shit ton of stupid projects

[sick hound]

what abt when u were a complete beginner ?

[unique heath]

i dont remember

[sick hound]

ah

[unique heath]

but also this

[sick hound]

i just bought a course

[unique heath]

why

[sick hound]

on udemy

it has rlly good ratings

[unique heath]

waste of money 😭

[sick hound]

so lets see

[unique heath]

programming courses are waste of money imo

!res

[night quarry]

Resources

The Resources page on our website contains a list of hand-selected learning resources that we regularly recommend to both beginners and experts.

[sick hound]

o

okay

[unique heath]

i suggest https://automatetheboringstuff.com which is a free ebook

[sick hound]

oo

ik that book

[unique heath]

and then make a shit ton of stupid projects

emphasis on shit ton

!e

__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__())).__add__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()))))((_:=__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()),__:=_.__add__(_).__lshift__(_.__add__(_).__add__(_).__lshift__(_)),____:=_.__add__(_),___:=____.__add__(_),_____:=__.__sub__(___.__add__(_).__add__(_).__lshift__(_.__add__(_))),_______:=__.__sub__(_.__add__(_).__lshift__(___).__sub__(_.__add__(_))),__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__add__(_).__lshift__(___.__add__(_)).__sub__(___.__add__(_).__add__(_))))().__getattribute__(__name__.__dir__().__getitem__(___.__lshift__(___)))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(___.__add__(_).__lshift__(___.__add__(_)).__add__(___.__add__(_))))(__builtins__.__getattribute__(__builtins__.__dir__().__getitem__(__builtins__.__dir__().__len__().__floordiv__(__name__.__dir__().__len__()).__lshift__(_.__lshift__(_.__add__(_))).__sub__(____))),[__.__sub__(____.__lshift__(___).__add__(____.__lshift__(____))), _______.__sub__(_.__add__(_).__lshift__(___)).__add__(___), _____, _____,_______.__sub__(___),_.__add__(___).__lshift__(___),__.__sub__(_.__lshift__(___)).__sub__(_), _______.__sub__(___),_______,_______.__sub__(___.__lshift__(_)), __.__sub__(___.__add__(_).__mul__(___.__add__(___).__add__(_)))]))).__getitem__(_.__sub__(____)))```

[night quarry]

@unique heath :white_check_mark: Your 3.12 eval job has completed with return code 0.

hello world

[unique heath]

so anyway

https://code.golf/collatz#python golf

hello world

(help i wrote that by hand)

[unique heath]

cereal made a deobfuscator here

it makes more sense when you see that

lmao

[quartz wave]

i'm improving it btw