esoteric-python

[fleet bridge]

if you have some immutable attributes of your instance, you can inline them into bytecode, so LOAD_NAME self; LOAD_ATTR x; becomes LOAD_CONST 'value';

[ashen quarry]

class Solution:
    def fib(self, n: int) -> int:
        # if (n == 0) or (n == 1):
        #     return n
        # else:
        #     c = [0, 1]
        #     for x in range(1, n):
        #         c.append(c[x-1] + c[x])
        #     return c[n]
        
        ##
        
        a = []
        for x in range(20):
            for y in range(20): 
                a.append(int(str(n)))
        b = True
        for x in range(20):
            b = False
            for y in range(20):
                if n in a:
                    b = False
                    if n in a:
                        b = True
                        if n in a:
                            b = False
                            if n in a:
                                b = True
                                if n in a:
                                    b = False
                                    if n in a:
                                        b = True
                                        if n in a:
                                            b = False
                                            if n in a:
                                                b = True
        b = False
        if b == False and b != True and (int(str(int(str(n)))) not in [3,4,5]):
            if b != True:
                if b == False:
                    if int(str(int(str(n)))) in [0,1]:
                            return int(str(int(str(n))))
        else:
            if b != False and b == True and (int(str(n)) not in [0,1]):
                for x in range(2):
                    c.append(x)
                for x in range(1,int(str(int(str(int(str(n))))))):
                    c.append(c[x-1+1-1+1*1-1]+c[x-1+1*1])
                return c[n]```

[fleet bridge]

It is possible to implement efficient currying by hardcoding first argument into bytecode.
So we can get rid of this ugly partial by hardcoding self into method

[fleet bridge]

If we can do that, hacks like this will be efficient:

l = [] # imagine it is not a list
la = l.append # inline self into append method
for i in range(100000000):
    la(i)

[versed eagle]

idk if it works on windows, but on Linux you can set a $PYTHONSTARTUP env variable, and it will run that python script before it does whatever else you tell it to do

so that might be helpful to whatever you're trying to do

[mint mural]

hey, how do i make a while true loop in 1 line, im golfing

[versed eagle]

{do_stuff() for i in iter(int,1)}

if you want a while true that runs forever

if you want to be able to break,

[mint mural]

yea, i've figured it out thank you

import time;solution = type("solution", (object, ), {"__init__": lambda self: setattr(self, "i", 0), "run": lambda self: setattr(self, "i", self.i + 1) or time.sleep(1) or print(self.i)})();{solution.run() for _ in iter(int,1)}

xD

[versed eagle]

_break=0;{(_break:=do_stuff()) for i in iter((lambda:_break),1)}

you can of course use conditionals to set break

and you can also have a variable for continuing

to get proper while loop behavior

[mint mural]

yea i get it, thank you

[versed eagle]

np, happy to help

[mint mural]

i was going to use something like this

for()in iter(list, 1): solution.run()

[versed eagle]

t=setattr;import time;s=type("solution",(),{"__init__":(lambda s:t(s,"i",0)),"run":(lambda s:t(s,"i",s.i+1)or time.sleep(1)or print(s.i))})();{s.run()for _ in iter(int,1)}

removed all the whitespace for you :P

[fleet bridge]

Doesnt work

[mint mural]

it does, i've tried it and it does work

[fleet bridge]

Ah, right

Hmm, im not sure

!e

for () in iter(list, 1): print(42)

[night quarry]

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

001 | 42
002 | 42
003 | 42
004 | 42
005 | 42
006 | 42
007 | 42
008 | 42
009 | 42
010 | 42
011 | 42
... (truncated - too many lines)

Full output: too long to upload

[mint mural]

told u

[versed eagle]

its a tuple of the looping vars that has 0 elements, so theres no looping variable
its the same concept as

for key, val in somedict.items():
    do_stuff()

where its a tuple of vars you're using to loop

[fleet bridge]

I know how it works

[versed eagle]

heres a random thingy i found
https://github.com/dutc/rwatch
basically, whenever a variable is accessed, it calls a function of your choosing with that variable as an argument, and the value the user gets returned to them is whatever that function returns

so thats fun

[old socket]

Maybe import hooks as well depending on how you want to run said code

[versed eagle]

exactly

you could probably also alias python to something like python -c "import sys; exec(sys.argv[1].replace("import", some_long_lamdba_that_acts_as_an_import_hook))"

[versed eagle]

thats pseudocode, but you probably get the idea

[old socket]

Had some fun with import hooks

It allowed me to run my languages source code into python while being a different language

[versed eagle]

ooh
do tell

what language was it

[old socket]

I made an interop language that transpiles to python bytecode

so it runs on the python vm

I just made the import hook read the file and then compile it into bytecode which then gets executed as a module

So that makes me able to run python code inside my language and also I can run my language inside of python

[versed eagle]

whats it called

[old socket]

Not public but looks something like this

[versed eagle]

ooh

fun

[old socket]

Yea super easy to make as well, all it does is transpile

[versed eagle]

cool

one day (when i find the time) im going to write a transpiler of c++ -> python
so that c++ can be interpreted
the problem is finding the time lol

[old socket]

Oh yea It took me some trial and error

That was because I didn't plan it out very well, so this time I added models for all python ast nodes then slowly implemented them all one by one

It actually took me less time to do that, than It did for me to figure out how to syntax highlight

[versed eagle]

lol

syntax highlight hard

[old socket]

Esp with nvim which was what I was doing it for

I had many problems with overlapping, etc

Shot myself in the foot with regions which was basically only did decorator highlighting

[fleet bridge]

>>> a = ast.parse('x = 1 + ...'); c = compile(a, '', 'exec', optimize=0); dis(c)
  1           0 LOAD_CONST               0 (1)
              2 LOAD_CONST               1 (Ellipsis)
              4 BINARY_ADD
              6 STORE_NAME               0 (x)
              8 LOAD_CONST               2 (None)
             10 RETURN_VALUE
>>> c = c.replace(co_consts=([1], [2], None))
>>> exec(c); x
[1, 2]

replacing .co_consts with arbitrary values works, it is very good

obviously, compile(string) cant put non-literals into .co_consts
compile(modded ast) raises TypeError: got an invalid type in Constant: list

[split violet]

import sys

class System:
    class terminal:
        class stdin:
            data = []
            
            def input():
                if len(System.terminal.stdin.data) == 0:
                    System.terminal.stdin.data += sys.stdin.readline().rstrip('\n').strip(" ").split()
                return System.terminal.stdin.data.pop(0)
                
            def inputln():
                if len(System.terminal.stdin.data) == 0:
                    ret_val = sys.stdin.readline().rstrip('\n')
                else:
                    ret_val = " ".join(System.terminal.stdin.data)
                System.terminal.stdin.data = []
                return ret_val
                
        class stdout:
            @staticmethod
            def print(data):
                sys.stdout.write(str(data))
            @staticmethod
            def println(data):
                sys.stdout.write(f'{str(data)}\n')

System.terminal.stdout.println(System.terminal.stdin.inputln())```

I need more

[old socket]

mfw when not esoteric

[mint mural]

mfw readable code (i did not expect this)

[burnt pasture]

is there such a thing as accidentally esoteric? In IRC, we once had someone show us a long file with this structure (many lines not shown here):

        while True:
            class MyClass:
                def Method(self):
                    for g in data:
                        if __name__ == '__main__':
                            do_something()

[earnest wing]

just needs a with and try to top it off!

[burnt pasture]

yeah, it was an eye-opener....

[versed eagle]

then put it all in a lambda

[quartz wave]

(in an exec in a lambda)

[versed eagle]

in an eval in a print

in every builtin function

[pearl berry]

Sounds useful, it should be buildin xD

[lavish dirge]

the output for just 1 line of code

[lime seal]

https://github.com/SinaKhalili/Folders.py

[sick hound]

oh god

[fleet bridge]

__inline_1__: Any
__inline_2__: Any
__inline_3__: Any

def _test1() -> None:
    def f(a: int, b: int) -> int:
        return a + b + 1
    g = inline_const(f, 1, 3)
    assert f(0, 0) == 1, f(0, 0)
    assert g(0, 0) == 3, g(0, 0)

def _test2() -> None:
    def f(a: int) -> int:
        return a + __inline_1__
    g1 = inline_var(f, '__inline_1__', 1)
    g2 = inline_var(f, '__inline_1__', 2)
    assert g1(1) == 2, g1(1)
    assert g2(1) == 3, g2(1)

def _test3() -> None:
    def f(a: int) -> int:
        return a + __inline_1__ + __inline_1__
    g1 = inline_var(f, '__inline_1__', 1)
    g2 = inline_var(f, '__inline_1__', 2)
    assert g1(1) == 3, g1(1)
    assert g2(1) == 5, g2(1)

def _test4() -> None:
    def f(a: int) -> int:
        return a + __inline_1__ + __inline_2__

    g1 = inline_var(inline_var(f, '__inline_1__', 1), '__inline_2__', 2)
    g2 = inline_var(inline_var(f, '__inline_1__', 3), '__inline_2__', 4)

    assert g1(1) == 4, g1(1)
    assert g2(1) == 8, g2(1)

i implemented some easy inline cases
inline_const just replaces one const with another
inline_var replases LOAD_GLOBAL var with LOAD_CONST const
it was easy...

[mint mural]

oh my god

[fleet bridge]

@quartz wave I think you will be interested ☝️

[lavish dirge]

@mint mural what

[mint mural]

cursed

[lavish dirge]

???

[quartz wave]

what does the 3rd argument do

[lavish dirge]

@mint mural is there something you didn't like

[mint mural]

i like it, the syntax is cool

[lavish dirge]

Oh! I thought you didn't like something about it.

[fleet bridge]

inline_const(func, before, after)
inline_var(func, varname, constvalue)

[lavish dirge]

@mint mural can you understand the logic

[mint mural]

yes

[lavish dirge]

That's the output for just a command for just 10 words

[mint mural]

dang haha

keep up the work, seems interesting and i like it

[sick hound]

ZZpwX1obfKJZQotcZ3FSDZZDWo5XZX09

bet u guys cant decrypt this

[earnest wing]

is this ROT

[sick hound]

[mint mural]

calculator = type("calculator", (object, ), {"__init__": lambda self: setattr(self, "operations", {"add": lambda x, y: x + y, "subtract": lambda x, y: x - y, "multiply": lambda x, y: x * y, "divide": lambda x, y: x / y}), "run": lambda self: print(self.operations[input("What do you want to do? ")](int(input("First number: ")), int(input("Second number: "))))})();calculator.run()

any opinions?

[arctic skiff]

useless

[mint mural]

simple calculator

[arctic skiff]

doesnt seems so

i cant do this simple calculation

i just told it to calculate the mass of sun

[mint mural]

seems like a you problem

[royal whale]

🤣

[last locust]

.wa calculate the mass of sun

[glass drum]

Wolfram Alpha

View original for a bigger picture.

[last locust]

Ez

[arctic skiff]

look lance is better than your calculator

exec("import sys;s=''\ndef o():\n\too=input('enter oprator [+ - / * ^] (leave blank for answer): ')\n\treturn oo\nwhile True:\n\ti=input('input number: ')\n\ttry:\n\t\tfloat(i)\n\t\ts+=i\n\texcept:\n\t\tprint('please enter a number')\n\twhile True:\n\t\tt=o()\n\t\tif t=='' or t.isspace():\n\t\t\tprint(eval(s))\n\t\t\tsys.exit()\n\t\telif t in ('+', '-', '*', '/', '^'):\n\t\t\tif t=='+':\n\t\t\t\ts+='+';break\n\t\t\telif t=='-':s+='-';break\n\t\t\telif t=='*':s+='*';break\n\t\t\telif t=='/':s+='/';break\n\t\t\telif t=='^':s+='**';break")
```here is one line better calc:)

@mint mural

@pallid flare

[arctic skiff]

(ik its just normal code packed in exec())

[pallid flare]

...

[mint mural]

this is nothing special

any idiot can do this

it's not better,

[arctic skiff]

Functionality is better than yours

Tho i will actually make it in 1 line

[pallid flare]

hmm

[mint mural]

my code is in 1 line

[sick hound]

lambda, my beloved

[mint mural]

for real

[vital dirge]

hey how much functionality of a program do you suppose we could stuff into just the decorator portion of the main class/function's declaration?

[versed eagle]

well you can decorate with lambdas, so all of it

you can do @(lambda...)

[quartz wave]

you can decorate with any expression actually

[versed eagle]

has to be a callable

doesnt it?

otherwise it will give a 'classname' object is not callable

ValueError iirc

[quartz wave]

yes but you can put any expression as a decorator

[versed eagle]

right, but it will only actually work if its callable

[quartz wave]

yes

[versed eagle]

and we're talking about it actually working

[quartz wave]

no

[versed eagle]

.

it would have to function

if it threw an error, it wouldnt function

[quartz wave]

i'm basically saying "all of it" but literally

you can put a tuple subscript as a decorator

or an and/or

[versed eagle]

yeah, it just has to be a callable

*the result of the expression

has to be a callable

that takes at least 1 argument

**unless you use some sort of fancy ctypes and/or ast stuff that changes how thingies work

[earnest wing]

not even that!

your whole program just needs to be finished and exit before the decorator gets called

if you don't like the exception that is

[versed eagle]

well yeah, but we're talking about the decorator being run
obviously if a call is never attempted it wont complain about it not being callable

[quartz wave]

print(globals().update(r={0:1,'a':1})or eval(''.join([i+([r.update(a=0)or g if g else''for _ in iter(lambda:r['a']and r[0],0)for t in[input('enter operator [+ - / * ^] (leave blank for answer): ')]for g in[r.__setitem__(0,0)and''if''==t or t.isspace()else t if t in'+-*/'else'**'if'^'==t else'']][-1],r.update(a=1))[0]for _ in iter(lambda:r[0],0)for i in[input('input number: ')]if r.pop('r',type('',(__import__('contextlib').ContextDecorator,),{'__enter__':int,'__exit__':lambda s,*a:isinstance(a[1],ValueError)and[r.update(r=print('please enter a number'))]})()(float)(i))])))

[finite blaze]

sexy

[night quarry]

Hey @sick hound!

It looks like you tried to attach a Python file - please use a code-pasting service such as https://paste.pythondiscord.com

[sick hound]

for some reason, pyright reports no typing error on this https://paste.pythondiscord.com/ipimowocid

and mypy only reports one error, which is the side effect of having subclasses of object not having the same constructor signature as the base class

[sick hound]

this sucks tbh, and the code highlighting is broken

there is not even an embed or something

[humble rune]

this is esoteric python not #help-dumpling

[next flame]

???

[sick hound]

• not asking for help
• this is esoteric as it is cursed

[sick hound]

"???" ?

[next flame]

code highlighting seems to be working fine

[sick hound]

on my computer it doesn’t

[next flame]

do you have js disabled

[sick hound]

why would I

by broken i didn’t mean that there was no syntax highlighting

I just meant that it doesn’t highlight properly

or for the least it is very basic highlighting

even discord does better

[next flame]

aside from builtins not being highlighted, I don't see a difference

[sick hound]

well that’s a major difference imo

anyway that doesn’t matter

[arctic skiff]

Nice

[arctic skiff]

you forgot semi colon:)

!e```py
import builtins
class foo:
def lshift(self, bar):
print(bar, end=''); return self
setattr(builtins, 'cout', foo())

####################
cout << "hello world";

[night quarry]

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

hello world

[fleet bridge]

Can you implement cin?

[maiden blaze]

#python-discussion message

[old socket]

!e ```py
@type.call
class cin:
def rshift(self, inp):
print(inp)

cin >> "Hello World"

[night quarry]

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

Hello World

[versed eagle]

cin is for input, not output

:(

[earnest wing]

me when I input into a string literal

[versed eagle]

i have done that quite a lot before lmao

[old socket]

oh shit lol woops

I have not done as much c++ as I have C

[versed eagle]

lol

its ok

[arctic skiff]

You also forgot ;

@type.__call__
class cin:
    def __rshift__(self, inp):
        return input(inp)

import builtins
class foo:
    def __lshift__(self, bar):
        print(bar, end=''); return self
setattr(builtins, 'cout', foo())

a=cin >> "input: " ;
cout << a;```

@fleet bridge @versed eagle

[fleet bridge]

doesnt work like in c++:

int a; string b;
cin >> a >> b;

[arctic skiff]

If you want this much syntax

You came to wrong person

[arctic skiff]

Tho cin >> a >> b; should work

But i cant make int a and string b

[versed eagle]

thats not how cin works

cin returns itself

and it splits on spaces

so if you do this

std::string str1;
std::string str2;
std::cin >> str1;
std::cin >> str2;

and on the first input the user types
two words
then str1 will be two and str2 will be words, because cin only reads until (and including) the next space

so what you have to do is create a buffer stored by cin, read into that, and then return from that until its empty
then read input again

[quartz wave]

std::cin << 😔

[versed eagle]

im tired af

give me a break

[versed eagle]

fixed

[arctic skiff]

Sorry not experienced with cpp

[arctic skiff]

No idea how i can do that

[fleet bridge]

How to get annotations of local vars?
Is it possible without parsing source code?

!e

def f()->None:
    x: int

print(f.__annotations__)

[night quarry]

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

{'return': None}

[restive void]

The annotations happen at runtime there, so they won't be on the f object at all

To be more precise: they're not evaluated at all

only at class/module level:

Annotating a local variable will cause the interpreter to treat it as a local, even if it was never assigned to. Annotations for local variables will not be evaluated.

[sacred umbra]

!e ```py
print("This is how you get the funny number in Python")
print((a:=((b:=([]==[]))+([]==[]))*(c:=(([]==[])+([]==[]))))aa+c+c+b)

[night quarry]

@sacred umbra :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | This is how you get the funny number in Python
002 | 69

[sacred umbra]

@full talon Thanks for the idea

[sacred umbra]

!e ```py
print((c:=(b:=(a:=((({},)is({},))==({'':''}is{'':''})))), [(b:=(a:=((({},)is({},))==({'':''}is{'':''}))+a))for _ in range(68)])[c][b-c-c])

[night quarry]

@sacred umbra :white_check_mark: Your 3.11 eval job has completed with return code 0.

69

[sacred umbra]

!e ```py
print((c:=(b:=(a:=((({},)is({},))==({'':''}is{'':''})))), [(b:=(a:=((({},)is({},))==({'':''}is{'':''}))+a))for _ in range(c+c*(c+c)(c+c)(c+c)(c+c)(c+c)*(c+c)+c+c+c)])[c][b-c-c])

[night quarry]

@sacred umbra :white_check_mark: Your 3.11 eval job has completed with return code 0.

69

[full talon]

np 😄

now extend it

and make a very HOOGE one

[fleet bridge]

Yes, i know that they are not evaluated, it is sad.
Isnt it better to put them in some place? (Before first func call, so they must be strings (or some other lazy evaluating object))
Like f.__local_annotations__ or f.__annotations__ = {'__locals__': {local vars annotations}, **{signature annotations}}

[arctic skiff]

from varname import argname
@type.__call__
class cout:
    def __lshift__(self,inp):
        print(inp, end='')
@type.__call__
class cin:
    def __rshift__(self,inp):
        i=argname('inp');globals()[i]=input()

##################################################


a=''
cin >> a;
cout << a;```

@versed eagle

@fleet bridge

heres cin

[fleet bridge]

This doesn't work if you call it from different module
You should modify parent frame globals, not your globals

Also:
cin >> cin;
cin >> x; # boom

Can i do a<<cin or a>>cout?

[earnest wing]

cin >> ㅤ.x (hangul filler character)

[sacred umbra]

okay sure @lunar ore

brain damage

Still no Pyfuck though

[full talon]

Guess what ill do: Convert JSFuck to PyFuck

[sacred umbra]

Well good luck to get it to work with any arbitrary code

Instead of just numbers

[sacred umbra]

!e ```py
print(f'{u((a:=((()==())+(()==())))aaaa+(aa*a))}{u(aaaaa*a-(aaa)-(aaa+a//a)-a)}{u(b:=(aaaaaa-(aaa)-(aa//a+a)))}{u(b)}{u(c:=(b+a+(a//a)))}{u(aaa*a)}{u(b-(aaaa)+aa+(aaa-(a//a)))}{u(c)}{u(b+(a**aa-a))}{u(b)}{u(b-(a**a*a))}')

[night quarry]

@sacred umbra :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 1, in <module>
003 | NameError: name 'u' is not defined

[sacred umbra]

wait I think I did something wrong

!e ```py
(u:=chr),
print(f'{u((a:=((()==())+(()==())))aaaa+(aa*a))}{u(aaaaa*a-(aaa)-(aaa+a//a)-a)}{u(b:=(aaaaaa-(aaa)-(aa//a+a)))}{u(b)}{u(c:=(b+a+(a//a)))}{u(aaa*a)}{u(b-(aaaa)+aa+(aaa-(a//a)))}{u(c)}{u(b+(a**aa-a))}{u(b)}{u(b-(a**a*a))}')

[night quarry]

@sacred umbra :white_check_mark: Your 3.11 eval job has completed with return code 0.

Hello World

[quartz wave]

is not the underscore challenge but impressive anyway so approved

[full talon]

how did u generate this

[sacred umbra]

A mix of True + True = 1 and the chr function

[full talon]

ok

[sacred umbra]

underscore challenge?

[sacred umbra]

I feel like with eval you could make it execute any code

Well there's your JSFuck

Where your only primitives are +, -, *, **, //, (, ), :=, chr and eval

It has to be floor division since chr can't accept floats

[full talon]

dang

so +-*/:= + chreval

[sacred umbra]

Well sure

It's not as minimalistic as JSFuck

But it's close enough I guess

[full talon]

altho

[versed eagle]

a while ago i wrote a proper std::cout with a buffer and whatnot

[full talon]

we could make u chr

[versed eagle]

basically, only vars that you create have to be underscores, anything you use thats builtin has to be a dunder, only tuples
its pretty cool

heres an example of it

[sacred umbra]

I just made chr u since it's short for Unicode

[versed eagle]

rules, part 1

rules part 2

[sacred umbra]

@solar dune Hey what if someone made a transpiler that transpiles Python to this subset of Python

Python to Python's own JSFuck

And what if I forced you to take witness in the transpiled code

u=chr
eval("spaghetti output here")

[solar dune]

do it

if it produces a single newline it is wrong

[sacred umbra]

So eval can't have newlines 👍

[old socket]

lisp it

[sick hound]

that would be crazy

[astral rover]

yes

classes are harder but functions arent too bad

[fleet bridge]

!e

x = __name__.__class__.__class__(__name__, __name__.__class__.__bases__.__class__(), __annotations__.__class__())
print(x)

[night quarry]

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

<class '__main__.__main__'>

[astral rover]

i was more thinking use build_class but that works too ig

[fleet bridge]

__loader__.__dict__['exec_module'].__func__.__class__(__loader__.__dict__['exec_module'].__func__.__code__.__class__(a lot of args), __annotations__.__class__())

how to get type(lambda:...) without lambdas, defs and getting pre-existing functions?

also i dont like object.__subclasses__()[...]

[earnest wing]

"Preexisting functions" covers quite a lot

You won't likely find a non-C function inside C builtins (outside of something like globals snooping)

[sacred umbra]

https://pastebin.com/XCpjrcHr

[night quarry]

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

<class '__main__.__main__'>

[sacred umbra]

Oh, Pastebin removed it

https://gist.githubusercontent.com/DenseReptile/40c0dbfddb2e5fdd8dfbd595f420c4d1/raw/ef413257fa25f8d78e7087589e1e273b1c639f4c/rip.py

- Copy the source code.
- Open the REPL.
- Paste into the REPL then enter.
- print(_("print('Hello World')"))
- Copy printed output
- Paste into REPL then press enter (again).
- :lemon_surprised:

@full talon

Okay if the REPL can't handle it you may need to paste into a script file instead

And yes, I've obfuscated/transpiled the source code into Pyfuck as a demo within itself

I still have the original saved though

Honestly this is really more of an obfuscator still than a true Pyfuck

Since it still depends on exec and chr

[full talon]

@sacred umbra what does the code do

it does nothing atm

also did you make pyfuck

[sacred umbra]

Question, did you paste it in a REPL or script file?

[full talon]

script file

it has an error halfway through the file

[sacred umbra]

VS Code can't parse it right

Too much for it I guess

[sacred umbra]

# The thing
print(_("print('Hello World')"))

[full talon]

@sacred umbra have you made an automated script for this

[sacred umbra]

That is the automated script

[full talon]

I mean to convert it into pyfuck code

[sacred umbra]

I ran the source code of the automated script into itself

[full talon]

so do u have smth that can make MY code look like that

[sacred umbra]

@full talon
Gonna ask you something while I type it out

So did you add the line in this reply below the source code of the obfuscated obfuscator/Pyfuck

[full talon]

yea

and it just printed more gibberish

[sacred umbra]

Now copy, paste then run that

[full talon]

whats the _() part

[sacred umbra]

The obfuscator/Pyfuck as a function

Hold on I'm just gonna copy and paste it over from Google Notes to my phone then paste the unobfuscated source code here

The unobfuscated source code of this glorified shitpost

def _(source):
    left,right="{","}"
    ucp_store=[]
    for char in source:
        ucp_store.append(ord(char))
    def _():
        str_store=[]
        for ucp in ucp_store:
            store=[]
            for _ in range(ucp):
                store.append("(()==())")
            resulting_string=f"{left}u({'+'.join(store)}){right}"
            str_store.append(resulting_string)
        return "".join(str_store)
    print(f"""
u=chr
exec(f\"{_()}\")
""".strip())

[sacred umbra]

Btw if you've copied, pasted and run the gibberish does it say Hello World

[full talon]

@sacred umbra I have made it a compiler

like an actual one

[full talon]

have you ever thought of MAYBE using multiplication

[sacred umbra]

No since this is just a quick and simple shitpost

Where I just make it look like JSFuck or Brainfuck

[full talon]

imma just, fix ir

as it ran out of memory

@sacred umbra my python is using 1GB for a simple hello world

holy frk

[night quarry]

Hey @full talon!

It looks like you tried to attach a Python file - please use a code-pasting service such as https://paste.pythondiscord.com

[full talon]

@sacred umbra it uses VERY big amounts of memory XD

and I need to multiply

for good perf

[sacred umbra]

I have 16 GB to spare so yeah

[full talon]

dang 300mgb for lorem

[sacred umbra]

I originally considered writing this in JavaScript

To make things worse

[burnt pasture]

suddenly everyone is using #esoteric-nicks making this channel even harder to read.

[vague cairn]

Maybe we've already done this one before, but:

Ok, so the lambda syntax evaluates to a function that can then be returned or assigned or whatever else I might wish to do with an expression with a function in it... For simplicity I will exemplify with:
x=lambda:...
which does the same thing as:
def x():...
which takes 1 fewer characters. it is also a syntax for creating a function, and for binding it to a name.

However this is esoteric-python, it seems to me that outside of line golf or character golf it should be vaguely as frequent for us to be abusing the second case for assignments as much as we abuse the first case for function definitions.

What is the shortest / most complex / most obscure way to use the def syntax for arbitrary assignments instead of function definition?

my starting entry is: ```py
def _(f): return f.code.co_consts[1]

@_
def greet(): return 'Hello World!'

print(greet)

[restive void]

@lambda f:f.__doc__
def name():"Grass"

[full talon]

@restive void

@lambda f:f()
def test():
    def test2():
        return "Hello World!"
    return test2

now test is test2

[vague cairn]

Impressive! not general purpose but works for strings.

[restive void]

@lambda f:eval(f.__doc__)
def foo():
    "23"

[vague cairn]

lol!

evil— I mean eval

[old socket]

!e ```py
@lambda f:f.defaults[0]
def foo(_=23): ...

print(foo)

[night quarry]

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

23

[fleet bridge]

i hate changing bytecode directly

here you can see minimap of several functions: inline_const, inline_var, inline_method and inline_attr
i didnt even start inlining bytecode, im just replacing some LOAD_ATTR, LOAD_GLOBAL, LOAD_METHOD... with faster LOAD_CONST

using these optimizations i got 50% speedboost of my code, which has a lot of small functions

if you change the order of decorator - it breaks

@inline_attr('_s', 'size', Struct('?').size)
@inline_attr('_s', 'unpack', Struct('?').unpack)
@inline_method('self', 'read', read)
def read_bool(self, /) -> bool:
    return _s.unpack(self.read(_s.size))[0]



Original code:
376           0 LOAD_GLOBAL              0 (_s)
              2 LOAD_METHOD              1 (unpack)
              4 LOAD_FAST                0 (self)
              6 LOAD_METHOD              2 (read)
              8 LOAD_GLOBAL              0 (_s)
             10 LOAD_ATTR                3 (size)
             12 CALL_METHOD              1
             14 CALL_METHOD              1
             16 LOAD_CONST               1 (0)
             18 BINARY_SUBSCR
             20 RETURN_VALUE

Optimized code:
376           0 NOP
              2 LOAD_CONST               3 (<built-in method unpack of _struct.Struct object at 0x00000161ACDBD260>)
              4 LOAD_CONST               2 (<function read at 0x00000161AD1FA290>)
              6 LOAD_FAST                0 (self)
              8 NOP
             10 LOAD_CONST               4 (1)
             12 CALL_FUNCTION            2
             14 CALL_FUNCTION            1
             16 LOAD_CONST               1 (0)
             18 BINARY_SUBSCR
             20 RETURN_VALUE

it gives you exactly 0% speedup (20%-50% actually, if you apply a lot of changes to different functions)

[serene stratus]

would be nice if you remove the NOPs

[fleet bridge]

it is harder than just replacing two opcodes with NOP+opcode

i should fix all jumps, maybe fix EXTENDED_ARG related to jumps
i think, in 3.11 i should also fix exception table

[serene stratus]

yes you'd have to but it's not that hard, I can give you my bytecode editor package if you want which handles all that

[serene stratus]

very interesting concept though, might make for some good obfuscation technique to disable decompilers

[quartz wave]

why don't you just remove the places of the NOPs

[quartz wave]

do i recreate this in 3.11 or 3.12

[fleet bridge]

i created it for 3.10, it will break in 3.11 because of a lot bytecode incompatible changes in 3.11

[night quarry]

Hey @fleet bridge!

It looks like you tried to attach a Python file - please use a code-pasting service such as https://paste.pythondiscord.com

[fleet bridge]

https://pastebin.com/HRcdpnCz
https://pastebin.com/vnk4XqRe
@quartz wave there is my code (it is VERY (very very) buggy, i would recommend not use it but rewrite it entirely)

[quartz wave]

ok so i rewrote the code for 3.11
still haven't implemented inline_attr or inline_method but here are some results ```py

from dis import dis
s = 5
@inline_globals(check_store=2)
... @inline_var('a', 5, inline_args=False)
... def g(a, b):
... global s
... print(s)
... s += 7
... print(s)
... return a + b
...
dis(g)
1 0 RESUME 0

5 2 LOAD_GLOBAL 1 (NULL + print)
14 LOAD_CONST 2 (5)
16 PRECALL 1
20 CALL 1
30 POP_TOP
32 LOAD_GLOBAL 1 (NULL + print)
44 LOAD_CONST 3 (12)
46 PRECALL 1
50 CALL 1
60 POP_TOP

7 62 LOAD_FAST 0 (a)
64 LOAD_FAST 1 (b)
66 BINARY_OP 0 (+)
70 RETURN_VALUE

s += 7 is completely factored out this shouldn't be happening so s can actually be assigned to
but the 12 in the second print() call is expected behaviour

also a isn't replaced because inline_args=False as you can see

[fleet bridge]

What is check_store for?

[quartz wave]

to stop inlining a name when there's a STORE_* with that name

if it's 2 then it means to consider constant assignments before stopping inlining

so basically it turns ```py
s = 5

@inline_globals(check_store=2)
@inline_var('c', 5, check_store=2, inline_args=False)
def g(a, b):
global s
print(s)
s += 7
print(s)
c += 2
res = a + b + c
c += a
res2 = a + b + c
return res, res2
intopy
def g(a, b):
global s
print(5)
s = 12
print(12)
res = a + b + 7
c += a # c = a + 7
res2 = a + b + c
return res, res2

[fleet bridge]

Interesting, but i dont see any usecases for that feature

Also, it is good if your decorated functions still works if you remove @inline_* decorators.
Because these decorators are very unstable. They will be broken in every new python version

What library are you using to manipulate bytecode? dis? bytecode? Something else?

[fleet bridge]

What inline_args is doing? It removes argument from function signature, it inline_args=True

[quartz wave]

https://github.com/thatbirdguythatuknownot/sniplections/blob/main/inline.py

[quartz wave]

inline_args=True basically means override the parameter name with the inline constant

[fleet bridge]

@inline_var('c', 5, check_store=2, inline_args=False)
def g(a, b):

i think here inline_args=False is useless because c is not a parameter

[quartz wave]

oh yeah

i forgot to edit that out from the time i used 'a'

[fleet bridge]

_replace_code(code, 0, 0)
💥 on 59 line

[quartz wave]

hmm

[quartz wave]

fixed

[fleet bridge]

fastest bugfix ever

[quartz wave]

5s bugfix

[fleet bridge]

you removed most useful decorator for debugging(

[quartz wave]

ok i'll add that

[quartz wave]

done

[fleet bridge]

fastest feature implementation ever

def cmp(f: F) -> F:
    print(f'Comparing {f.__qualname__}...')
    dis.dis(f)
    print()
    print()
    while hasattr(f,'__orig_func__'):
        f = f.__orig_func__  # type: ignore[attr-defined]
        dis.dis(f)
        print()
        print()

    return f

builtins.cmp = cmp  # type: ignore[attr-defined]

@cmp
@inline_var('x', 5)
@inline_const(None, 123)
def f() -> None:
    print(x)

445           0 LOAD_GLOBAL              0 (print)
              2 LOAD_CONST               1 (5)
              4 CALL_FUNCTION            1
              6 POP_TOP
              8 LOAD_CONST               0 (123)
             10 RETURN_VALUE


445           0 LOAD_GLOBAL              0 (print)
              2 LOAD_GLOBAL              1 (x)
              4 CALL_FUNCTION            1
              6 POP_TOP
              8 LOAD_CONST               0 (123)
             10 RETURN_VALUE


445           0 LOAD_GLOBAL              0 (print)
              2 LOAD_GLOBAL              1 (x)
              4 CALL_FUNCTION            1
              6 POP_TOP
              8 LOAD_CONST               0 (None)
             10 RETURN_VALUE

this is breaking functions qualnames
add result.__qualname__ = f.__qualname__ to fix

[quartz wave]

k wait

done

i gtg now

[fleet bridge]

is inline_args=True removing arg from signature?

[sullen kayak]

hii esoteric gangsters

import random;import os;print(''.join([['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0'][random.randint(0,len(['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0'])-1)]for i in range(len(['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0']))]));os.system("pause")

any way to make this harder to read?

[royal whale]

Oh, yes...
always

rdmoan=__import__("".join([(a:=["r", "d", "m", "o", "a", "n",][:])[0], a[-1-1], a[-3+2], a[1], a[-(6//2)], a[2]]); exec("sys tropmi; subprocess sa ssecorpbus tropmi"[::-1]); sys.stdout.write(''.join([['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0'][rdmoan.randint(0,len(['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0'])-1)]for i in range(len(['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0']))]));ssecorpbus.call("pause",shell=-(-(-(False))))

for example

That's a comparatively small change
You could do much more

@sullen kayak

[sullen kayak]

what the haill

couldn't you use the ascii library

to get rid of the alphabet and whatever

what the fuck are oyu just setting the

oh my god

[royal whale]

...you asked...

so I answered

[sullen kayak]

no i'm just

SO confused

LMAO

is there any way to obfuscate the letters + numbers

or do they help with the obfuscation

[floral meteor]

small change: ```py
rdmoan=import("".join([(a:=["r", "d", "m", "o", "a", "n",][:])[0], a[-1-1], a[-3+2], a[1], a[-(6//2)], a[2]]); exec("sus sa sys tropmi; subprocess sa ssecorpbus tropmi"[::-1]); sus.stdout.write(''.join([['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0'][rdmoan.randint(0,len(['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0'])-1)]for i in range(len(['a','b','c','d','e','f','g','h','i','j','k','l','m','n','o','p','q','r','s','t','u','v','w','x','y','z','1','2','3','4','5','6','7','8','9','0']))]));ssecorpbus.call("pause",shell=-(-(-(False))))

[sullen kayak]

it's you again!!

what projects are you workin on

esoteric python florida man

[floral meteor]

I made 4 dimensional snek

with a classic 20th century gui for the start menu

hard mode is actually hard, easy mode is so easy you can actually play it without instinctual 4 dimensional comprehension

I think I might make a pause menu

instead of esc to quit

[royal whale]

Hmm
Enlightening

[floral meteor]

I almost confused myself then I saw the lambda wxyz=[8,8,8,8];k=lambda:4-[*wxyz].count(1);

n=S.N;(w,x,y,z),s,d=S.h,n.n,n.h;t=w+(d==2)-(d==3),x+(d==1)-(d==0),y+(d==6)-(d==7),z+(d==5)-(d==4) this is how you do unreadability

[quartz wave]

no

it's just allowing the inlining of the variable when it's an argument

i think it's actually better if i'd named it inline_arg

[fleet bridge]

What happens if you try @inline_var('arg', 0, inline_args=False)
?

[quartz wave]

if it works as expected it should just return the original function if arg is a parameter of the function

[sullen kayak]

does anybody have any idea how i resolve that?

this is my whole program

import random; import sys; import webbrowser; for i in range(10): webbrowser.open(f"https://www.google.com/search?q={''.join([[i for i in sys.argv[1]][random.randint(0,len([i for i in sys.argv[1]])-1)] for i in range(len([i for i in sys.argv[1]]))])}&tbm=isch",new=2)

i would like to keep it all on one and the same line

[restive void]

not possible like that. You can just turn it into a list comprehension

import webbrowser; [webbrowser.open(f"https://{i}.example.com") for i in range(10)]

[sullen kayak]

oh

thank you

[onyx jacinth]

this regex checks for prime number 😮
/^1?$|^(11+?)\1+$/

i was like what ? how?

but it does work

[earnest wing]

the magic is in the backreferences

[onyx jacinth]

pumping lemma is too OP

haha

[earnest wing]

😄

[serene stratus]

https://github.com/Svenskithesource/pySpy

might be something you guys like

[buoyant tulip]

is there something i can hook into to do some action when a class object is initialized? i.e. i want to be able to write class Foo(MyMetaHook): ..., and MyMetaHook gives me a way to take some action that acts on the Foo class (not an instance of it)

context: i want to add little unit tests to jupyter notebook cells; i'd like to be able to write class TestFooCell(CellTestFixture): ... and have it automatically do the equivalent of unittest.main()

aha, looks like __init_subclass__ does the trick

[lavish dirge]

https://github.com/Zconj095/EdgeLoreScripts_Model_001

My progress so far on developing the language

[fleet bridge]

use metaclass
metaclass can change class creation process

[craggy hamlet]

Here's my first attempt at making some esoteric code and being brave enough to post it (python beginner 😅)

Queue = type("Queue", (object,), {'__init__' : lambda _: _.__setattr__("arr", []),'isEmpty' : lambda _ : (not True) == len(_.__getattribute__("arr")),'enqueue' : lambda _, __ : _.__getattribute__("arr").__getattribute__("insert")((not True), __),'dequeue' : lambda _ : _.__getattribute__("arr").__getattribute__("pop")(),'size' : lambda _ : len(_.__getattribute__("arr"))
})
q = Queue()

q.enqueue(4)
print(q.dequeue())

[quartz wave]

approved

[royal whale]

!e

Queue = type("Queue", (object,), {'__init__' : lambda _: _.__setattr__("arr", []),'isEmpty' : lambda _ : (not True) == len(_.__getattribute__("arr")),'enqueue' : lambda _, __ : _.__getattribute__("arr").__getattribute__("insert")((not True), __),'dequeue' : lambda _ : _.__getattribute__("arr").__getattribute__("pop")(),'size' : lambda _ : len(_.__getattribute__("arr"))
})
q = Queue()

q.enqueue(4)
print(q.dequeue())

[night quarry]

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

4

[royal whale]

Nice! It's esoteric enough that I can't tell what it does by glancing at it
10/10!

[craggy hamlet]

❤️ thanks for all the kind responses!

[finite blaze]

with open("przyklad.txt", "r") as f:
  for x in f:
    if int(x[::-1]) % 17 == 0: print(x)

any shorter way to write it?

i guess i could open it inside for but then it wouldn't close?

[versed eagle]

[print(x)for x in open("whatever.txt")if int(x[::-1])%17==0]```

r is the default open mode, no need to call open with it

if you really want the try except in the with, you can do this

with open("whatever.txt")as _:[print(x)for x in _ if int(x[::-1])%17==0]

[quartz wave]

with open("whatever.txt")as _:[int(x[::-1])%17<1==print(x)for x in _]
# and if there are negative numbers
with open("whatever.txt")as _:[int(x[::-1])%17==0==print(x)for x in _]

[versed eagle]

I always forget about comparison chaining lmao

nice

[quartz wave]

for this you can do ```py
[int(x[::-1])%17<1==print(x)for x in open("whatever.txt")]

or

[int(x[::-1])%17==0==print(x)for x in open("whatever.txt")]

[night quarry]

Hey @lavish dirge!

You either uploaded a .txt file or entered a message that was too long. Please use our paste bin instead.

[lavish dirge]

https://paste.pythondiscord.com/raw/yacupewufu

ACCURACY = THE POINT AND RANGE OF A MEASURED AMOUNT OF CAPABILITY A POSSIBILITY CAN HAPPEN AND DETERMINE COME INTO EFFECT
RADIUS = SET RANGE OF A CENTERED POINT TO THE END DESTINATION
DIAMETER = SET RANGE POINT FROM START TO MIDDLE TO THE END WHILE PASSING THE RADIUS
CONVERT = CHANGE FORM AND OR CHARACTER AND OR FUNCTION
PATTERN = REPEATING METHOD
WRITE = ENSCRIBE FROM LOOKING AT WORDS
READ = DESCRIBE FROM LOOKING AT A PATH OF WORDS
SPAN = MEASURED LIMITED RANGE
VIBRATION = PARTS THAT MOVE BACK AND FORTH AT A GIVEN SPEED
TRANSFORM = MAKE A CHANGE IN FORM
SYNCHRONIZE = LINK AND SEND THE SAME RESULT TO ALL SOURCES
SCAN = ANALYZE A SPECIFIC WORD OR FIELD AND OR GIVE DATA ON THE ASKED INFORMATION TO SEARCH FOR
ANALYZE = READ AND LOOK OVER
CALCULATE = GIVE A DESIGNATED OF A CALCULATES DESCRIPTION FOR A NUMBER AND GIVE ANSWER FOR ALL OF VALUE
LIMIT = SET DEFINED AMOUNT FOR KNOWLEDGE WITH A GIVEN POWER LEVEL
RECALL = GAIN THE ABILITY TO VIEW PAST MEMORY INSIDE BRAIN
REACH = GRAB TO PULL INWARDS
PREDICT = GIVE PERFECT VALUE
REPEAT = CYCLE SAME EFFECT AGAIN INTO SAME FREQUENCY
RECOGNIZE = RECALL FROM AN EARLIER POINT WITHIN TIME
ENCODE = COMPRESS CODE
DECODE = DECOMPRESS CODE
RECODE = COMPRESS CODE ONCE MORE

LOOP = BIND IN A CYCLE
MEASURE = TAKE IN THE AMOUNT AND DISTANCE OF
ANSWER = SOLUTION TO A PROBLEM
SOLUTION = FINAL OUTCOME TO AN FORMULA
PROBLEM = UNFINISHED SOLUTION
SEARCH = FIND AND LOCATE SOMETHING
ASK = STATE A QUESTION
TIME = MEASUREMENT IN WHICH CURRENT REALITIES MUST PASS
SPACE = CONTAINER IN WHICH TIME MUST PASS THROUGH
UPLOAD = TRANSFER3 INTO DESCRIBED LOCATION
DOWNLOAD = TRANSFER3 TO CURRENT DEVICE
SIDELOAD = TRANSFER3 TO ALL DEVICES WITH STATUS OF STATED SET LOCATION
CLONE2 = MAKE AN IDENTICAL COPY OF
SYNCHRONIZE = LINK AND SEND THE SAME RESULT TO ALL SOURCES
ENCODE = COMPRESS CODE
DECODE = DECOMPRESS CODE
RECODE = COMPRESS CODE ONCE MORE
SETTING = A MEASUREMENT COMMAND THAT CAN BE ADJUSTED AND BY AN OPERATOR
ADJUST = EDIT AND MODIFY
EDIT = CHANGE AND OR MODIFY TO ADJUST TO A SPECIFIED PURPOSE
WORK = PRODUCING EFFORT TO FINISH A TASK
WORKLOAD = THE AMOUNT OF WORK
COMMAND = ORDER TO BE GIVEN
LINK = BRING TOGETHER AND ATTACH TO
BIND = EDIT AND MODIFY
LEVEL = NUMBER AMOUNT OF OR SIZE
UNIT = STORAGE CONTAINER
DIMENSION = NUMBER OF GIVEN AXIS POINTS
NUMBER = ARITHMETICAL VALUE THAT IS EXPRESSED BY A WORD AND OR SYMBOLE AND OR FIGURE REPRESENTING A PARTICULAR QUANTITY AND USED IN COUNTING AND MAKING CALCULATIONS AND OR FOR SHOWING ORDER IN A SERIES OR FOR IDENTIFICATION
FREQUENCY = REPEATED PATTERN AND OR SETTING

POWER = AMOUNT
STRENGTH = LEVEL INTENSITY
CALIBRATE = SCALE WITH A STANDARD SET OF READINGS THAT CORRELATES THE READINGS WITH THOSE OF A STANDARD IN ORDER TO CHECK THE INSTRUMENT AND ITS ACCURACY
PUBLIC = ACCESS TO ALL OF CREATORS INTERIOR DOMINION
PRIVATE = HIDDEN TO EVERYONE BUT CURRENT2 USER2
PERSONAL = EXCLUSIVE TO THE CREATOR
ESCAPE = RETURN TO SOURCE PLACE2
RETURN = GO BACK
CONSTANT = ALWAYS IN EFFECT
CYCLE = PROCESS OF REPEATING AN EVENT CONTINUOUSLY IN THE SAME ORDER
MEASUREMENT = AN ACT TO CALCULATE AND GIVE A SPECIFIC LENGTH ON SOMETHING
CALCULATOR = A DEVICE USED TO CALCULATE INFORMATION AND ANALYZE SET TASKS AS A ROOT VALUE OF LOGIC
WAVELENGTH = A SET OF WAVE PATTERNS GIVEN FREQUENCY FORMAT IN A LENGTH OF A WAVE VALUE DETERMINED BY A PREVIOUS VALUE EFFECT
LENGTH = HOW LONG A MEASURED DIMENSIONAL OBJECT IS EXTENDED
LATTICE = INTERLACED STRUCTURE AND OR PATTERN
LOCATION = SPECIFIED AREA
LINE2 = CHOSEN DIRECTION THAT IS SET IN A SINGLE PATH
WAVE2 = CONTINUAL FLUCTUATION OF FREQUENCY AND OR PATTERN
WIDTH = MEASUREMENT OF SOMETHING FROM SIDE TO SIDE
HEIGHT = THE LENGTH OF RAISING OR LOWERING IN A VERTICAL PATH
HERTZ = DEFINED SOUND WAVE FREQUENCY
MEASURE = TAKE IN THE AMOUNT AND DISTANCE OF

Those are the mathematical variables the language has
just some and not a complete list yet

[burnt pasture]

@lavish dirge the pastebin was plenty, thanks. what does this have to do with esoteric python? Are you looking for help/feedback of some kind?

[lavish dirge]

Input and feedback. What others think of it. And how others see it could be used if I made it as a smaller library/module for it to connect to the full language with

What others see within its potential as well

[burnt pasture]

you should find another channel to discuss it. When you get there, be ready to explain how these definition (in all caps, why?) will be used as a "language"?

[next flame]

why spam

[earnest wing]

Oh you've made some progress in the language

Ididn't realize how many keywords there were

[lavish dirge]

@earnest wing you like it

[fleet bridge]

Is it a fortran code?

[lavish dirge]

@fleet bridge that is python believe it or not

[worldly trail]

This is my attempt at esoteric: https://paste.pythondiscord.com/raw/mihehusefu

[maiden river]

!e

a,b,c = 4,3,6

g = (a if a>c else c)if a>b else(b if b>c else c)

print(g)

[night quarry]

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

6

[maiden river]

single liner for greatest of three variables

Can we make one for 4 variables ?

!e

a,b,c,d = 4,3,6,8

z=(
((a if a>d else d)
if a>c else c)
if a>b else
((b if b>d else d)
if b>c else c)
)

print( z)

[night quarry]

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

6

[maiden river]

!e Working now : )

a,b,c,d=3,2,5,8

z=(

((a if a>d else d)
if a>c else (c if c>d else d))

if a>b else

((b if b>d else d)
if b>c else (c if c>d else d))
)

print(z)

[night quarry]

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

8

[restive void]

max(3,2,5,8)

[craggy hamlet]

!e

def _(w, x, y, z):
    a = [0] * 2
    a[0] = w
    a[1] = x 
      

    b = bool(a[0] - a[1] + abs(a[0] - a[1]))
    if b: 
        b = 1
    else: 
        b = 0
  

    a[0], a[1 - b] = a[1 - b], a[0]
      
 
    a[1] = y 
    b = bool(a[0] - a[1] + abs(a[0] - a[1])) 
    if b: 
        b = 1
    else: 
        b = 0
    a[0], a[1 - b] = a[1 - b], a[0]
    a[1] = z 
    b = bool(a[0] - a[1] + abs(a[0] - a[1]))
    if b: 
        b = 1
    else: 
        b = 0
    a[0], a[1 - b] = a[1 - b], a[0]
      
    return a[0]  
print(_(3,2,5,8))

[night quarry]

@craggy hamlet :white_check_mark: Your 3.11 eval job has completed with return code 0.

8

[craggy hamlet]

how about without using < > or max?

[restive void]

min(map(__import__("operator").neg, numbers)) :P

[craggy hamlet]

I'll generalize to not cheating :_)

[restive void]

you do realize what channel we're in? :P

But alright, I think this wouldn't count as cheating:

def max(numbers):
    orig = numbers.copy()
    while True:
        indices = set()
        for i in range(len(numbers)):
            if numbers[i]:
                numbers[i] -= 1
                indices.add(i)
        if len(indices) == 1:
            return orig[indices.pop()]

[maiden river]

wah ? 😳

[craggy hamlet]

This is maximum of 4 numbers.

[maiden river]

yes ik, but the algo

[quartz wave]

max_=lambda n:[v.update(a=2)or o[d.pop()]if len(d:={n.__setitem__(i,n[i]-1)or i for i in range(len(n))if n[i]})==1 else""for o in[n[:]]for v in[{}]for _ in iter(lambda:not v,0)][-1]
``` note that it won't work for negative values

[last locust]

You can even remove the space here: ==1 else :)

It breaks syntax highlighting but will still work

[quartz wave]

will produce a warning in 3.11

probably an error in 3.12

[last locust]

!e ```py
def foo(y):
return 1if y==5else 0

print(foo(5))
print(foo(3))```

[night quarry]

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

001 | <string>:2: SyntaxWarning: invalid decimal literal
002 | <string>:2: SyntaxWarning: invalid decimal literal
003 | 1
004 | 0

[last locust]

Huh, so it does

They're trying to stop us smh

[quartz wave]

it is an error in 3.12 nvm just tested

[last locust]

I like this feature though smh

It's a cool niche

[finite blaze]

How does it work?

==print()?

[quartz wave]

comparison chaining

[next flame]

sorted

[finite blaze]

But why == executes print?

[quartz wave]

hidden and

x<y==z
``` is equivalent to ```py
x<y and y==z

[finite blaze]

So if z is a print it is executed

I didn't know that python works like this

Cool thing, thanks!

Nvm I'm stupid lol. This is such an obvious thing

Other languages do the same thing lol

[quartz wave]

a majority of the other languages don't

they do (x < y) == z

[finite blaze]

I mean executing z

If it is a function

For a second i thought that it is some black magic

Btw, anyone tried golfing sieve of erathosthenes?

[serene stratus]

I think other languages would rather raise an error but I'm not sure

[fleet bridge]

i love this part: ==print(x)

[outer ledge]

ain't no way this works....

i need bleach

this channel is a war crime

forbidden knowledge

[finite blaze]

def SieveOfEratosthenes(n):
 r,m=range,n+1
 x,p=[1]*m,2
 while p*p<=n:
  if x[p]==1:
   for i in r(p*p,m,p):x[i]=0;p+=1
  [x[p]==1==print(p)for p in r(2,m)] 
SieveOfEratosthenes(20)

my attempt at golfing Sieve of Eratosthenes

[craggy hamlet]

!e

import re
z = 40
n=1
t=1
while n<z:
    t+=1
    if not re.match(r'^1?$|^(11+?)\1+$',"1"*t):
        print(t)
        n+=1

[night quarry]

@craggy hamlet :white_check_mark: Your 3.11 eval job has completed with return code 0.

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

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

[craggy hamlet]

what about we implement this algorithm for primes

[versed eagle]

i dont understand math please explain

math is scary

well actually
i sort of understand what its doing
just not how its doing it

Pn is the nth prime, then it does some fancy equation to get that prime

[craggy hamlet]

It's a really smart formula

Let me find the video on it that I watched

https://youtu.be/j5s0h42GfvM

It does some weird thing where it doesn't actually make primes so to speak, it eliminates things that AREN'T primes.

[fleet bridge]

it is at least O(2^n) time complexity
we can calculate at most 100 primes using formula

[fleet bridge]

!e lazy approach

import math as m

p=lambda n:1+sum(int((n/sum(map(int,map(2.0.__rpow__,map(m.cos,map(m.pi.__mul__,[(m.factorial(j)+1.0)/j for j in range(1,i+1)]))))))**(1/n))for i in range(1,2**n+1))

for i in range(1, 20):
    print(f'p_{i:<2} = {p(i)}')

[night quarry]

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

001 | p_1  = 3
002 | p_2  = 5
003 | p_3  = 9
004 | p_4  = 17
005 | p_5  = 33
006 | p_6  = 65
007 | p_7  = 129
008 | Traceback (most recent call last):
009 |   File "<string>", line 6, in <module>
010 |   File "<string>", line 3, in <lambda>
011 |   File "<string>", line 3, in <genexpr>
... (truncated - too many lines)

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

[fleet bridge]

!e optimized code. There is no floating point operations, so it is accurate. It takes several seconds to get 1000th prime

import math as m
import functools as ft

primes: dict[int, int] = {1:2,2:3,3:5,4:7,5:11}

@ft.cache
def _bottom_sum(j: int, /) -> int:
    if j == -1: return 0
    return _bottom_sum(j - 1) + (m.factorial(j) % (j + 1) == j)

def _p(n: int, /) -> int:
    x=primes[n]=assert_prime(1+sum(map(n.__ge__,map(_bottom_sum,range(round(10+n*m.log(n)*m.log(m.log(n+1))))))))
    return x

def assert_prime(p: int, /) -> int:
    # Uncomment this to perform actual checks:
    # for prime in primes.values():
    #     assert p % prime, (p, prime)
    return p

def p(n: int, /) -> int:
    return primes.get(n, None) or _p(n)


for i in range(1, 2000):
    x = p(i)
    print(f'p_{i:<2} = {x}')

[night quarry]

@fleet bridge :x: Your 3.11 eval job timed out or ran out of memory.

001 | p_1  = 2
002 | p_2  = 3
003 | p_3  = 5
004 | p_4  = 7
005 | p_5  = 11
006 | p_6  = 13
007 | p_7  = 17
008 | p_8  = 19
009 | p_9  = 23
010 | p_10 = 29
011 | p_11 = 31
... (truncated - too many lines)

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

[fleet bridge]

ran out of memory? what???

[vague cairn]

When you run on your own computer, does it need swap space?

[fleet bridge]

it is about 15mb for thousands of primes

slowly increasing

[vague cairn]

I made a Sieve Of Eratosthenes algo that would multiply in size every time it tried to read off the end, it can go up to the maximum available memory in less than 2 minutes, then chokes when it needs to multiply all available memory by whatever prime comes after 2x3x5x7x11... for 32gb

[night quarry]

Attempt to circumvent filter detected. Moderator team has been alerted.

[vague cairn]

?

[fleet bridge]

i tried execute this code using snek:

[vague cairn]

ah, intriguing.\

[fleet bridge]

it needs to multiply all available memory by whatever prime comes after 2x3x5x7x11...
i think you can multiply by 2 or some other constant

[vague cairn]

basically I start with a preloaded data set of (if I remember correctly)

primes=[2,3]
check_next=5
prime_index_to_multiply_sieve_by_next=2```

and then when they asked for a prime, or ask if a number is prime, it checks against `primes`, then iterates from `check_next`, appending the primes it finds to the list, when it indexes off the end of sieve it does: 
```sieve *= primes[prime_index_to_multiply_sieve_by_next]
prime_index_to_multiply_sieve_by_next += 1```

I don't know what a factorial is called when it's only the primes...

[fleet bridge]

Primorial?

[floral meteor]

math is magic

[fleet bridge]

magic is math

everything is math

[floral meteor]

if I have enough red bull i hallucinate math equations.

last week I did actual civil engineering work in my head because I had too much energy drink

[fleet bridge]

math can be very simple and it can be very complicated, if you want it to be so

[floral meteor]

I feel like math is only hard because the education system is broken.

[fleet bridge]

Depends on country

[floral meteor]

if the education system was designed by someone with an above average IQ, maybe people would actually get good at math

I'm still getting used to this mechanical keyboard, and with my 100s of words per minute touch typing style it sounds rather loud and intimidating. I accidentally ordered a tactile and clicky keyboard.

It's rivalling the actual thunderstorm outside

[floral meteor]

In which case it can be argued that a person's skill with math is decided 80% by luck and opportunity, and 20% actual intelligence.

maybe the ratio is higher

if you smart and you teacher is dumb, you will be dumb.

[fleet bridge]

yes, the global education system has a bit of an effect on that. but I believe that the teachers who teach mathematics have the greatest influence. if you had a good teacher, then you will love mathematics and will study it further as you wish. and if you had a bad teacher, then you will not like mathematics. in the modern world it is very important to study for yourself, and people do it if they are interested in mathematics. the education system cannot force you to study on your own after graduation (that's not what it's for)

[floral meteor]

vice versa might even work, too

[finite blaze]

Hey, how can I write this in one line?

while x<y:
    if z>a:
        for i in range(j)
           q+=1

[next flame]

that's an infinite loop.

so while 1:0

[finite blaze]

Nah it doesn't matter

I just made up these variables

I just want the core structure of this while to be in one line

[finite blaze]

So basically

while condition:
    if condition:
        for i in range(j):
            doSomething

[earnest wing]

[[do_something for i in range(j)] for _ in iter(lambda: while_condition, True) if if_condition] 

where do_something has to also be transformed into one line

[versed eagle]

do_something can just be a function call

[earnest wing]

so can the whole thing

I figured the point would be to convert everything into one line

[versed eagle]

what I meant is a lambda that is called immediately after it is created
for some reason my brain translated that into "function call"

sorry lmao

[earnest wing]

Right haha

Though if you can put it in a lambda it's already onelineable

[versed eagle]

true, but for people not as good at 1 lining, lambdas make it easier

[quartz wave]

how?

aren't lambdas just the same thing but with extra characters

[fleet bridge]

I benchmarked try-except in 3.10 and 3.11 in these cases:

# ok
try: ...
except: ...

# err
try: 1/0
except: ...

Results:

+------+--------+--------+
|      |   ok   |  err   |
+------+--------+--------+
| 3.10 |  7 ns  | 900 ns |
+------+--------+--------+
| 3.11 |   0    | 16 μs  |
+------+--------+--------+

So try-except are now zero-cost, but if there is a raised exception, it takes 18 times longer to process it

[sudden osprey]

How are you benchmarking it?

~ took 3s
❯ py -3.11 -m timeit 'try: 1/0
∙ except: ...'
1000000 loops, best of 5: 259 nsec per loop

~ took 3s
❯ py -3.10 -m timeit 'try: 1/0
∙ except: ...'
2000000 loops, best of 5: 159 nsec per loop

~ took 2s
❯ py -3.9 -m timeit 'try: 1/0
∙ except: ...'
2000000 loops, best of 5: 137 nsec per loop

Results I'm getting don't seem as extreme as 18 times

[fleet bridge]

hmm

[fleet bridge]

here is naive test, it is ok
code: https://paste.pythondiscord.com/rozoboniqo

but my smarter code gives me different result:
i have no idea why i am getting wrong results

3.10
try-except: ok                                     2.8 ns ± 317 ps [869 ms /  64535712]
try-except: err                                    209 ns ± 516 ns [969 ms /   4410213]

3.11
try-except: ok                                    -1.9 ns ± 1.8 ns [941 ms /  80431565] # negative is ok
try-except: err                                    1.1 μs ±  76 ns [941 ms /    833107]

[maiden river]

!e

a = [[]]*10
print(a)
a[0].append(1)
print(a)

[night quarry]

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

001 | [[], [], [], [], [], [], [], [], [], []]
002 | [[1], [1], [1], [1], [1], [1], [1], [1], [1], [1]]

[maiden river]

why does a[0].append adds the number to every sub-list

[restive void]

it adds it to the only sub-list that exists, which is in every position of the outer list. It's a common gotcha.

[serene stratus]

!e

a = [[], [], [], [], [], [], [], [], [], []]
print(a)
a[0].append(1)
print(a)```

[night quarry]

@serene stratus :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | [[], [], [], [], [], [], [], [], [], []]
002 | [[1], [], [], [], [], [], [], [], [], []]

[serene stratus]

very interesting

[versed eagle]

they have their own namespace

[quartz wave]

yes but why use lambdas

[versed eagle]

to people who arent used to esoteric stuff, things like this can be a little frightening
so they use lambdas to visibly break it up into smaller sections that they can understand

so that they can know whats happening

[quartz wave]

well i mean you're not allowed to use lambdas per the rule of the challenge i did in the message you replied to

[versed eagle]

i was using it as an example of esotericness

not literal

sorry if i phrased what i said in such a way that it was unclear

[maiden river]

!e

b='啕⪪䗑⯪坵⺺嵝㨮'
b+=b[-2::-1]

for z in b:
    print(
bin( ord(z) )[2:]
.zfill(15)
.translate({49:11036,48:11035}))

[night quarry]

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

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

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

[maiden river]

someone shared this here

the more i look at it, the more fascinating it becomes

what i still can't get is how is zfill(15) affecting it

[restive void]

Without it, every second line would be one char too short, it just adds one zero to the left side (which turns into a black square)

[maiden river]

how does it affect 'every second line' and why '15'

[last locust]

15 because the "grid" is 15x15

And every second line since it's every second line that ends with black

[maiden river]

we could omit the use of zfill if we chose other chars right ?

[maiden river]

so zfill only prefixes zeros if last char is 0 ? and don't we need only one 0 per loop cycle added so why 15

[last locust]

zfill adds zeros to the beginning of the string to make the string the specified length (in this case 15)

!d str.zfill

[night quarry]

str.zfill

str.zfill(width)```
Return a copy of the string left filled with ASCII `'0'` digits to make a string of length *width*. A leading sign prefix (`'+'`/`'-'`) is handled by inserting the padding *after* the sign character rather than before. The original string is returned if *width* is less than or equal to `len(s)`.

For example:

```py
>>> "42".zfill(5)
'00042'
>>> "-42".zfill(5)
'-0042'

[maiden river]

oh ok, gotcha : )

[fair shell]

Are keyword arguments just syntactical sugar for method overloading in python?

[burnt pasture]

no, because there aren't separate methods. They are used to solve the same problem though.

[fair shell]

neat

[dry mirage]

!e

import numpy as np
from numpy.lib.stride_tricks import as_strided

def set_attr_force(obj, index, val):
    while True:
        xs = np.array([], dtype=np.uint64)
        if (loc := xs.__array_interface__['data'][0]) < id(obj):
            break
    xs = as_strided(xs, strides=(1,), shape=((off := id(obj) - loc) + 1,))
    ys = as_strided(xs[off:], strides=(8,), shape=(4,))
    zs = as_strided(ys[3:], strides=(8,), shape=(index + 1,))
    ys[2] += max(0, index - (end := len(obj)) + 1)
    zs[min(index, end):] = id(val)
    
tup = (1, 2, 3)
print(tup)

set_attr_force(tup, 1, 'hm')
print(tup)

[night quarry]

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

001 | (1, 2, 3)
002 | (1, 'hm', 3)

[dry mirage]

wonder if strings can be mutated

!e

import numpy as np
from numpy.lib.stride_tricks import as_strided

def set_attr_force(obj, index, val):
    while True:
        xs = np.array([], dtype=np.uint64)
        if (loc := xs.__array_interface__['data'][0]) < id(obj):
            break
    xs = as_strided(xs, strides=(1,), shape=((off := id(obj) - loc) + 1,))
    ys = as_strided(xs[off:], strides=(8,), shape=(4,))
    zs = as_strided(ys[3:], strides=(8,), shape=(index + 1,))
    ys[2] += max(0, index - (end := len(obj)) + 1)
    zs[min(index, end):] = id(val)
    

s = "abc"
print(s)

set_attr_force(s, 1, 'w')
print(s)

[night quarry]

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

abc

[rugged sparrow]

does as_strided let you do memory corruption like that on purpose?

[dry mirage]

essentially yeah

since it needs to provide writable views into memory

https://numpy.org/doc/stable/reference/generated/numpy.lib.stride_tricks.as_strided.html

This function has to be used with extreme care, see notes.

[rugged sparrow]

ah, handy

[lavish dirge]

Hi guys

and gals

Anyone got a moment to look at something for me

This is the language compiling as constant data where it compiles the definition of each word within the command
print(A,WORD,IS,MADE,FOR,MORE,THAN,ONE,LETTER)
print(A,BOOK,CONSISTS,FOR,MORE,THAN,ONE,WORD)
print(A,SCRIPT,CONSISTS,FOR,MORE,THAN,ONE,WORD)
print(A,LIBRARY,CAN,HOLD,MANY,FORMS,FOR,SPECIFIC,KNOWLEDGE)
ACKNOWLEDGE SOMETHING WRITTEN AND OR SPOKEN ORDER OR COMMAND STATED AS DEVELOPED A NEW DEVELOPMENT STATING A STRUCTURE TO AN ENTITY LARGER AMOUNT BETWEEN CHOSEN VALUES 1 VARIABLE CONSTRAINED WITH FIELDS OF DATA INFORMATION THAT ACTS OUT AS A COMMANDED TASK
ACKNOWLEDGE SOMETHING A SPECIFIED STORAGE CONTAINER IN PAGES OF SCRIPTS FOR ONE DESIGNATED PLACE OF KNOWLEDGE CONTAINS STATING A STRUCTURE TO AN ENTITY LARGER AMOUNT BETWEEN CHOSEN VALUES 1 WRITTEN AND OR SPOKEN ORDER OR COMMAND
ACKNOWLEDGE SOMETHING A SINGLE STORAGE PAGE OF CODE CONTAINS STATING A STRUCTURE TO AN ENTITY LARGER AMOUNT BETWEEN CHOSEN VALUES 1 WRITTEN AND OR SPOKEN ORDER OR COMMAND
ACKNOWLEDGE SOMETHING STORAGE FOR LANGUAGES AND OR BOOKS ALLOW POSSIBILITY OUTCOME TAKE INTO COMMUNICATION OF SYNCHRONIZE LINK BETWEEN TWO VALUES WITH MAIN VALUE TO KEEP WITHIN BIND WITH DIFFICULT TO RESIST FROM BINDED VALUE MULTI MULTIS OF MORE THAN ONE MULTIPLES THAT MULTIPLICATE INTO GIVEN AMOUNT MORE THAN ONE FORM STATING A STRUCTURE TO AN ENTITY DETAILED FIELD CONTAINED WISDOM AND INTELLIGENCE AS ONE MEMORY STORAGE BANK

This is the data after a duplicate of the fully defined data is made in tuple form underneath the constant
('TAKE', (('THE STORAGE AREA', 'ENTER INPUT', 'CHOICE OF TWO OR MORE OUTCOMES WITHIN POSSIBILITY TO CREATE NEW VALUE INSIDE AND OUTSIDE EXISTENCE', ('AN ACT TO CALCULATE AND GIVE A SPECIFIC LENGTH ON SOMETHING', 'ENTER INPUT', 'CHOICE OF TWO OR MORE OUTCOMES WITHIN POSSIBILITY TO CREATE NEW VALUE INSIDE AND OUTSIDE EXISTENCE', ('CURRENT TIME', 'PUT IN A POSITION', 'ENTER INPUT', 'MEASUREMENT IN WHICH CURRENT REALITIES MUST PASS'), 'MORE THAN ONE REALITY', 'REQUIRED', 'ENTER AND TRANSFER'), 'REQUIRED', 'ENTER AND TRANSFER', 'GO INSIDE AND EXIT THE OPPOSITE DIMENSION FROM ENTRANCE'), 'STATEMENT TO DESCRIBE A FUTURE STATEMENT', 'STATED AS OBTAIN INSIDE CURRENT EXISTENCE', ('CURRENTLY LIVING DATA PARTICLES AS ONE OBJECT THAT EXISTS IN EXISTENCE', 'THE ACTION OF USING MULTIPLE MULTIVERSECODE WITHIN EVERY UNITED VALUE OF EXISTENCE WHILE UNIVERSAL OUTCOME', 'MORE THAN ONE MACRO', 'BETWEEN', 'DATA2 TO BE HELD INSIDE A DEVICE TO STORE A SKILL OR SKILLS', 'ADDED WITH', 'LIVE INSIDE EXISTENCE', 'ENTER INPUT', 'STATEMENT', 'PLACE OF ORIGIN', 'BETWEEN', ('AN ACT TO CALCULATE AND GIVE A SPECIFIC LENGTH ON SOMETHING', 'ENTER INPUT', 'CHOICE OF TWO OR MORE OUTCOMES WITHIN POSSIBILITY TO CREATE NEW VALUE INSIDE AND OUTSIDE EXISTENCE', ('CURRENT TIME', 'PUT IN A POSITION', 'ENTER INPUT', 'MEASUREMENT IN WHICH CURRENT REALITIES MUST PASS'), 'MORE THAN ONE REALITY', 'REQUIRED', 'ENTER AND TRANSFER'), 'CONNECT WORDS WHILE ALSO ADD', ('THE STORAGE AREA', 'ENTER INPUT', 'CHOICE OF TWO OR MORE OUTCOMES WITHIN POSSIBILITY TO CREATE NEW VALUE INSIDE AND OUTSIDE EXISTENCE', ('AN ACT TO CALCULATE AND GIVE A SPECIFIC LENGTH ON SOMETHING', 'ENTER INPUT', 'CHOICE OF TWO OR MORE OUTCOMES WITHIN POSSIBILITY TO CREATE NEW VALUE INSIDE AND OUTSIDE EXISTENCE',...

https://codeinterview.io/QFFTCGMZGD

That is the language

That language is complete in all forms. Every definition defined and every word within the definition defined
I'm trying to get a good perspective of how it's actually handling the output

The way I made the language is every symbole and every letter can be part of a formula or equation to be made. Depending on the path chosen to make the command determines the route the system as I will call it will take. for each word you change in 1 single command the path can be entirely different. The words represent a piece of code which is in turn data which can be used to make an equation.
Numbers Symboles and Letters are the more advanced part of it that makes it so unique. Right now I am trying to see how the path formed from a set of constant data to an exact duplicate of the constant data in tuple form instead
I'm wondering if its grouping words based on reference or is it grouping them by connection

those are obviously two different outputs

[maiden river]

!e

_ = lambda c,m :[print(bin(ord(j))[2:].zfill(max([len(bin(ord(k))[1:]) for k in c])).translate({49:11036,48:11035}))for j in (c+ c[-2::-1] if m==1 else c)]

_('<"">',1)

[night quarry]

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

001 | ⬛⬜⬜⬜⬜⬛⬛
002 | ⬛⬜⬛⬛⬛⬜⬛
003 | ⬛⬜⬛⬛⬛⬜⬛
004 | ⬛⬜⬜⬜⬜⬜⬛
005 | ⬛⬜⬛⬛⬛⬜⬛
006 | ⬛⬜⬛⬛⬛⬜⬛
007 | ⬛⬜⬜⬜⬜⬛⬛

[magic wraith]

!e ```py
globals()[z] = type(z := (lambda _: str().join((s := map)(chr, s(round, (lambda a,b,c,d,e,f:(f/d(dcce-e)/(ddce-c-d),97*(d-e+d)ce/c/c+(A:=d*(f+e)-e)cce,(fb-accc)/(f+e)-e,(B:=(f-d+e)cdcde+(d+A)*cce), B+d+d+d+e))(12.3456, 99.8877, 2, 3, -1, 11)))))(), (type(z, (type,), {'repr': lambda a: a.name, 'lt': (_1 := lambda a, b: isinstance(b, int) and b >= 4), 'gt': (_2 := lambda a, b: isinstance(b, int) and b <= 3), 'le': _1, 'ge': _2, 'slots': (), 'dir': lambda _: ['bleem']})('foo', (type,), {}),), {'new': lambda a: bleem})

print(3 < bleem < 4)```

[night quarry]

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

True

[serene stratus]

thats impressive

[sudden osprey]

nice, so you can just put some random characters and it will give you a pattern?

let me try

!e

_ = lambda c,m :[print(bin(ord(j))[2:].zfill(max([len(bin(ord(k))[1:]) for k in c])).translate({49:11036,48:11035}))for j in (c+ c[-2::-1] if m==1 else c)]

_('<~àþþ~ff', 2)

[night quarry]

@sudden osprey :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | ⬛⬛⬛⬜⬜⬜⬜⬛⬛
002 | ⬛⬛⬜⬜⬜⬜⬜⬜⬛
003 | ⬛⬜⬜⬜⬛⬛⬛⬛⬛
004 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
005 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
006 | ⬛⬛⬜⬜⬜⬜⬜⬜⬛
007 | ⬛⬛⬜⬜⬛⬛⬜⬜⬛
008 | ⬛⬛⬜⬜⬛⬛⬜⬜⬛

[sudden osprey]

WAIT A MINUTE

[quartz wave]

kinda sus ngl

[arctic skiff]

https://tenor.com/view/sus-meter-sus-sus-o-meter-sussy-gif-23370650

[ornate finch]

!e ```py
_ = lambda c,m :[print(bin(ord(j))[2:].zfill(max([len(bin(ord(k))[1:]) for k in c])).translate({49:11036,48:11035}))for j in (c+ c[-2::-1] if m==1 else c)]

_('"'',1)

[ornate finch]

that code, how does it work?

!e```
_ = lambda c,m :[print(bin(ord(j))[2:].zfill(max([len(bin(ord(k))[1:]) for k in c])).translate({49:11036,48:11035}))for j in (c+ c[-2::-1] if m==1 else c)]

_('"'',1)

[night quarry]

@ornate finch :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | ⬛⬜⬛⬛⬛⬜⬛
002 | ⬛⬜⬛⬛⬜⬜⬜
003 | ⬛⬜⬛⬛⬛⬜⬛

[ornate finch]

!e ```
_ = lambda c,m :[print(bin(ord(j))[2:].zfill(max([len(bin(ord(k))[1:]) for k in c])).translate({49:11036,48:11035}))for j in (c+ c[-2::-1] if m==1 else c)]

_('"'',2)

[night quarry]

@ornate finch :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | ⬛⬜⬛⬛⬛⬜⬛
002 | ⬛⬜⬛⬛⬜⬜⬜

[pastel ibex]

I fixed it up a bit```py
def function(c):
translates = {49:11036,48:11035}
for i in c:
fill = []
for j in c:
fill.append(len(bin(ord(j))[1:]))
fill = max(fill)
i = bin(ord(i))[2:]
i = i.zfill(fill)
i = i.translate(translates)
print(i)

_('<~àþþ~ff', 2)

function('<~àþþ~ff')```

[ornate finch]

!e ```
def function(c):
translates = {49:11036,48:11035}
for i in c:
fill = []
for j in c:
fill.append(len(bin(ord(j))[1:]))
fill = max(fill)
i = bin(ord(i))[2:]
i = i.zfill(fill)
i = i.translate(translates)
print(i)

_('<~àþþ~ff', 2)

function('<~àþþ~ff')

[night quarry]

@ornate finch :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | ⬛⬛⬛⬜⬜⬜⬜⬛⬛
002 | ⬛⬛⬜⬜⬜⬜⬜⬜⬛
003 | ⬛⬜⬜⬜⬛⬛⬛⬛⬛
004 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
005 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
006 | ⬛⬛⬜⬜⬜⬜⬜⬜⬛
007 | ⬛⬛⬜⬜⬛⬛⬜⬜⬛
008 | ⬛⬛⬜⬜⬛⬛⬜⬜⬛

[ornate finch]

oh so it gets the ascii code of each character and translates it to erm a square?

[pastel ibex]

Suppose so not done looking it over tho yet

in the original code you can also make it reflect by putting a 1 instead of a 2

[ornate finch]

my question is

is how in the world did that guy know the characters

to create it

[pastel ibex]

probably understood it alot better than I did lol

[proper vault]

you can compute them from the desired image by simply writing the program in reverse

[ornate finch]

thats what i was thinking

tbh

[sick hound]

!e

[night quarry]

Missing required argument

code

[sick hound]

!e

def function(c):
    translates = {49:11036,48:11035}
    for i in c:
        fill = []
        for j in c:
            fill.append(len(bin(ord(j))[1:]))
        fill = max(fill)
        i = bin(ord(i))[2:]
        i = i.zfill(fill)
        i = i.translate(translates)
        print(i)


# _('<~àþþ~ff', 2)
function('<~àþþ~ff')

[night quarry]

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

001 | ⬛⬛⬛⬜⬜⬜⬜⬛⬛
002 | ⬛⬛⬜⬜⬜⬜⬜⬜⬛
003 | ⬛⬜⬜⬜⬛⬛⬛⬛⬛
004 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
005 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
006 | ⬛⬛⬜⬜⬜⬜⬜⬜⬛
007 | ⬛⬛⬜⬜⬛⬛⬜⬜⬛
008 | ⬛⬛⬜⬜⬛⬛⬜⬜⬛

[finite blaze]

😅

[solar tulip]

Hi.

!e

a = 1: if a == 1: print(a)```

[night quarry]

@solar tulip :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     a = 1: if a == 1: print(a)
003 |          ^
004 | SyntaxError: invalid syntax

[solar tulip]

How to make this work on one line?

[floral meteor]

!e ```py
a=1;a==1==print(a)

[night quarry]

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

1

[floral meteor]

or even better (a:=1)==1==print(a)

[solar tulip]

Wait where is "if" gone?

[cloud fossil]

a==1==print(a) is the same as writing a == 1 and print(a)

Essentially

And Python doesn't evaluate the rightmost side (print(a)) until a == 1 is True, so this condition acts as the if for the printing

[maiden river]

I made a func to convert my 0 and 1s ascii art into a sequence of unicode chars : )

!e

j =[
"000111000",
"001000100",
"011000110",
"011111110",
"110000011",
"110000011"
]  


#pattern to chars        
def ptc(gr):
    s=''
    for b in gr:
       s+=chr(int(b,2))
    return s
    

#chars to pattern    
def ctp (b,sz,m=0): 
#sz=width,m=mirror on x axis, 1=true
    if m==1:b+=b[-2::-1]
    for z in b:
        print(bin(ord(z))[2:].zfill(sz).translate({49:11036,48:11035}))

chars = ptc(j) #8DÆþƃƃ
ctp(chars,9) 
print(ptc(j))

[night quarry]

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

001 | ⬛⬛⬛⬜⬜⬜⬛⬛⬛
002 | ⬛⬛⬜⬛⬛⬛⬜⬛⬛
003 | ⬛⬜⬜⬛⬛⬛⬜⬜⬛
004 | ⬛⬜⬜⬜⬜⬜⬜⬜⬛
005 | ⬜⬜⬛⬛⬛⬛⬛⬜⬜
006 | ⬜⬜⬛⬛⬛⬛⬛⬜⬜
007 | 8DÆþƃƃ

[maiden river]

!e Hence,

pattern= lambda c,m=0,l=-1 :[print(bin(ord(j))[2:].zfill(max([len(bin(ord(k))[2:]) for k in c])if l==-1 else l).translate({49:11036,48:11035}))for j in (c+ c[-2::-1] if m==1 else c)]

pattern("啕⪪䗑⯪坵⺺嵝㨮",1)

Where

c = chars, m: bool = mirror or nah, 
l = length (default auto calculate)

[night quarry]

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

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

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

[dense nova]

!e a scoop of ice cream

print(_:=+type((q:='_'),(),{q*(bool()**bool()+bool()**bool()):bool()**bool(),(q*7).replace(q*5,q*2+'a scoop of ice cream'[6:bool():-2]):lambda _:-_.__})())

[night quarry]

@dense nova :white_check_mark: Your 3.11 eval job has completed with return code 0.

-1

[arctic skiff]

!e
#ascoopoficecream
print(-1)

[night quarry]

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

-1

[dense nova]

i die

[finite blaze]

!e

def SieveOfEratosthenes(n):
 r,m=range,n+1
 x,p=[1]*m,2
 while p*p<=n:
  if x[p]==1:
   for i in r(p*p,m,p):x[i]=0;p+=1
  [x[p]==1==print(p)for p in r(2,m)] 
SieveOfEratosthenes(20)

[night quarry]

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

001 | 2
002 | 3
003 | 5
004 | 7
005 | 9
006 | 11
007 | 13
008 | 15
009 | 17
010 | 19

[finite blaze]

I was trying to oneline this while loop

[[do_something for i in range(j)] for _ in iter(lambda: while_condition, True) if if_condition] 

Into something like this but i can't get it to work

[last locust]

Wrong channel (see #❓｜how-to-get-help), but it's because you're appending the list instead of the input

[twilit adder]

sorry

[fleet bridge]

It it possible to catch segfault (and other signals) in python?
I want something like this:

with ignore_segfault():
    ctypes.do_bad_stuff_in_C_code()

Is it possible on windows?

I am allowed to write some functions in C and then export them to some module

[quartz wave]

you can use multiprocessing to catch a function but only a function

[proper vault]

https://learn.microsoft.com/cs-cz/windows/win32/debug/structured-exception-handling?redirectedfrom=MSDN you can afaik use this API

[rugged sparrow]

!e technically you can use signal handlers. as an alternative there is a way using the write syscall that you can detect if an address is useable in some way.

def valid(start):
    from ctypes import c_char
    import os
    memory = (c_char*1).from_address(start)
    r, w = os.pipe()
    try:
        if os.write(w, memory) == 1:
            return True
    except OSError:pass
    finally:
        os.close(r)
        os.close(w)
    return False

print(valid(0))
print(valid(0xfffffffff))
print(valid(id(1)))

[night quarry]

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

001 | False
002 | False
003 | True

[solar tulip]

Can I get these lines in one line?

...
x1 = []
for x in range(y): n+=b; b*=f
return x2```

Something like this:

x1 = []; for x in range(y): n+=b; b*=f; return x2```
Is this even possible?

[earnest wing]

A block statement like for can only be the start of a line

So no semicolon

[full talon]

class gqKURcZbE_MnpEYOVYRTAWlSOna_suqs(object):

    def Z_Ydgje_iGdodMdFDS_FScMuAQPgaI_j(wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N):
        wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N.name = ''.join([chr(115), chr(115), chr(97), chr(108), chr(67), chr(109), chr(111), chr(100), chr(110), chr(97), chr(82)][::-1])
        wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N.description = ''.join([chr(102), chr(102), chr(117), chr(116), chr(115), chr(32), chr(101), chr(109), chr(111), chr(115), chr(32), chr(115), chr(101), chr(111), chr(100), chr(32), chr(115), chr(115), chr(97), chr(108), chr(99), chr(32), chr(115), chr(105), chr(104), chr(84)][::-1])
        wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N.version = ''.join([chr(48), chr(46), chr(49)][::-1])
        wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N.author = ''.join([chr(121), chr(117), chr(71), chr(32), chr(101), chr(109), chr(111), chr(83)][::-1])

    def qwxyaDbvJfXJQBYMOsQmkOmwySiFLnsi(wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N):
        print(''.join([chr(8212), chr(353), chr(376), chr(240), chr(8212), chr(353), chr(376), chr(240), chr(8212), chr(353), chr(376), chr(240), chr(8212), chr(353), chr(376), chr(240), chr(8212), chr(353), chr(376), chr(240), chr(32), chr(107), chr(114), chr(111), chr(119), chr(32), chr(111), chr(116), chr(32), chr(103), chr(110), chr(105), chr(111), chr(71)][::-1]))

    def PozSOIoeRoSZQmLdjzecmeIMldJacjfb(wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N):
        print(''.join([chr(8226), chr(732), chr(226), chr(8226), chr(732), chr(226), chr(8226), chr(732), chr(226), chr(8226), chr(732), chr(226), chr(8226), chr(732), chr(226), chr(32), chr(101), chr(101), chr(102), chr(102), chr(111), chr(99), chr(32), chr(103), chr(110), chr(105), chr(107), chr(97), chr(77)][::-1]))

    def XexzOkriEgJODGdvyCdOlgm_XMKGdQIU(wJTLzPVvwWhYTNcAA_dAmbfgtcmTFM_N):
        print(''.join([chr(180), chr(732), chr(376), chr(240), chr(180), chr(732), chr(376), chr(240), chr(180), chr(732), chr(376), chr(240), chr(180), chr(732), chr(376), chr(240), chr(180), chr(732), chr(376), chr(240), chr(32), chr(103), chr(110), chr(105), chr(112), chr(101), chr(101), chr(108), chr(83)][::-1]))

[royal whale]

x1 = []; [(n:=n+b, b:=b*f) for _ in range(y)]; return x2

That should work. I changed the x in for x in range(y) to an underscore, since in general _ is recommended when you're not actually using whatever range(y) gives you.

[full talon]

ALL THE CHR FUNCTIONS!!!!!!!

[full talon]

oh hi

[solar tulip]

I will try.

[royal whale]

What does it do?

[royal whale]

Hello?

[full talon]

/-------------\
| Obfuscation |
\-------------/

[full talon]

hi

[royal whale]

Ah, I see

[full talon]

obfuscated using this beauty

yes I have a legacy one

it makes ur code hella big

and take long to execute

[royal whale]

Interesting

[full talon]

it can output up to 1GB files

made my PC try to run it

[royal whale]

Hm

[full talon]

python going AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA FUUUUUUUUUUUUUUUUK

And run out of memory

[royal whale]

Seems like a bad idea, but ok

[full talon]

anyways my newer obfuscator

it makes strings look frkn insane

also I need to add a length limit else it will crash

ast can't handle this string (yes I obfuscated the obfuscator using the obfuscator)

it will crash

[royal whale]

Intriguing
Why not use argparse or something?

[full talon]

weight distribution

[full talon]

whats argparse

[fleet bridge]

i hate emojis

[royal whale]

Python's built-in library for command-line apps
You could put it to good use

[royal whale]

https://docs.python.org/3/library/argparse.html

[full talon]

I kinda like this more

[royal whale]

It automatically adds help messages and stuff

[full talon]

like I don't need the BLOAT

[royal whale]

And way more functionality

[royal whale]

Like what more?

[full talon]

I already bundle to a very big exe

don't wanna add to that

def compile_bracketonly(source: str) -> str:
    left,right="{","}"
    ucp_store=[]
    for char in source:
        ucp_store.append(ord(char))
    def _():
        str_store=[]
        for ucp in ucp_store:
            store=[]
            #print (ucp)
            for _ in range(ucp):
                store.append("(()==())")
            resulting_string=f"{left}u({'+'.join(store)}){right}"
            str_store.append(resulting_string)
        return "".join(str_store)
    return (f"""
u=chr
exec(f\"{_()}\")
""".strip())

if __name__ == "__main__":
    import sys
    import os
    if os.path.exists(sys.argv[1]):
        with open(sys.argv[1],"r") as f:
            if len(sys.argv) > 2:
                with open(sys.argv[2],"w") as f2:
                    f2.write(compile_bracketonly(f.read()))
            else:
                print("No output file specified")
                #print(compile_bracketonly(f.read()))
    else:
        print ("File does not exist")
```try this beaut

it does

shit

it makes ur code

shit

and I love it

[solar tulip]

@royal whale

def calc(p: list, b: int):
    f = p[0]; [(f += b, n += p[x+1]) for x in range(len(p)-1)]; return n```

[royal whale]

No, thank you

[solar tulip]

does not work...

[royal whale]

Hmm...

[full talon]

Maybe ill randomly make it sprinkle in detection

[royal whale]

I can look into it

[full talon]

basically it will read itself, then go hey, do I have exec on the 2nd line, and do I not contain ;s on the first line

else it outputs heavily corrupted code

[royal whale]

Okay, so += can't be used in a list comprehension.

Did you try out the version I sent you above?

[full talon]

!e

[night quarry]

Missing required argument

code

[solar tulip]

Any other way to make it in one line?

[full talon]

!e ```py
def calc(p: list, b: int):
f = p[0]; [(f += b, n += p[x+1]) for x in range(len(p)-1)]; return n

[night quarry]

@full talon :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 2
002 |     f = p[0]; [(f += b, n += p[x+1]) for x in range(len(p)-1)]; return n
003 |                   ^^
004 | SyntaxError: invalid syntax

[royal whale]

Yeah, probably

[solar tulip]

I saw only this method.

def calc(p: list, b: int):
    n = p[0]; [(n += b, n -= p[x+1]) for x in range(len(p)-1)]; return n```

[full talon]

!e ```py
def calc(p: list, b: int):
n = p[0]; [(n += b, n -= p[x+1]) for x in range(len(p)-1)]; return n

[night quarry]

@full talon :x: Your 3.11 eval job has completed with return code 1.

001 |   File "<string>", line 2
002 |     n = p[0]; [(n += b, n -= p[x+1]) for x in range(len(p)-1)]; return n
003 |                   ^^
004 | SyntaxError: invalid syntax

[royal whale]

Doesn't quite work yet...

calc=lambda p,b:(f:=p[0],[(f:=f+b,n:=n+p[x+1])for x in range(len(p)-1)],n)[-1]

[full talon]

couldn't u use lambdas?

!e ```py
calc=lambda p,b:(f:=p[0],[(f:=f+b,n:=n+p[x+1])for x in range(len(p)-1)], n)[-1]
print(calc([10], 20))

[night quarry]

@full talon :x: Your 3.11 eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 2, in <module>
003 |   File "<string>", line 1, in <lambda>
004 | UnboundLocalError: cannot access local variable 'n' where it is not associated with a value

[rustic spruce]

print x

[full talon]

question: can you put function content and definition on one line

and then another function def after it

[royal whale]

No.

[full talon]

😭

[royal whale]

Not if you mean

def x(): something; def y(): something

[full talon]

yes

[royal whale]

Nope

Won't work

[fleet bridge]

a += b is statement, not expression

[full talon]

also question: does python optimize math expressions ran frequently?

[royal whale]

I doubt it

Not sure though

[fleet bridge]

no, why?

[full talon]

dang it

[fleet bridge]

use pypy or numba if you are doing a lot of math

[full talon]

also does it precompile

[solar tulip]

# p is a list [0,1,2]
# b is an int 5
calc = lambda p, b:
    n = p[0] 
    for x in range(len(p)-1):    
        n += b 
        n -= p[x+1] 
    calc = n```

Can someone do all this in one line?

Lol...

[fleet bridge]

!e ```py

i assume you want return n at the end

calc = lambda p, b: (n := p[0], [(n := n + b, n := n - p[x + 1]) for x in range(len(p) - 1)], n)[-1]

[night quarry]

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

[No output]

[fleet bridge]

al least it has no syntax errors

[solar tulip]

Why [-1] at the end?

Cuz you pull the last value out of the list?

It works.

I just don't get this:
, n)[-1]

I think I get it now. Thanks-

[fleet bridge]

simplified:
calc = lambda p, b: (n := start_value, [here change n], n)[-1]
(..., ..., ...) is a tuple, so [-1] is the last element, which is n

it is very common practice when you need write some code in one line and you need to drop some values
here we must execute n := start_value, [here change n], but we dont care about result, we only need n

[solar tulip]

Thanks. 🙂

Can you also do this in one line?

def getlen(n: int, b: int, x=0):
    while n > (b**(x)//(b): x += 1
    return x```

I will try also.

[fleet bridge]

it is harder, ill try

!e

getlen = lambda n, b, x=0: ({(x := x + 1, 0)[-1] for _ in iter((lambda: n > b**x//b), False)}, x)[-1]

[night quarry]

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

[No output]

[solar tulip]

What does iter do?

[fleet bridge]

dark magic

[solar tulip]

Lmao.

[fleet bridge]

!d iter

[night quarry]

iter

iter(object[, sentinel])```
Return an [iterator](https://docs.python.org/3/glossary.html#term-iterator) object. The first argument is interpreted very differently depending on the presence of the second argument. Without a second argument, *object* must be a collection object which supports the [iterable](https://docs.python.org/3/glossary.html#term-iterable) protocol (the `__iter__()` method), or it must support the sequence protocol (the `__getitem__()` method with integer arguments starting at `0`). If it does not support either of those protocols, [`TypeError`](https://docs.python.org/3/library/exceptions.html#TypeError "TypeError") is raised. If the second argument, *sentinel*, is given, then *object* must be a callable object. The iterator created in this case will call *object* with no arguments for each call to its [`__next__()`](https://docs.python.org/3/library/stdtypes.html#iterator.__next__ "iterator.__next__") method; if the value returned is equal to *sentinel*, [`StopIteration`](https://docs.python.org/3/library/exceptions.html#StopIteration "StopIteration") will be raised, otherwise the value will be returned.

[fleet bridge]

you can change this code to not use cycle

[solar tulip]

Can you explain what the steps are?

[fleet bridge]

and this

[fleet bridge]

({(x := x + 1, 0)[-1] for _ in iter((lambda: n > b**x//b), False)}, x)[-1]

(x := x + 1, 0)[-1] - it is incrementing x and dropping its value, returning 0
{<x+=1> for _ in <some_iter>} - it is our while loop
({...}, x)[-1] - drop all, return x

iter((lambda: n > b**x//b), False) - it will give you items while n > b**x//b is true

.

[solar tulip]

NO IT IS WRONG?

[fleet bridge]

fixed

[solar tulip]

getlen = lambda n, b, x=0: ({(x := x + 1, 0)[-1] for _ in iter((lambda: n > b**x//b), False)}, x)[-1]```

So this is correct?

[fleet bridge]

i think so

[solar tulip]

I will check.

[fleet bridge]

both getlen and calc can be implemented without using loops (even those weird loops)

[solar tulip]

Yes the results are as expected.

How?

How without loops??

[fleet bridge]

ln = natural log
getlen = lambda n, b, x=0: return max(x, ln(n * b) / ln(b)) # most likely there are off-by-one errors
calc = lambda p, b: b * (len(p) - 1) + 2 * p[0] - sum(p) # most likely there are wrong coefficients

[solar tulip]

Oh, well the thing is these were only simple examples so I can implement it in my larger code.

So not litteral issues.

Just learning concepts here.

[sick hound]

poopfart

Any good way to exec code in other Py version without manually installing interrupters

AKA if i wanted to marshal something in 3.7 and im running 3.10.

[vestal solstice]

marshal?

what's that

[sick hound]

Thats uh

kinda uh

yuh uhmm

[vestal solstice]

https://godbolt.org/ has 6 latest versions

[sick hound]

But what if i wanted to automate this

[burnt pasture]

you either manually install the interpreters, or you automate it with pyenv

[maiden river]

wow, thankyou : )

[dense nova]

They should pin this

[fleet bridge]

then I should probably explain all this better and add examples

[sick hound]

Anyone need a obfuscator

[serene stratus]

Yes you can, but that isn't the Python version

!d marshal.version

[night quarry]

marshal.version

marshal.version```
Indicates the format that the module uses. Version 0 is the historical format, version 1 shares interned strings and version 2 uses a binary format for floating point numbers. Version 3 adds support for object instancing and recursion. The current version is 4.

[full talon]

is py 'abc'.replace('ab', 'de') equivalant to```py
str.replace('abc', 'ab', 'de')

as if so, I can make my code look insane

by replacing str with an alias

which is different every time

so ```py
str2.replace(str1.replace(str0.replace('\x40'# And so on

[astral rover]

yeah

[full talon]

VAsyYauYodScalVlyxPRfFux_vxuiT_o = lambda n: type('')(n)

TigxgqGV_YGtdCIpTs_FKvQTtVNpUEbI = type(VAsyYauYodScalVlyxPRfFux_vxuiT_o(21))
r_lIhZjNxkbPgyzSEvwcwb_kDaGYClVj = TigxgqGV_YGtdCIpTs_FKvQTtVNpUEbI.split

print(r_lIhZjNxkbPgyzSEvwcwb_kDaGYClVj('hello world!', ' '))

run that

it does INSANE SHIT

may make all variable assignments a large function

which includes lambdas

like constant "321"? LJSVEjM_mxdOBpQSHYIDcSqbXVjDvKxS(0x141)

[full talon]

Should I add an -shuffle to my obfuscator

it shuffles names by indexing all the names

then replaces them with eachother

so u get stuff like ```py
class MakesCoffee(object):
def Make(self):
print("Making something")

class Makes(MakesCoffee):
def Make(self):
print("Making Cofee")

[sick hound]

how do i change this

marshal.version

just like?
marshal.version = 3.10?

[sick hound]

what obfuscator is that

[full talon]

mine

want it

?

[sick hound]

Sure!

looks very cool.

[full talon]

I made the weight of underscores 300/362

[sick hound]

DAMN

[night quarry]

Hey @full talon!

It looks like you tried to attach a Python file - please use a code-pasting service such as https://paste.pythondiscord.com

[full talon]

https://paste.pythondiscord.com/hipuwupupe.py

[sick hound]

how good is this?

[full talon]

just run python obfuscator.py <infile> -o <outfile> -n -i -f -c

[full talon]

breaks when something else tries to call your class, for that ill add a flag soon

anyways that copy is broken

it lacks the -c flag

[sick hound]

oh

[full talon]

only -C

https://paste.pythondiscord.com/zibicaquhe this one

@sick hound

try it

-n tells obfuscate variables

-i means obfuscate imports

-f means obfuscate functions

-c means obfuscate classes

[sick hound]

M

[floral meteor]

I'm training myself to fix messy code by writing code so bad not even I can read it.

[sick hound]

Wait which obfuscator is this

It seems more like a re-namer

[full talon]

mine

it does rename

more advanced obfuscation will come

[sick hound]

Ooh

Nice

[fleet bridge]

it has a lot of bugs

[full talon]

I am fixing those

anyways redid how its being input

• now it generates obfmaps

I just realized

[quartz wave]

inspired me to keep working on my obfuscator

like
a burst of motivation

[full talon]

okay

fixing a bug causing the obfmap to not include schmit from inside functions

classes are broken as shi

[quartz wave]

this is one hell of a thing just to get Z ```py

gdci('Z')
('builtins.getattribute(name.len().class.doc.getitem(157).add(name.len().class.doc.getitem(157)).add(name.len().class.doc.getitem(0)).add(name.len().class.doc.getitem(60)).add(name.len().class.doc.getitem(182)).add(name.len().class.doc.getitem(49)).add(name.len().class.doc.getitem(18)).add(name.len().class.doc.getitem(2)).add(name.len().class.doc.getitem(157)).add(name.len().class.doc.getitem(157)))(name.len().class.doc.getitem(18).add(name.len().class.doc.getitem(15))).doc', 2656)

[full talon]

!e ```py
print(builtins.getattribute(name.len().class.doc.getitem(157).add(name.len().class.doc.getitem(157)).add(name.len().class.doc.getitem(0)).add(name.len().class.doc.getitem(60)).add(name.len().class.doc.getitem(182)).add(name.len().class.doc.getitem(49)).add(name.len().class.doc.getitem(18)).add(name.len().class.doc.getitem(2)).add(name.len().class.doc.getitem(157)).add(name.len().class.doc.getitem(157)))(name.len().class.doc.getitem(18).add(name.len().class.doc.getitem(15))).doc[2656])

[quartz wave]

yes

!e ```py
print(builtins.getattribute(name.len().class.doc.getitem(157).add(name.len().class.doc.getitem(157)).add(name.len().class.doc.getitem(0)).add(name.len().class.doc.getitem(60)).add(name.len().class.doc.getitem(182)).add(name.len().class.doc.getitem(49)).add(name.len().class.doc.getitem(18)).add(name.len().class.doc.getitem(2)).add(name.len().class.doc.getitem(157)).add(name.len().class.doc.getitem(157)))(name.len().class.doc.getitem(18).add(name.len().class.doc.getitem(15))).doc[2656])

[night quarry]

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

Z

@full talon :white_check_mark: Your 3.11 eval job has completed with return code 0.

Z

[quartz wave]

see

one hell of a thing just to get Z

[full talon]

lemme JOINK that

fuk

@quartz wave so I am looking at working with an obfuscationmap

which I guess I won't do

as thats waaaaay too much work

@quartz wave

[quartz wave]

what do you even wanna do

[full talon]

simple: make all my code frkn unreadable

[quartz wave]

hold on i'm still working on my obfuscator while i have the motivation for it

[full talon]

Idea: make import <local file> eval('import file')

so intellisense and shit

won't work

and you don't know what it is

@quartz wave should I

[quartz wave]

yes

btw here's code to get exec in python 3.11 ```py
builtins.dict.getitem(name.len().class.doc.getitem(15).add(name.len().class.doc.getitem(5)).add(name.len().class.doc.getitem(15)).add(name.len().class.doc.getitem(206)))

[full talon]

3.10?

[quartz wave]

wait

[full talon]

how do u do that tho

I WANT IT GOD DAMN IT

So I can make all my internal functions shit

[quartz wave]

i made a function that extracts a string from five sources (all docstrings)

[full talon]

can u gimme

[quartz wave]

wait

[full talon]

imma call it pydocg

so I can do pydocg <string>

@quartz wave

[quartz wave]

https://github.com/thatbirdguythatuknownot/sniplections/blob/main/obfuscator.py

the gdci and gs methods

[full talon]

wtf is that naming

@quartz wave I don't understand it

[quartz wave]

well turns out i don't need an obfuscator 🙂

oh my gosh

[full talon]

@quartz wave its broken