esoteric-python

[sick hound]

[1] 21495 segmentation fault (core dumped) python

[rugged sparrow]

It looks like the strings are all being placed one after another in memory, cpython can use memory pools for allocation, so it's likely that the compiler allocates a pool large enough for all of the objects, then places all of them into that memory pool right after one another

[quartz wave]

btw it seems like the PyDictValues part of interned when i investigated further

[rugged sparrow]

That's entirely likely

Memory allocation is a funky platform dependent thing

If you remind me about this in like 2 weeks I can do more investigating

[coral sun]

hi

[quartz wave]

hello

[night quarry]

Nah.

Sorry, you can't do that here!

[coral sun]

I'm studying fastAPI

[quartz wave]

ok

[turbid lodge]

data compression library, but i can't see the code.. so i will assume its sus

[quartz wave]

it's qualifier/tests.py in the code jam 2022 qualifier

[quartz wave]

<frame>.f_code?

[floral meteor]

!remind @floral meteor 2w remind chilaxan memory allocation investigation

[night quarry]

I'm sorry Dave, I'm afraid I can't do that.

Sorry, you can't do that here!

[floral meteor]

Awwww

This channel's the only place I don't ignore pings

[quartz wave]

!e ```py
hing=type('',(),{'bool':lambda s:False not in globals().values()})()
if not hing is False:
print("oh ok")
g = False
if not hing is False:
print("hmm")
else:
print("ok")

[night quarry]

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

001 | oh ok
002 | hmm

[quartz wave]

???

[knotty bronze]

fastAPI works by taking advantage of modern web technologies like asyncio and websockets. It is a high-performance web framework that allows developers to build web applications quickly and easily. fastAPI is built on top of the asyncio library, which makes it easy to write asynchronous code.

[floral meteor]

Comparison of objects does not call bool

And I think is has precedence over not

It's like ==, but for the id of the object pointer

[quartz wave]

oh wait operator precedence

[earnest wing]

yep, not being looser than is is sometimes unexpected

[floral meteor]

if(not hing)is False
is different to
if not hing is False
which is equivalent to
if not(hing is False)

[quartz wave]

!e ```py
hing=type('',(),{'bool':lambda s:False not in globals().values()})()
if (not hing) is False:
print("oh ok")
g = False
if (not hing) is False:
print("hmm")
else:
print("ok")

[night quarry]

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

001 | oh ok
002 | ok

[regal meadow]

def fibonacci():
    """display the fibonacci sequence"""
    a = 1
    b = 0
    c = 0
    termlimit = int(input('What term do you want to go to?\n'))
    if termlimit == 0:
        c = 0
        print('the 0th term in the sequence will be ', c)
    else:
        returnme = []
        n = 0
        while n != termlimit:
           c = a + b
           a = b
           b = c
           returnme.insert(0,c)
           n = n + 1
        print(returnme)
        
"""main function begins here"""
fibonacci()
``` is there anyway I could improve upon this function?

[earnest wing]

remove all the newlines

[fleet bridge]

dont use two or more letters for variable names
better dont use variables at all

[sick hound]

Do it with generator. It will be proper esoteric python

[fleet bridge]

make FibonacciEnterpriseEdition

[sick hound]

def fib(n): a, b = 0, 1 for _ in range(n): yield a a, b = b, a + b

[old socket]

Indentation is wrong

Also how is this esoteric

[sick hound]

Generator

[old socket]

How is that esoteric

Also indentation is still wrong

you have a non-working fib

[royal whale]

!e

from functools import reduce
fib=lambda n:reduce(lambda x,_:x+[x[-1]+x[-2]],range(n-2),[0,1])
print(fib(5))

[night quarry]

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

[0, 1, 1, 2, 3]

[royal whale]

Not my creation, but significantly more esoteric

[sick hound]

I havent learned lamdar yet

[royal whale]

Ah. They are truly one of the worthiest of pursuits in Python. They allow confusion to ascend to a new dimension when properly (improperly?) handled.

[old socket]

To ascend lambda is to understand how to use ctypes to curse python

code: import ctypes
me:

[unreal echo]

!d curses

[night quarry]

curses

The curses module provides an interface to the curses library, the de-facto standard for portable advanced terminal handling.

While curses is most widely used in the Unix environment, versions are available for Windows, DOS, and possibly other systems as well. This extension module is designed to match the API of ncurses, an open-source curses library hosted on Linux and the BSD variants of Unix.

Note

Whenever the documentation mentions a character it can be specified as an integer, a one-character Unicode string or a one-byte byte string.

Whenever the documentation mentions a character string it can be specified as a Unicode string or a byte string.

[sick hound]

print ("dog")

[floral meteor]

it's just digging into lower level code to customise how python is interpreted

[floral meteor]

I've been looking into string encoding, it somehow makes the data longer

it's like there's bits randomly inserted

I tried reconstructing a string from its bytes, but about every three or four bytes, a bit seems to be inserted

[floral meteor]

Either that or I somehow reconstructed it wrong

[quartz wave]

what did you try to do

[void charm]

!e print("Lol")

[night quarry]

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

Lol

[earnest wing]

str encoding is adaptive, isn't it?

sometimes utf-32, sometimes better

[inland axle]

has anyone ever played with the pypy error messages and just made the dumbest possible fix until it converges to a valid program?

pypy3.8
Python 3.8.13 (7.3.9+dfsg-1, Apr 01 2022, 21:41:47)
[PyPy 7.3.9 with GCC 11.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>>> arr = [j for i in a, b for j in i]
  File "<stdin>", line 1
    arr = [j for i in a, b for j in i]
                       ^
SyntaxError: invalid syntax (expected ']')
>>>> arr = [j for i in a], b for j in i]
  File "<stdin>", line 1
    arr = [j for i in a], b for j in i]
                                      ^
SyntaxError: unmatched ']'
>>>> arr = [[j for i in a], b for j in i]
  File "<stdin>", line 1
    arr = [[j for i in a], b for j in i]
                             ^
SyntaxError: invalid syntax (expected ']')
>>>> arr = [[j for i in a], b ]for j in i]
  File "<stdin>", line 1
    arr = [[j for i in a], b ]for j in i]
                                        ^
SyntaxError: unmatched ']'
>>>> arr = [[[j for i in a], b ]for j in i]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
NameError: name 'i' is not defined
>>>> i = None
>>>> arr = [[[j for i in a], b ]for j in i]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: 'NoneType' object is not iterable
>>>> i = ()
>>>> arr = [[[j for i in a], b ]for j in i]

[floral meteor]

Move data from a str +48 to an int +24, then call bin on the int, then reconstruct the string from the 1s and 0s

Basically, the int gets possessed by a str so I can turn the string into binary, but confirming it worked shows weird results

The int containing data for a str of length 32 should contain length data of \x08, \0, \0, \0, \0, \0, \0, \0

Hence that's what's allocated

[restive void]

*UCS-4

[earnest wing]

[restive void]

(same thing, but officially...)

[quartz wave]

assuming the string is UCS-1 then that's correct

[floral meteor]

so it's efficienter in memory to not use strings

[quartz wave]

wait a string of length 32 shouldn't make an int with length data of 8

there's the limit that for each digit, the digit must be in range(2**30)

[floral meteor]

!e I found a better way to convert to binary anyway

from ctypes import*
s = "Hello, World!"
b = ''.join(map(lambda x:bin(x)[2:],b''.join((c_char*len(s)).from_address(id(s)+48))))
print(b)
print(eval('0b'+b))

[night quarry]

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

001 | 1001000110010111011001101100110111110110010000010101111101111111001011011001100100100001
002 | 176009217592051565591370017

[floral meteor]

wdym digit?

is this a 2**30 base system?

[quartz wave]

it's so easy to do that in a C extension ```c
PyBytes_FromStringAndSize(PyUnicode_DATA(s), PyUnicode_GET_LENGTH(s) * PyUnicode_KIND(s))

[quartz wave]

yes

[floral meteor]

whyyyyyy

[quartz wave]

2**30 is size 2 already btw

!e ```py
from ctypes import c_ssize_t
def g(x):return c_ssize_t.from_address(id(x)+16)
print(g(0), g(1), g(230), g(260))

[night quarry]

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

c_long(0) c_long(1) c_long(2) c_long(3)

[floral meteor]

!e ```py
from ctypes import*
el_numero = 8**32 - 1
print(c_long.from_address(id(el_numero)+16).value)
print((c_byte32).from_address(id(el_numero)+24))

[night quarry]

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

001 | 4
002 | -1 -1 -1 63 -1 -1 -1 63 -1 -1 -1 63 63 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0

[floral meteor]

wait im supposed to do 2**8**32 to get 32 bytes

!e ```py
from ctypes import*
el_numero = (2**8)32 - 1
print(c_longlong.from_address(id(el_numero)+16).value)
print((c_byte*32).from_address(id(el_numero)+24))

[night quarry]

@floral meteor :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | 1
002 | -1 31 0 0 116 127 0 0 1 0 0 0 0 0 0 0 -96 -31 -75 1 116 127 0 0 1 0 0 0 0 0 0 0

[floral meteor]

now why is it saying length data of one when it's huge?

and why is a byte in the range -128, 128 instead of unsigned?

[quartz wave]

it's 8191

8191 is so much less than 2**30

[floral meteor]

oh so it's just -1, 31, 0, 0

so this 2**30 system is why it acts weird when i try to port data between str and int objects?

[quartz wave]

probably

[floral meteor]

so what's this 16777216 number doing here?

what happens two the last two bits in these c_longs?

are they void for interpreter optimisation?

[quartz wave]

if c_long has a size of 4 bytes then that's encompassing ```py
[0, 0, 0, 0, 1, 0, 0, 0] # digit array as byte list
^ ^ ^ ^

[floral meteor]

i nede more cofe

so what happens when i make 4 bytes higher than 1073741823 in an integer memory?

does it just ignore the offending two bits on the left? does it ignore any two offending byte start bits?

[quartz wave]

what?

[floral meteor]

does it corrupt the data?

[quartz wave]

oh you mean filling it up

[floral meteor]

what bits are used to make the 2**30?

[quartz wave]

one digit it's totally fine

haven't tested with 2

[floral meteor]

because 4 bytes can fit 32 bits, but each 4 bytes here are 30 bit digits.

what two bits is it ignoring?

[quartz wave]

testing with 2 digits, it seems to ignore the most significant 2 bits (bit 30 and bit 31)

[floral meteor]

far left?

or furthest right c_long, but far left of that?

[quartz wave]

i have no idea what you just said

[floral meteor]

what is happening here?

[quartz wave]

idk

also isn't that just one digit

[floral meteor]

I have summoned an integer from an evil parallel dimension

I made a two digit number and made both digits 255

[quartz wave]

>>> a = 2**30-1
>>> c_uint.from_address(id(a)+24).value = -1
>>> a
4294967295
>>> a
4294967295
>>> a+0
4294967295
``` ???

[floral meteor]

but 4294967295+0 = -1

[quartz wave]

can't reproduce

even with exactly that code

[floral meteor]

because python math from evil parallel dimension says so

*evil cackling*

[quartz wave]

why can't i reproduce

[floral meteor]

!e ```py
from ctypes import*
el_numero = 2**30-1
(c_long*2).from_address(id(el_numero)+24)[:] = -1,-1
print(repr(el_numero))
print(repr(el_numero+0))

[quartz wave]

what kind of system are you in

oh wait

run in 3.10

[night quarry]

@floral meteor :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | 4294967295
002 | -1

[quartz wave]

so 3.10 has the problem

[floral meteor]

HA!

[quartz wave]

3.11 fixed it but how 🤔

[floral meteor]

in the 2**30 system it shouldn't work because all 32 bits are 1

[quartz wave]

!e 3.10 ```py
from ctypes import*
el_numero = 2**30-1
c_uint.from_address(id(el_numero)+24).value = -1
print(repr(el_numero))
print(repr(el_numero+0))

[night quarry]

@quartz wave :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | 4294967295
002 | -1

[quartz wave]

you only need to set one digit

[floral meteor]

and that's how i made a string corrupter function a week or two ago

using this bug

now the function of the corrupter has been explained to me, and now I know exactly 1 in 16 bits of information are lost in the corruption process

[quartz wave]

if python integer base is 2**15 then that's exactly correct

[floral meteor]

so unlike a cipher, there's no way to restore the information, as there's 2 possibilities every two characters as to what the original could be?

am I mathing this right?

or is the loss of information too systematic to be classified as corruption?

[quartz wave]

i mean you could always look at the digit array as a byte array and find out what they are
but when you do bin() they're just ignored so probably yeah

[floral meteor]

this is starting to sound like the topic should be called Malware Design 106: Data Corruption.

so how do I recover the corrupted data?

[quartz wave]

i mean you could always look at the digit array as a byte array and find out what they are

[floral meteor]

problem is that the way the corrupter does it, it makes a new object from the +0

which is why when I port the bytes back to the str instance, it becomes corrupted

[quartz wave]

(in 3.10)

[floral meteor]

!e ```py
s = """
Beautiful is better than ugly.
Explicit is better than implicit.
Simple is better than complex.
Complex is better than complicated.
Flat is better than nested.
Sparse is better than dense.
Readability counts.
Special cases aren't special enough to break the rules.
Although practicality beats purity.
Errors should never pass silently.
Unless explicitly silenced.
In the face of ambiguity, refuse the temptation to guess.
There should be one-- and preferably only one --obvious way to do it.
Although that way may not be obvious at first unless you're Dutch.
Now is better than never.
Although never is often better than right now.
If the implementation is hard to explain, it's a bad idea.
If the implementation is easy to explain, it may be a good idea.
Namespaces are one honking great idea -- let's do more of those!
"""
from ctypes import c_char

def corrupt(victim):
l = len(victim)
type_base = (c_char8).from_address(id(victim)+8)
odata = (c_charl).from_address(id(victim)+48)
type_base.value = bytes((c_char8).from_address(id(0)+8))
a = victim + 1
odata.value = bytes((c_charl).from_address(id(a)+48))
type_base.value = bytes((c_char*8).from_address(id('')+8))

corrupt(s)
print(s)

[night quarry]

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

001 | 
002 | Be!vti&vl )t b%ute2!th!o u'my.
003 | Exp,jci4!is bet4fr 4ian imp,jci4/
004 | S)npl%!is bet4fr 4ian com0mex.
005 | Co-qle8!is bet4fr 4ian com0mic!ued.
006 | Fl!u i3!be4uer tha.!ne3ued.
007 | Sp!sse is "ftt%s t(bn $fns%/
008 | R%bda"jli4z c/vnt3/
009 | S0fci!m c!tes are.(t 3qec)bl %oou'i t/!br%bk 4ie 2vle3/
010 | A,uho5hh 0sac4jca,jty bea4t p5sit9/
011 | E2sor3!sh/vld nev%s p!ts 3jle.uly.
... (truncated - too many lines)

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

[floral meteor]

nope

it's successfully corrupted in 3.11

[quartz wave]

so here's the problem with 3.10 https://github.com/python/cpython/blob/3.10/Objects/longobject.c#L3070-L3072

[night quarry]

Objects/longobject.c lines 3070 to 3072

if (Py_ABS(Py_SIZE(a)) <= 1 && Py_ABS(Py_SIZE(b)) <= 1) {
    return PyLong_FromLong(MEDIUM_VALUE(a) + MEDIUM_VALUE(b));
}```

[quartz wave]

here's the fix in 3.11 https://github.com/python/cpython/blob/main/Objects/longobject.c#L3271-L3273

[night quarry]

Objects/longobject.c lines 3271 to 3273

if (IS_MEDIUM_VALUE(a) && IS_MEDIUM_VALUE(b)) {
    return _PyLong_FromSTwoDigits(medium_value(a) + medium_value(b));
}```

[floral meteor]

so why does this look like the Zen of Python from a laptop in the Upside Down?

it should either look fine or look completely different, not like some echo of the original crying for help

[languid hare]

i like it

it's like the warning at the entrance of a cursed tomb

[floral meteor]

!e ```py
print('Yaa+pv \x08b'G%s v/xs 4p c5sse any"pdy who wou,e o0fn 4iis gra6f, 3p n/cod9!wi,m o0fn )u. 60 y%brs old.')

[night quarry]

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

Yaa+pv b'G%s v/xs 4p c5sse any"pdy who wou,e o0fn 4iis gra6f, 3p n/cod9!wi,m o0fn )u. 60 y%brs old.

[royal whale]

I must say the #esoteric-python channel has some very creative people

[long hamlet]

Fractal rendered under 2 sec with pure python

Code here https://github.com/TechnoTanuki/Python_BMP/blob/main/Hello_BarnsleyTree.py

[versed eagle]

esoteric how

[royal whale]

Please don't keep advertising your repo on here, we've seen it before. Also that's more of a #user-interfaces topic.

It's neat but off-topic

[long hamlet]

100% python grounds up and yes I wont

[long hamlet]

Not UI thou maybe media

[royal whale]

At any rate, certainly not #esoteric-python

[versed eagle]

being 100% python doesn't make it esoteric
it is cool, but it's not esoteric, which means that this isn't the correct channel

[rapid sparrow]

maybe it doesn't neatly fit into any of the channels? and i don't really feel like it's advertising

[long hamlet]

Have you written a pure python bmp parser and done math in the complex plane

[versed eagle]

allow me to rephrase then
it may be esoteric from a math standpoint, but that doesn't make it esoteric python

as such, this isn't the right channel for it

[quartz wave]

about 70% of this channel is normal code

not everything in #esoteric-python is esoteric

[versed eagle]

true but we can at least try to keep it organized

:P

[haughty sapphire]

class Myclass(type):
  pass

[quartz wave]

great now deobfuscate this ```py
n=512;print((lambda:0).class(compile([*[locals()['']for locals()['']in[[].len().class.class.base.subclasses()[2+2+({}!=[])]]][0].dict.values()]locals()[''].basicsize-"####".count('#'),'','eval'),locals())())

[versed eagle]

that's not valid python
locals()[""] for locals()[""] in

also locals()[""] doesn't exist

unless you're doing some sort of trickery

but what you're doing is creating a function object

using a bytes object

and _basicsize_ of something that I dont think exists

[quartz wave]

!e ```py
n=512;print((lambda:0).class(compile([*[locals()['']for locals()['']in[[].len().class.class.base.subclasses()[2+2+({}!=[])]]][0].dict.values()]locals()[''].basicsize-"####".count('#'),'','eval'),locals())())

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "<string>", line 1, in <module>
003 | KeyError: ''

[quartz wave]

ok

!e ```py
n=512;print((lambda:0).class(compile([*[globals()['']for globals()['']in[[].len().class.class.base.subclasses()[2+2+({}!=[])]]][0].dict.values()]globals()[''].basicsize-"####".count('#'),'','eval'),globals())())

[night quarry]

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

512

[quartz wave]

@versed eagle how about now

anything??

[versed eagle]

give me a moment I'm on mobile

[quartz wave]

oh i found the problem

[versed eagle]

how does globals()[""] work

[quartz wave]

python

[versed eagle]

it shouldn't

[quartz wave]

python

[versed eagle]

I'm confused

[quartz wave]

you should be

!e ```py
n=512;print((lambda:0).class(compile([*[s['']for (s:=[].class.class.base.subclasses()[108].create_module.globals['sys']._getframe(1).f_locals)['']in[[].len().class.class.base.subclasses()[2+2+({}!=[])]]][0].dict.values()]locals()[''].basicsize-"####".count('#'),'','eval'),globals())())

[versed eagle]

you can't give a var that name

[old socket]

Is this supposed to happen?

[night quarry]

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

001 |   File "<string>", line 1
002 | SyntaxError: comprehension inner loop cannot rebind assignment expression target 's'

[quartz wave]

you can

[versed eagle]

without assigning to globals directly

*afaik

[quartz wave]

well you can't normally

[versed eagle]

how

[quartz wave]

🤷

[versed eagle]

then how does your code work

[quartz wave]

it just does

!e ```py
n=512;print((lambda:0).class(compile([*[[].class.class.base.subclasses()[108].create_module.globals['sys']._getframe(1).f_locals['']for [].class.class.base.subclasses()[108].create_module.globals['sys']._getframe(1).f_locals['']in[[].len().class.class.base.subclasses()[2+2+({}!=[])]]][0].dict.values()]locals()[''].basicsize-"####".count('#'),'','eval'),globals())())

[night quarry]

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

512

[versed eagle]

I'm so confused

[old socket]

What's it even supposed to do

[versed eagle]

but, here's a somewhat deobfuscated version

!e

n=512;print((lambda:0).__class__(compile([*[globals()['']for globals()['']in[bytes]][0].__dict__.values()][globals()[''].__basicsize__-4](globals()[''],'6e20616e6427272e6a6f696e285b6368722834382b6e2f2f31302a2a6925313029666f72206920696e2072616e6765285f5f696d706f72745f5f28226d61746822292e6c6f673130286e292e5f5f7472756e635f5f28292b31295d2e5f5f72657665727365645f5f2829296f72273027'),'','eval'),globals())()) 

[night quarry]

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

512

[versed eagle]

is less obfuscated

slightly

[quartz wave]

ok but do you understand that

[versed eagle]

no

I understand the parts I removed

[quartz wave]

exactly

[versed eagle]

but I didn't touch the stuff I didn't understand

I didn't wanna break it

[quartz wave]

btw the printed 512 isn't a number

[versed eagle]

!e

n=512;print(type((lambda:0).__class__(compile([*[globals()['']for globals()['']in[bytes]][0].__dict__.values()][globals()[''].__basicsize__-4](globals()[''],'6e20616e6427272e6a6f696e285b6368722834382b6e2f2f31302a2a6925313029666f72206920696e2072616e6765285f5f696d706f72745f5f28226d61746822292e6c6f673130286e292e5f5f7472756e635f5f28292b31295d2e5f5f72657665727365645f5f2829296f72273027'),'','eval'),globals())()))

[night quarry]

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

<class 'str'>

[versed eagle]

huh

[quartz wave]

!e this is the absolute deobfuscation ```py
n=512;print(f"{n}")

[night quarry]

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

512

[versed eagle]

print(str(512))

im just gonna accept that you're a hyperdimensional magic being and try to not have a headache over this

[quartz wave]

could say that for most of the experts in #esoteric-python (including jack the mod)

[versed eagle]

True but very few people are active here

is you and death.py mostly

[quartz wave]

it may not work in <3.11

[versed eagle]

with hedge and fork also being there sometimes

[old socket]

My ipython is 3.10.4

Probably

[versed eagle]

and then I'm bumbling about just learning about fishhook like two days ago Lmao

my brain isn't ready for this stuff

[quartz wave]

it had more active participants some weeks earlier
including Ryuga, whitespace, chilaxan(?), Olivia, and JavaLim

[versed eagle]

oh yeah I saw lim

chilaxan is the one who made fishhook right?

[quartz wave]

yes

also very knowledgeable about python internals

[vague cairn]

When the weather is this hot, I'm brain dead by the time I get home from work, so during this season I mostly lurk or do slightly less mentally taxing things. I'm not gone. Just quieter than at other times.

I'm still thankful for the puzzlers that other people are putting out there.

[floral meteor]

I code crazier and more often when it's hot

It's colder at this time of year down in this hemisphere.

[earnest wing]

I read every message here

👻

[lyric fern]

!e

a=str|int|bool;b:a='512';print(int(b));[0];

...

[earnest wing]

Can you fetch all the defined operator strings (binary, unary, boolean, in place, ...) at runtime?

[plucky plover]

!d token

[night quarry]

token

Source code: Lib/token.py

This module provides constants which represent the numeric values of leaf nodes of the parse tree (terminal tokens). Refer to the file Grammar/Tokens in the Python distribution for the definitions of the names in the context of the language grammar. The specific numeric values which the names map to may change between Python versions.

The module also provides a mapping from numeric codes to names and some functions. The functions mirror definitions in the Python C header files.

[plucky plover]

this?

[earnest wing]

there's tokens that aren't operators

tbf at tokenization time python shouldn't expose operator specific info since there's ambiguity

e.g. * is not just an operator

I was looking in ast but couldn't find any public api

horrible idea: scrape the documentation of operator functions to fetch the op strings

[plucky plover]

or parse the sphinx inventory file

[earnest wing]

or fetching the operator classes from ast and unparsing them

[earnest wing]

!e ```py
import ast
import tokenize

tokens = []
for token in tokenize.EXACT_TOKEN_TYPES:
try:
x = ast.parse(f"a {token} b")
except SyntaxError:
continue
match x.body[0]:
case ast.Expr(ast.BinOp() | ast.BoolOp() | ast.UnaryOp() | ast.Compare()) | ast.AugAssign():
tokens.append(token)
print("lol", tokens)

[night quarry]

@earnest wing :white_check_mark: Your 3.11 eval job has completed with return code 0.

lol ['!=', '%', '%=', '&', '&=', '*', '**', '**=', '*=', '+', '+=', '-', '-=', '/', '//', '//=', '/=', '<', '<<', '<<=', '<=', '==', '>', '>=', '>>', '>>=', '@', '@=', '^', '^=', '|', '|=']

[earnest wing]

there's binary ones ig

except without the keyword operators

so not many at all tbh

[floral meteor]

I discovered new error

i used alt-f4 on a program using multiprocessing, now i can't rerun it lol

[royal whale]

What? There's not an @ or @= operator!

What is this sorcery?

[restive void]

sure there is

!e doesn't throw SyntaxError:

lambda: a @ b

[night quarry]

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

[No output]

[split salmon]

what

[restive void]

!pep 465

[night quarry]

**PEP 465 - A dedicated infix operator for matrix multiplication**

Link

Status

Final

Python-Version

3.5

Created

20-Feb-2014

Type

Standards Track

[royal whale]

what

[earnest wing]

I've been considering making a library for the following:

import functools
def xor(left, right):
    return (left or right) and not (left and right)

assert (==)(True `xor` True, False)

haskell-style conversions between infix and prefix functions

the implementation itself is fairly simple (just a token transformation + partial expression parser), but I have a problem:

I want to do this through a codec (# coding: operator-syntax or something), but it would need to be registered before your code is parsed

So is there a proper way for a library to register a codec before any code runs? Maybe using PYTHONPATH?

Maybe by monkeypatching the standard library somehow?

(I'm posting here because this is definitely something you shouldn't be doing )

[astral rover]

I mean lambda managed it

With import experessions

[earnest wing]

Oh good point

I'll see how that repo does it

Oh no is it .pth files

oh no

python packaging is such a hopeless mess, i love it

[versed eagle]

if you're doing in python, shouldn't you call left.__xor__(right)?

[earnest wing]

nope, that's different

clearer example:

def to(left, right):
    return right(left)

assert -1 `to` abs == 1

[versed eagle]

huh?

what

what are the backticks doing here

and to is a function, but you aren't calling it?

do you mean something like this?

def to(left, right):
    return right(left)

assert to(-1, abs) == 1

[earnest wing]

nope, I mean exactly what I wrote :>

that's the whole point of what I'm making :>

[versed eagle]

then why was it like this

im confused

[earnest wing]

.

[unreal echo]

!pypi lib-CST

!pypi libCST

[night quarry]

libcst v0.4.7

A concrete syntax tree with AST-like properties for Python 3.5, 3.6, 3.7, 3.8, 3.9, and 3.10 programs.

[quartz wave]

that reminds me of julia

[unreal echo]

what the heck

you ever wonder why god is in heaven?

because this channel is on earth

[unreal echo]

!E
print((lambda a, b: a @ b)(1, 2))

[night quarry]

@unreal echo :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 |   File "<string>", line 1, in <lambda>
004 | TypeError: unsupported operand type(s) for @: 'int' and 'int'

[unreal echo]

WHATTTT

[tidal canyon]

hello increasingly verbose Python

[quartz wave]

it only works with numpy arrays

[quartz wave]

hello

[tidal canyon]

ho

hi

[restive void]

!e

class F:
 def __matmul__(s, o):
  return 42

print(F() @ "hello") 

[night quarry]

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

42

[royal whale]

I'm not sure whether to call that brilliant or loathsome.

[tulip harbor]

Anyone familiar with how to debug lower level socket behavior? On MacOS 12.4 running python3.9.10, sockets hang for quite some time before completing. If I run the same code in either a curl format, OR on a Ubuntu machine, the timings are what I would expect.
Exs:

In [55]: %time requests.get('https://www.google.com/')
CPU times: user 14.9 ms, sys: 5.23 ms, total: 20.1 ms
Wall time: 1min 15s
Out[55]: <Response [200]>

Curl:

$ time curl -s "https://www.google.com/" > /dev/null
curl -s "https://www.google.com/" > /dev/null  0.01s user 0.01s system 4% cpu 0.504 total

[unreal echo]

does this count?

stairs = lambda c,t: map(lambda i:'\n'+i*' '+(t-i)*c,range(t,-1,-1))
print(*stairs('#',5))```

i don't think it does

it's pretty readabe

just golfed

[serene stratus]

Uhm this isn't really the channel for it

[tulip harbor]

But Esoteric: intended for or likely to be understood by only a small number of people with a specialized knowledge or interest.

[earnest wing]

Maybe you'll have some luck in #internals-and-peps or #networks

[bronze agate]

to golf it further you can ditch the variable assignment by calling the lambda in the expression like (lambda: None)(), you can also move your print inside the lambda

[versed eagle]

!e

# here's my version
# I call it the pyramid scheme
# not golfed but imo done in a fun way
@(lambda _:(lambda __,___:map((lambda _____:_____*" "+(___-_____)*__+"\n"),range(___,-1,-1))))
def stairs(c,d):...
print(*stairs("$",5))

[night quarry]

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

001 |      
002 |      $
003 |     $$
004 |    $$$
005 |   $$$$
006 |  $$$$$

[versed eagle]

!e

print(*(lambda c,t: map(lambda i:'\n'+i*' '+(t-i)*c,range(t,-1,-1)))('#',5)) 

[night quarry]

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

001 | 
002 |       
003 |     # 
004 |    ## 
005 |   ### 
006 |  #### 
007 | #####

[versed eagle]

golfed?

[restive void]

!e

print("\n".join(f"{'#'*n: >5}"for n in range(6)))

[night quarry]

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

001 |      
002 |     #
003 |    ##
004 |   ###
005 |  ####
006 | #####

[finite blaze]

'#'*n: >5

whats going on here?

nvm, I see now

!e

for n in range(6):print(f"{'#'*n:>5}")

[night quarry]

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

001 |      
002 |     #
003 |    ##
004 |   ###
005 |  ####
006 | #####

[quartz wave]

1. we don't discuss networking here
2. we're supposed to worsen your current condition

[floral meteor]

for _ in range(6):print(f"{'#':>{_}}")

[quartz wave]

??

[fleet bridge]

!e idk why

print(*map('\n'.__add__, map(lambda x: eval('x.__name__'), (type('#'*_,(),{}) for _ in range(10)))))

[night quarry]

@fleet bridge :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 | #########

[fleet bridge]

!timeit

import gc

gc.collect()
(*map('\n'.__add__, map(lambda x: eval('x.__name__'), (type('#'*_,(),{}) for _ in range(1000)))),)

[night quarry]

@fleet bridge :white_check_mark: Your 3.10 timeit job has completed with return code 0.

20 loops, best of 5: 18.9 msec per loop

[floral meteor]

!e ```py
for n in range(6):print(f"{' ':#>{n+1}}")

[night quarry]

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

001 |  
002 | # 
003 | ## 
004 | ### 
005 | #### 
006 | ##### 

[fleet bridge]

!e

print('''#
 #
  #
   #
    #''')

[night quarry]

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

001 | #
002 |  #
003 |   #
004 |    #
005 |     #

[fleet bridge]

the best

[floral meteor]

I right aligned lol I'm dum

[quartz wave]

!e```py
for n in range(6):print(f"{' ':#^{n+1}}")

ok

[night quarry]

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

001 |  
002 |  #
003 | # #
004 | # ##
005 | ## ##
006 | ## ###

[floral meteor]

!e ```py
for n in range(1,7):print(f"{' ':#<{n}}")

[night quarry]

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

001 |  
002 |  #
003 |  ##
004 |  ###
005 |  ####
006 |  #####

[floral meteor]

I'm typing on phone, don't judge

!e for n in range(1,7):print(f"{' '*(6-n):#<{n//2+5}}")

[night quarry]

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

001 |     #
002 |    ###
003 |   ####
004 |  ######
005 | #######
006 | ########

[floral meteor]

Eh idk

!e for n in range(6):print(' '(5-n),n'#')

[night quarry]

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

001 |       
002 |      #
003 |     ##
004 |    ###
005 |   ####
006 |  #####

[wheat apex]

!e 3.10

for i in range(5):
     print("#"*((i+1)**2))

[night quarry]

@wheat apex :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | #
002 | ####
003 | #########
004 | ################
005 | #########################

[wheat apex]

!e 3.10

for i in range(20):
     print("#"*((i+1)**2))

[night quarry]

@wheat apex :white_check_mark: Your 3.10 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/kajiduragi.txt?noredirect

[wheat apex]

wow

!e 3.10

for i in range(5):
     print((1 + round(0.25*i)) * "#")

[night quarry]

@wheat apex :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | #
002 | ##
003 | ###
004 | ####
005 | #####

[wheat apex]

!e 3.10

for i in range(5):
     print((i + 1 + round(0.49*i)) * "#")

[night quarry]

@wheat apex :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | #
002 | ##
003 | ####
004 | #####
005 | #######

[floral meteor]

!e hiding a variable inside a function. ```py
f=lambda n=[0]:n.setitem(0,n[0]+1)or n[0];print('\n'.join([(5-n)' '+'#'(2*n-1)for n in iter(f,6)]))

[night quarry]

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

001 |     #
002 |    ###
003 |   #####
004 |  #######
005 | #########

[floral meteor]

I feel like I made a ```py
def state_iterator(initial_state, halting_state, step_function):
return iter(lambda state=[initial_state]:state.setitem(0,step_function(*state))or state[0],halting_state)

which could make something as basic as ```py
def range(*args):
start, stop, step = args[0]if len(args)>1 else 0,args[0]if len(args)==1 else args[1],args[2]if len(args)>2 else 1
return state_iterator(start, stop, lambda x:x+step)

or any state machine.

!e iter can be used for a breaking condition when iterating ```py
print(*iter(range(16).iter().next,8))

[night quarry]

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

0 1 2 3 4 5 6 7

[floral meteor]

!e here's a possibly unnecessarily complicated fibonacci script to ruin your day. ```py
def state_iterator(initial_state, halting_state, step_function): return iter(
lambda state=[initial_state]:state.setitem(0,step_function(*state))or state[0],
halting_state
)

class HaltingState:
def init(self, f):self.f=f
def eq(self, k):return not not self.f(k)

fib = lambda maximum, init_state=[1,0]:state_iterator(
init_state[0], HaltingState(lambda k:k>=maximum),
lambda a,b = init_state[1:]: (b[0],b.setitem(0,a+b[0]))[0]
); *map(print,fib(69)),

[night quarry]

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

001 | 0
002 | 1
003 | 1
004 | 2
005 | 3
006 | 5
007 | 8
008 | 13
009 | 21
010 | 34
011 | 55

[floral meteor]

!e and here's a relatively nicer one. ```py
fib = lambda n:map(lambda _,a=[0,1]:(a[0],a.setitem(slice(0,2),(a[1],a[0]+a[1])))[0],range(n))
print(*fib(8))
print(*fib(16))

[night quarry]

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

001 | 0 1 1 2 3 5 8 13
002 | 0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610

[worldly trail]

!e

import ctypes,sys;_=list(zip(list([str(_) for _ in range(4)[1:]]), [_ for _ in range(4)[1:]]));sys.stdout.write(str(ctypes.cast(int(hex(id(_)),base=16),ctypes.py_object).value))

[night quarry]

@worldly trail :white_check_mark: Your 3.11 eval job has completed with return code 0.

[('1', 1), ('2', 2), ('3', 3)]

[worldly trail]

Is it possible longer? 🙂

[serene stratus]

Wdym longer

[lyric fern]

!e

print(True**False/True*True//True*False+1-True)

[night quarry]

@lyric fern :white_check_mark: Your 3.11 eval job has completed with return code 0.

0.0

[lyric fern]

!e

true = True 
false = False

a = True

if a == true:
    print(true)
else:
    print(false)

[night quarry]

@lyric fern :white_check_mark: Your 3.11 eval job has completed with return code 0.

True

[restive void]

!e

true, false, null = True, False, None
class json:
    loads = eval

print(json.loads('{"some": true, "thing": null, "24": [2,1,3]}'))

[night quarry]

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

{'some': True, 'thing': None, '24': [2, 1, 3]}

[maiden blaze]

tsk tsk

[pastel sparrow]

...that feels illegal

[sick hound]

Well of course that isn't completely the same

[last locust]

That's how you know it belongs here 😄

[sick hound]

{
    "some": true,
}

Is invalid

But it would be accepted as a python dict itself when evaluated like that

[restive void]

Yes, it's an overly permissive JSON parser.

[wheat river]

aint no one cares about json syntax

[earnest wing]

Is it actually superset of JSON parsers?

Like, assuming UTF-8?

[restive void]

Not quite, no. Different escape sequences.

[earnest wing]

right, \/ etc parse differently

[restive void]

That, and I think surrogate characters?

[earnest wing]

lone surrogates don't parse in python?

[restive void]

They do, but surrogates are just those codepoints, so you will never get the "real" character back.

> "\ud83c\udf09"
'🌉'

>>> "\ud83c\udf09"
'\ud83c\udf09'
>>> print("\ud83c\udf09")
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
UnicodeEncodeError: 'utf-8' codec can't encode characters in position 0-1: surrogates not allowed

!e vs.

import json
print(json.loads(r'"\ud83c\udf09"'))

[night quarry]

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

🌉

[sick hound]

Better way of doing this?

    if song.lower() in data['eras']['wod']:
        print("From wod")
    elif song.lower() in data['eras']['drfl']:
        print("from drfl")
    elif song.lower() in data['eras']['gbgr']:
        print("From gbgr")
    elif song.lower() in data['eras']['outsiders']:
        print("From outsiders")
    elif song.lower() in data['eras']['juicethekidd']:
        print("From juicethekidd")
    elif song.lower() in data['eras']['nineninenine']:
        print("From nineninenine")
    else:
        print("song not in database")

[quartz wave]

    sl = song.lower()
    for name, x in map(lambda y: (y, data['eras'][y]), ('wod', 'drfl', 'gbgr', 'outsiders', 'juicethekidd', 'nineninenine')):
        if sl in x:
            print(f"From {name}")
            break
    else:
        print("song not in database")

[sick hound]

I meant a better easier way not completely mind boggling 🤣

[quartz wave]

it's literally better than doing what you just did

[sick hound]

Also that returns all the songs in the era not the actual era

[restive void]

wrong channel, then

[sick hound]

What channel should I go to..

[restive void]

#❓｜how-to-get-help

[quartz wave]

ok wait

[quartz wave]

fixed

[sick hound]

What does map do?

[restive void]

How about:

!e

data = {
    "eras": {
        "wod": ["a wod song", "another wod song"],
        "gbgr": ["sounds grumpy"],
    }
}


for song in ("unknown song", "A WOD song", "sounds GrUmPy"):
    v = eval(' or '.join(f"song.lower() in data['eras']['{c}'] and '{c}'" for c in ("wod","gbgr")))
    print(v and f"From {v}" or "song not in database")

[night quarry]

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

001 | song not in database
002 | From wod
003 | From gbgr

[quartz wave]

apply the function throughout the iterable

map(a, (b, c, d)) is just a short way of doing [a(x) for x in (b, c, d)]

except it returns a generator

[burnt pasture]

so it's the same as (a(x) for x in (b, c, d))

[quartz wave]

yeah actually it's more like that

[fleet bridge]

map is a lot faster than generator-comprehension

because comprehensions are compiled to bytecode and map is implemented in C

also map better fits for this channel because you can do something funny like map(1.__radd__, a) instead of (x+1 for x in a)

[burnt pasture]

but map has to invoke a function, which has overhead. are you sure it's faster?

[fleet bridge]

comprehensions (and for-loops) can unpack values: for (k, v) in d.items()
map cannot

[burnt pasture]

perhaps it depends if the comprehension is also invoking a function

[fleet bridge]

i can test it in this example: map(1.__radd__, a) vs (x+1 for x in a)
give me a minute

[burnt pasture]

it will be hard to generalize "map is faster", it will depend a lot on what the map does.

also, those two things aren't equivalent

oops, they are

[fleet bridge]

>py -m timeit -s "a = [*range(1000)]" "list((x+1 for x in a))"
5000 loops, best of 5: 55.2 usec per loop

>py -m timeit -s "a = [*range(1000)]" "list(map(1 .__radd__, a))"
10000 loops, best of 5: 36.8 usec per loop

[burnt pasture]

cool

[fleet bridge]

if i want to do (f(x * 2 + 1) for x in a) with map, i must nest a lot of maps:
map(f, map(1 .__radd__, map(2 .__rmul__, a)))

in this case map is slower (i guess)

[burnt pasture]

if we weren't in #esoteric-python i would have a comment about that choice

[fleet bridge]

C:\Users\denba>py -m timeit -s "a = [*range(1000)]; f = lambda x: x * 0.5" "list(map(f, map(1 .__radd__, map(2 .__rmul__, a))))"
2000 loops, best of 5: 140 usec per loop

C:\Users\denba>py -m timeit -s "a = [*range(1000)]; f = lambda x: x * 0.5" "list((f(x*2 + 1) for x in a))"
2000 loops, best of 5: 148 usec per loop

hmm, almost no differense

also list((generator compr)) is slower than [list compr]

!timeit

a = [*range(10**4)]; f = lambda x: x - 2
``` ```py
list((f(x*2 + 1) for x in a))

[night quarry]

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

200 loops, best of 5: 1.49 msec per loop

[fleet bridge]

!timeit

a = [*range(10**4)]; f = lambda x: x - 2
``` ```py
[f(x*2 + 1) for x in a]

[night quarry]

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

200 loops, best of 5: 1.24 msec per loop

[fleet bridge]

!timeit

a = [*range(10**4)]; f = lambda x: x - 2
``` ```py
list(map(f, map(1 .__radd__, map(2 .__rmul__, a))))

[night quarry]

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

100 loops, best of 5: 2.3 msec per loop

[fleet bridge]

!timeit

a = [*range(10**4)]; f = lambda x: x - 2
``` ```py
[*map(f, map(1 .__radd__, map(2 .__rmul__, a)))]

[night quarry]

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

100 loops, best of 5: 2.29 msec per loop

[fleet bridge]

1. list-comp (1.24 msec)
2. list(gen-compr) (1.49 msec)
3. [*map] (2.29 msec)
4. list(map) (2.30 msec)

[burnt pasture]

i didn't realize #esoteric-python is also #microoptimizations 🙂

[fleet bridge]

sometimes i do this optimizations, its funny

nested maps are challenging to write and challenging to read, I enjoy it

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "list((str(x) for x in a))"
200 loops, best of 5: 1.11 msec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "list(map(str, a))"
500 loops, best of 5: 742 usec per loop

i think i get the rule:

• 1 map is faster than list-compr
• 2 maps are (almost) the same as list-compr
• 3 maps and more alre slower than list-compr

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "list(map(int, map(str, a)))"
200 loops, best of 5: 1.48 msec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "list((int(str(x)) for x in a))"
200 loops, best of 5: 1.88 msec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "[int(str(x)) for x in a]"
200 loops, best of 5: 1.68 msec per loop

important note: im using cpython 3.10

only creation of map and generator:```py
C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "(int(str(x)) for x in a)"
1000000 loops, best of 5: 208 nsec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]" "map(int, map(str, a))"
2000000 loops, best of 5: 163 nsec per loop

I think I've flooded this channel enough

[quartz wave]

probably because map() also acts like zip()

!e ```py
a = range(0, 20, 2)
b = range(10)
print(*map(lambda x, y: x + y, a, b))

[night quarry]

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

0 3 6 9 12 15 18 21 24 27

[fleet bridge]

TIL: map can accept more than one iterable

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]; b = [*range(0, 2*10**4, 2)]" "[a + b for a, b in zip(a, b)]"
500 loops, best of 5: 423 usec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]; b = [*range(0, 2*10**4, 2)]" "map(int.__add__, a, b)"
2000000 loops, best of 5: 118 nsec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]; b = [*range(0, 2*10**4, 2)]; add = __import__('operator').add" "map(add, a, b)"
5000000 loops, best of 5: 93.5 nsec per loop

hmm, i expect 3-rd example to be slower than 2-nd
why operator.add is faster than int.__add__?

ok, i forgot to iterate over map 😃

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]; b = [*range(0, 2*10**4, 2)]; add = __import__('operator').add" "[*map(add, a, b)]"
1000 loops, best of 5: 311 usec per loop

C:\Users\denba>py -m timeit -s "a = [*range(10**4)]; b = [*range(0, 2*10**4, 2)]" "[*map(int.__add__, a, b)]"
500 loops, best of 5: 524 usec per loop

operator.add(a, b) - 311 usec
a + b - 423 usec
int.__add__(a, b) - 524 usec

i expect int.__add__ to be faster than operator.add, and operator.add to be faster than a + b

[night quarry]

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

hello world

[versed eagle]

!e

(lambda ____:[_:=(lambda:[].__class__.__base__.__subclasses__().__getitem__([[{}]].__mul__([[[{}]]].__str__().__len__()).__str__().__len__().__sub__([[]].__len__().__bool__()).__xor__([[]].__str__().__len__()))())].__add__([__:=_().__str__().__mul__(_())]).__add__([___:=__import__([___:=([___:=_().__invert__().__neg__()].__add__([___:=___.__lshift__(___)]).__add__([___:=___.__lshift__(___)]).__add__([___:=___.__lshift__(___.__sub__(_().__invert__().__neg__().__invert__().__neg__()))]).__add__([___:=___.__or__(___.__floordiv__(_().__invert__().__neg__().__invert__().__neg__()))]).__add__([___:=___.__add__(___.__floordiv__([].__str__().__len__().__add__({}.__str__().__len__()).__add__(_().__invert__().__neg__().__add__(_().__invert__().__neg__()))))]).__add__([___:=___.__sub__([[]].__str__().__len__())]).__add__([__builtins__.__dir__().__str__().__getitem__(___.__add__(_().__invert__().__neg__())).__add__(__builtins__.__dir__().__str__().__getitem__(___)).__add__(__builtins__.__dir__().__str__().__getitem__(___.__add__(_().__invert__().__neg__())))])).__getitem__(_().__invert__())].__getitem__(_().__invert__()))]).__getitem__(_().__invert__()).__stdout__.__getattribute__(___.__stdout__.__dir__().__getitem__(_().__invert__().__neg__().__lshift__(_().__invert__().__neg__()).__lshift__(_().__invert__().__neg__().__lshift__(_().__invert__().__neg__())).__add__(_().__invert__())))(____))("hello world")

[night quarry]

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

hello world

[versed eagle]

i have done a hello world

[vague cairn]

!e ```py
x = {i:str(i+1) for i in range( 10)}

#not impossible.
print(list(map(repr, x.items())))

[night quarry]

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

["(0, '1')", "(1, '2')", "(2, '3')", "(3, '4')", "(4, '5')", "(5, '6')", "(6, '7')", "(7, '8')", "(8, '9')", "(9, '10')"]

[vague cairn]

Oh, I misunderstood what you were talking about.

Yeah, you're right.

[quartz wave]

!e ```py
(((:=(:=().class.base.subclasses().getitem(((____:=(:=().class.itemsize)).add()).mul(_______:=(:=.sub(:=((),).len().invert().neg()))).add(.add()))).dict).getitem((:=().class(builtins.dict).getitem)(.mul(.add())).add((:=loader.module.getitem(:=().len())).add(builtins.repr().getitem(builtins.dict.getitem((.pow().add(.add(:=.invert().neg()))))(_____,.add(_____))))))(:=__.getitem(().class(loader.dict).getitem(:=.invert().neg().invert())).wrapped(,(_____:=loader.doc.getitem)().add(()).add((:=(:=__.mul(:=.neg()))).add()).add((.sub().mul().add())).getattribute((:=().class(loader.name.class.dict).getitem)())(((:=.add()),)))),).mul((.sub())).add((.getattribute(name.getattribute((.add(.add(.add()))))()),))).getitem(.sub())()

[night quarry]

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

Hello world!

[quartz wave]

only parentheses, no constant strings, no constant integers, no constant booleans, only names with underscores, and no operators (except for :=)

[sour grove]

YOU LEGEND

[earnest wing]

and you don't need a walrus either if you just inline everything

and you don't need tuple literals either if you're using dunder methods

[quartz wave]

ok

[vital dirge]

Hello all. Ah, one of my favorite channels on this server. So entertaining, these uses and mis-uses of decorators and lambdas.

[vital dirge]

👏 👏 👏 🙂

[floral meteor]

!e ```py
from fishhook import hook
@hook(tuple)
def call(self, f, *g):
r = f(*self)
if g:return self.class.call((r,),*g)
else:return r

(6,9)(lambda x,y:10*x+y,lambda x:x,print)

[night quarry]

@floral meteor :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | <string>:8: SyntaxWarning: 'tuple' object is not callable; perhaps you missed a comma?
002 | 69

[floral meteor]

!e 3.10

from fishhook import hook
@hook(tuple)
def __call__(self, *args):
  for f in self:args = f(*args),
  try:arg,=args
  except:return args
  else:return arg

() (lambda x,y:x*10+y,lambda x:x+1,print) (6,8)

[night quarry]

@floral meteor :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | <string>:9: SyntaxWarning: 'tuple' object is not callable; perhaps you missed a comma?
002 | 69

[floral meteor]

callable tuples to replace
a = f(e(d(c(b(*args)))))
with
a = (b,c,d,e,f)(*args)

[versed eagle]

wait
,=?

since when is that a thing

[floral meteor]

since a long time ago

[versed eagle]

oh is it tuple unpack

[floral meteor]

made specifically for single-element tuples

[versed eagle]

I thought it was augm assign with comma

at first

[versed eagle]

does it work in 3.11?

[floral meteor]

!e ```py
(a, b), c = (1, 2), 3
print(a, b, c)
(a,) = (4,)
b,=5,
print(a,b)

[night quarry]

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

001 | 1 2 3
002 | 4 5

[floral meteor]

fishhook isn't installed in 3.11 in the bot

no reason why it shouldn't work in 3.11

[versed eagle]

yeah it's tuple unpack
sorry, I'm tired lmao

[versed eagle]

huh

[floral meteor]

is a user created package, not builtin

[versed eagle]

yeah i know

but still

[floral meteor]

packages are like DLCs in games. fishhook is like a modding system

[versed eagle]

yes
ik
but still, it's weird that it's installed on 1 but not the other

[floral meteor]

ctypes is like cheat codes

[versed eagle]

that's what I'm confused about, not about what fishhook is

[floral meteor]

idk

haven't installed it yet

[versed eagle]

what's the () for

() (lambda...

[floral meteor]

literally nothing

works without it, works with it

[versed eagle]

huh

[raw sparrow]

Print('hel')

Wot

print('hel')

[floral meteor]

what is this?

[raw sparrow]

Idk

Im just borred because i want h3elp

And im messing aorund because in 1- mins imma go out

10

[floral meteor]

!e try torturing python like ```py
import ctypes
ctypes.c_long.from_address(id(4)+24).value += 1
print( 2 + 2 )

[night quarry]

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

5

[raw sparrow]

!e

[night quarry]

Missing required argument

code

Command Help

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

*Run Python code and get the results.

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

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

By default your code is run on Python's 3.11 beta release, to assist with testing. If you run into issues related to this Python version, you can request the bot to use Python 3.10 by specifying the python_version arg and setting it to 3.10.

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

[raw sparrow]

!e

print('hello world')

OMG

oh right

I can't sned it

It's a very long message

[night quarry]

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

hello world

[floral meteor]

but also remember this isn't #bot-commands this is just for python torture

[edgy lava]

guyzz , my mandelbrot render

where is the channel for maths related stuff btw?

[floral meteor]

depends how esoteric it is.

I'd say math stuff is esoteric as many avoid mathematics.

[floral meteor]

!e 3.10

from ctypes import py_object
class Variable:
  def __init__(self, name):self.s=self.__name__=name;self.names={name}
  def __repr__(self):return self.s
  def __str__(self):return self.s
  def make_function(self):
    return eval(f'lambda {", ".join(self.names)}:{self!r}')
  def __call__(self, *a, **k):return self.make_function()(*a, **k)
  def __mul__(self, other):
    a, b = self.s!=self.__name__,type(other)==__class__
    self.s = f"{'('*a}{self!r}{')'*a}*{'('*b}{other!r}{')'*b}"
    if b:self.names|=other.names
    return self
  def __rmul__(self, other):
    a, b = self.s!=self.__name__,type(other)==__class__
    if b:other.names|=self.names;return self
    else:
      self.s=f"{other!r}*{'('*a}{self!r}{')'*a}"
      b and self.names.__ior__(other.names);return self
  def __add__(self, other):
    a, b = self.s!=self.__name__,type(other)==__class__
    if b:self.names|=other.names
    self.s = f"{'('*a}{self!r}{')'*a}+{'('*b}{other!r}{')'*b}"
    return self
  def __radd__(self, other):
    a, b = self.s!=self.__name__,type(other)==__class__
    if b:other.names|=self.names;return self
    else:
      self.s=f"{other!r}+{'('*a}{self!r}{')'*a}"
      b and self.names.__ior__(other.names);return self
  ...

class Dict(dict):
  def __missing__(self, key):
    import builtins
    try:return builtins.__dict__[key]
    except:return Variable(key)

py_object.from_address(id(globals())+8).value = Dict

y = (10*x+9)
print(f"y = {y!r}")
print(y(x=6))

[night quarry]

@floral meteor :x: Your 3.10 eval job has completed with return code 139 (SIGSEGV).

001 | y = (10*x)+9
002 | 69

[floral meteor]

I made algebra.

[edgy lava]

Nice

[floral meteor]

... denotes a fuckton of dunders I didn't bother implementing but pretend they exist.

someone here probably knows how to simplify this using tokens

as for brackets around an operation that already has precedence, that requires implementation of order of operations.

@edgy lava an fun project would be adding an implementation to Variable such that

  def __getattribute__(self, attr):
    getattr = object.__getattribute__
    if attr[0]=='d':
      d, varname = attr
      return __class__.integrate(varname)
    else:return getattr(self, attr)

...

y = 3*x + 4
Y = y.dx
print(Y) # (3*x**2)/2 + 4*x + c

and add a modification to globals for variables starting with d such that

y = 3*x + 4
y_ = dy/dx
print(y_) # 3

a package somewhere has gotta implement integration and deriving.

[unreal echo]

print(*map(lambda x: (x[0], *x[1]), zip((0 for i in [[1,2],[3,4],[5,6]]), [[1,2],[3,4],[5,6]])))``` tbh when i come up with a one liner for someones problem i just cross post it here

[fleet bridge]

!e

print(*map(lambda x: (x[0], *x[1]), zip((0 for i in [[1,2],[3,4],[5,6]]), [[1,2],[3,4],[5,6]])))

[night quarry]

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

(0, 1, 2) (0, 3, 4) (0, 5, 6)

[fleet bridge]

!e

print(*map((0,).__add__, map(tuple, [[1,2],[3,4],[5,6]])))

[night quarry]

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

(0, 1, 2) (0, 3, 4) (0, 5, 6)

[kindred needle]

!e

print(*[tuple([0]+list(range(7))[i:i+2]) for i in range(1, 7, 2)])

[night quarry]

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

(0, 1, 2) (0, 3, 4) (0, 5, 6)

[pastel sparrow]

!e

print(*[(0,*t)for t in[[1,2],[3,4],[5,6]]])

[night quarry]

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

(0, 1, 2) (0, 3, 4) (0, 5, 6)

[edgy lava]

second render of mandelbrot set

[dawn kayak]

!e ```py
import('ctypes').c_bool.from_address(id(True)+24).value = False
assert 1 == 1

[wide crow]

Doesn't this give the same effect without having to define every operation?

class Alg:
    def __init__(self, expr, var):
        self.expr = expr
        self.var = var
    
    def __call__(self, arg):
        new_expr = self.expr.replace(self.var, str(arg))
        return eval(new_expr)

f = Alg('3*x + 5 + x**2 + a', var='x')
a = 6
y = f(4)
print(y)  # 39

[night quarry]

@dawn kayak :warning: Your 3.10 eval job has completed with return code 0.

[No output]

[sullen kayak]

hey how do i handle function parameter conversion in a lambda function?

[restive void]

import sys
import inspect
from contextlib import contextmanager
from collections import defaultdict


deferred_lines = defaultdict(list)


def defer_trace(frame, event, arg):
    if event != "line":
        return defer_trace
    with open(inspect.getsourcefile(frame)) as f:
        lines = f.readlines()
    line = lines[frame.f_lineno - 1].strip()
    if line.startswith("defer; "):
        deferred_lines[frame].append(line.strip()[len("defer; "):])
        frame.f_lineno += 1
    else:
        pass
    return defer_trace


@contextmanager
def deferable():
    frame = sys._getframe().f_back.f_back
    sys.settrace(lambda *a:None)
    frame.f_trace = defer_trace
    try:
        yield
    finally:
        for line in deferred_lines[frame]:
            exec(line, frame.f_globals, frame.f_locals)
        deferred_lines.pop(frame)
        sys.settrace(None)


with deferable():
    print("Hello")
    defer; print("Bye")
    print("Middle")

Prints Hello\nMiddle\nBye\n

[floral meteor]

noice

[floral meteor]

defeats the entire purpose. anyone can interpret strings.

[edgy lava]

||print('hello')||

print('hello')

[dense nova]

put the lambda in brackets

[quartz wave]

you can't

[sullen kayak]

conflicting advice

[quartz wave]

1. there's no such thing as a "function parameter conversion" because python doesn't do anything about those
2. they're annotations and you can't put annotations in lambdas

[sullen kayak]

also i learned that indeed.

just annotations for the editor

that can help with type hints and such.

[quartz wave]

!e @sullen kayak ```py
def a(x: int, y: int) -> int:
return x + y

print(a("b", "c"))

[night quarry]

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

bc

[sullen kayak]

yeah, it's just a type hint.

both of them

-> int and x: int

[quartz wave]

ye

[sullen kayak]

i got that.

[quartz wave]

then also you can't put them in a lambda

[dense nova]

I agree with @quartz wave on this one, i misinterpreted your question. sorry

[sullen kayak]

i should probably ask my question here, i want this:

to do the same thing as: this

but i don't know how i'm gonna get just controls out of get_output

this has to be able to be imported in another file and ran just fine.

[quartz wave]

add and controls after the tuple

or do (...)[2]

[sullen kayak]

like so?

[quartz wave]

yea try that

[sullen kayak]

man if this works lmao

wayy a syntax error

[quartz wave]

what python version are you in

[sullen kayak]

in this case..

i think this venv is 3.7.9?

[quartz wave]

you need 3.8 for walrus operator

[sullen kayak]

damnit lmfao

i can't update it.

so uhh

what do i use instead 😬

setattr self target_location?

then access self.target_location later on?

[floral meteor]

if i can't use a walrus, i just use a list like a cell so that I can setitem inline

[floral meteor]

you know you can embed code here with
```py
stuff
```
right?

[sullen kayak]

yeah but i'm not a fan of the discord syntax highlighting.

LMAO

[floral meteor]

I'm not a fan of image compression

[sullen kayak]

yk if you just click on it

but ight

[floral meteor]

if i click on it, it's the same size

[sullen kayak]

anyway that function wasn't the problem fortunately

[floral meteor]

just moved to the middle

embed ur code.

not all of us are young and can read microscopic code

[sullen kayak]

oh my guh it works

LMFAOO

IT WORKS

i just added and None after the init tuple of statements.

[quartz wave]

btw you can do ```py
*map(setattr, (self.controls,)*4, ("steer", "boost", "throttle", "handbrake"), (steer_toward_target(...), True, 1.0, ...)),

[sullen kayak]

does it make my code harder to read?

[quartz wave]

idk

test it out

[sullen kayak]

is there any way i can drop the line count here at all?

to zoom in a little.

[floral meteor]

Even with microscope it still looks illegible

[sullen kayak]

Ok.

[floral meteor]

why are you making an empty class MyBot then inheriting from it? it seems kinda pointless

[sullen kayak]

hm

[floral meteor]

instead of pass you could at least add a docstring explaining why it exists

[sullen kayak]

adding a docstring would add lines

[floral meteor]

no it won't

class A:"""DOCSTRING"""

[sullen kayak]

automatically does this.

wait lmfao

semicolon fixed it

[floral meteor]

remember I said instead of pass?

not as well as pass

[sullen kayak]

there we go, a short and concise explanation.

[floral meteor]

pass is a beginner's gimmick, not anything meaningful

... is much cuter and does the same thing

it's like pass 2.0 it can even go inline

[sullen kayak]

so 3 lines is the lowest line count i'll be getting?

[floral meteor]

nope

[sullen kayak]

.-.

[floral meteor]

you can get rid of the second line

and thus be able to semicolon the remaining two lines together

[sullen kayak]

class MyBot(BaseAgent):

    def __init__(self, name, team, index):
        super().__init__(name, team, index)
        self.active_sequence: Sequence = None
        self.boost_pad_tracker = BoostPadTracker()

    def initialize_agent(self):
        # Set up information about the boost pads now that the game is active and the info is available
        self.boost_pad_tracker.initialize_boosts(self.get_field_info())

    def get_output(self, packet: GameTickPacket) -> SimpleControllerState:
        """
        This function will be called by the framework many times per second. This is where you can
        see the motion of the ball, etc. and return controls to drive your car.
        """

this is the class i'm trying to make.

[floral meteor]

embed the what you have instead of screenshotting it

[floral meteor]

embed this

[sullen kayak]

import math; from rlbot.agents.base_agent import BaseAgent, SimpleControllerState; from rlbot.utils.structures.game_data_struct import GameTickPacket; from util.boost_pad_tracker import BoostPadTracker; from util.orientation import Orientation, relative_location; from util.drive import steer_toward_target; from util.vec import Vec3
class MyBot():"""class for rlbot to run."""
Derived = type("Derived", (MyBot,BaseAgent), {"__init__": lambda self, name, team, index: (super(Derived, self).__init__(name, team, index), setattr(self, "boost_pad_tracker", BoostPadTracker()))and None, "initialize_agent": lambda self: self.boost_pad_tracker.initialize_boosts(self.get_field_info()), "get_output": lambda self, packet: (self.boost_pad_tracker.update_boost_status(packet), setattr(self, "target_location", min([[i.location, Vec3(packet.game_cars[self.index].physics.location).dist(i.location)] for i in self.boost_pad_tracker.boost_pads if i.is_active == True], key=lambda i: i[1])[0]), setattr(self, "controls", SimpleControllerState()),setattr(self.controls, "steer", steer_toward_target(packet.game_cars[self.index], self.target_location)), setattr(self.controls, "boost", True), setattr(self.controls, "throttle", 1.0),setattr(self.controls, "handbrake", abs(math.atan2(relative_location(Vec3(packet.game_cars[self.index].physics.location), Orientation(packet.game_cars[self.index].physics.rotation), self.target_location).y, relative_location(Vec3(packet.game_cars[self.index].physics.location), Orientation(packet.game_cars[self.index].physics.rotation), self.target_location).x)) > 1.5), )and self.controls})

[floral meteor]

noice

[sullen kayak]

Readable.

[quartz wave]

remove some more spaces

[floral meteor]

i'm already on it

import math;from rlbot.agents.base_agent import BaseAgent as A,SimpleControllerState;from rlbot.utils.structures.game_data_struct import GameTickPacket;from util.boost_pad_tracker import BoostPadTracker;from util.orientation import Orientation,relative_location;from util.drive import steer_toward_target;from util.vec import Vec3;B=type("B",(),{});Derived=type("Derived",(B,A),{"__init__":lambda s,name,team,index:(super(Derived,s).__init__(name,team,index),setattr(s,"boost_pad_tracker",BoostPadTracker()))and None,"initialize_agent":lambda s: s.boost_pad_tracker.initialize_boosts(s.get_field_info()),"get_output":lambda s,packet:(s.boost_pad_tracker.update_boost_status(packet),setattr(s,"target_location",min([[i.location,Vec3(packet.game_cars[s.index].physics.location).dist(i.location)]for i in s.boost_pad_tracker.boost_pads if i.is_active==True],key=lambda i:i[1])[0]),setattr(s,"controls",SimpleControllerState()),setattr(s.controls,"steer",steer_toward_target(packet.game_cars[s.index],s.target_location)),setattr(s.controls,"boost",True),setattr(s.controls,"throttle",1.0),setattr(s.controls,"handbrake",abs(math.atan2(relative_location(Vec3(packet.game_cars[s.index].physics.location),Orientation(packet.game_cars[s.index].physics.rotation),s.target_location).y,relative_location(Vec3(packet.game_cars[s.index].physics.location),Orientation(packet.game_cars[s.index].physics.rotation),s.target_location).x))>1.5),)and s.controls})

[quartz wave]

shorten some names

and strings

[floral meteor]

for that i'd need to use an actual IDE

[sullen kayak]

i didn't think that was possible

holy hell LMFAO

[floral meteor]

type('',(),{})

makes new class

[sullen kayak]

do you think i could like

Derived = type("Derived", (type("MyBot",(),{}),BaseAgent), {"__init__": la

pass the creation of the mybot class in the first argument of the type that makes derived?

[floral meteor]

yes

[sullen kayak]

W

[floral meteor]

wait why does MyBot even need to exist?

[sullen kayak]

i just need a class that inherits BaseAgent.

[floral meteor]

Derived does that

[sullen kayak]

mybot isn't necessary, come to think of it.

[floral meteor]

get rid of it lol

(BaseAgent,)

[sullen kayak]

yeah

[floral meteor]

don't forget the comma

import math;from rlbot.agents.base_agent import BaseAgent as A,SimpleControllerState;from rlbot.utils.structures.game_data_struct import GameTickPacket;from util.boost_pad_tracker import BoostPadTracker;from util.orientation import Orientation,relative_location;from util.drive import steer_toward_target;from util.vec import Vec3;B=type("B",(),{});D=type("D",(B,A),{"__init__":lambda s,n,t,i:(super(D,s).__init__(n,t,i),setattr(s,"b",BoostPadTracker()))and None,"initialize_agent":lambda s: s.b.initialize_boosts(s.get_field_info()),"get_output":lambda s,p:(s.b.update_boost_status(p),setattr(s,"target_location",min([[i.location,Vec3(p.game_cars[s.index].physics.location).dist(i.location)]for i in s.b.boost_pads if i.is_active],key=lambda i:i[1])[0]),setattr(s,"c",SimpleControllerState()),setattr(s.c,"steer",steer_toward_target(p.game_cars[s.index],s.target_location)),setattr(s.c,"boost",1),setattr(s.c,"throttle",1),setattr(s.c,"handbrake",abs(math.atan2(relative_location(Vec3(p.game_cars[s.index].physics.location),Orientation(p.game_cars[s.index].physics.rotation),s.target_location).y,relative_location(Vec3(p.game_cars[s.index].physics.location),Orientation(p.game_cars[s.index].physics.rotation),s.target_location).x))>1.5),)and s.c})

[sullen kayak]

yeah i put that there right after you said so

god this is ridiculous but you guys are amazing lmfao

what motivates you to make one of the easiest languages a pain in the backside to read?

[floral meteor]

if ...==True is the same as if ... the ==True is wasting time, keyboard life and pixels

same for the 0 in 1.0 you can just 1.

[floral meteor]

i refactored this a little, but i may have missed a few variables. If you get any NameErrors, just change the name to their first letter

[floral meteor]

bored, and it's an easy language so it's easy to abuse.

[quartz wave]

can just 1

[floral meteor]

idk what types the thingy takes

[quartz wave]

that's literally the same thing isn't it

[floral meteor]

sometimes these packagey things can be a bit fussy

"float can't be interpreted as int" or vice versa

pain in the hindquarters

[quartz wave]

😔

[sullen kayak]

it takes everything

i can set controls.throttle to fucking True if i wanted.

[floral meteor]

then why'd you use a bloody float?

[sullen kayak]

readability instinct

lmao

[floral meteor]

readability is for the weak

[quartz wave]

boost can be 1 then??

[quartz wave]

setattr(s.c,"boost",1)

[sullen kayak]

yes

it's a Truthy value.

[floral meteor]

I'm seeing a lot of setattrs that can be replaced with map, setattr and three arrays

[sullen kayak]

will it make it harder to read? 👑

[floral meteor]

obviously

XD

(setattr(a0,b0,c0),setattr(a1,b1,c1),setattr(a2,b2,c2),setattr(a3,b3,c3),setattr(a4,b4,c4))

or

(*map(setattr,[a0,a1,a2,a3,a4],[b0,b1,b2,b3,b4],[c0,c1,c2,c3,c4]),)

[sullen kayak]

[floral meteor]

and yes map can be used with more than two arguments

any time you use something you import exactly once, just use __import__ on the spot instead of importing at the start of the file

[sullen kayak]

?

[floral meteor]

wat

[sullen kayak]

LMFAO

[floral meteor]

where'd you get that from?

[sullen kayak]

idk

intellisense.

[floral meteor]

__import__('math')

!e ```py
print(import('math').pi)

[quartz wave]

setattr(s.c,"steer",steer_toward_target(p.game_cars[s.index],s.target_location)),setattr(s.c,"boost",1),setattr(s.c,"throttle",1),setattr(s.c,"handbrake",abs(math.atan2(relative_location(Vec3(p.game_cars[s.index].physics.location),Orientation(p.game_cars[s.index].physics.rotation),s.target_location).y,relative_location(Vec3(p.game_cars[s.index].physics.location),Orientation(p.game_cars[s.index].physics.rotation),s.target_location).x))>1.5)
``` to ```py
locals().update(x=p.game_cars[s.index]),locals().update(y=x.physics,z=s.target_location),*map(setattr,[s.c]*4,("steer","boost","throttle","handbrake"),(steer_toward_target(x,z),1,1,abs(math.atan2(relative_location(Vec3(y.location),Orientation(y.rotation),z).y,relative_location(Vec3(y.location),Orientation(y.rotation),z).x))>1.5))
``` ?

[night quarry]

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

3.141592653589793

[floral meteor]

ah no

[quartz wave]

also remove trailing comma

[floral meteor]

when has locals().update ever worked unless chilaxan has had at it?

[quartz wave]

how about globals().update

[floral meteor]

that works

[quartz wave]

ok do it

[floral meteor]

why use locals().update whatever when you cann

(lambda x,y:f(x,y))(1,2)

[quartz wave]

bruh ```py
("steer","boost","throttle","handbrake")
"steer boost throttle handbrake".split()

[floral meteor]

lmao

[quartz wave]

man if he was on 3.8 we could've used an assignment expression

[floral meteor]

if you want it shorter, interpret the whole file as utf-16, then embed it as a string into exec(b"...".decode("UTF-8"))

[sullen kayak]

i'm on 3.7.9

[floral meteor]

reduced character count but the bytes might be more

[sullen kayak]

and i can't upgrade it.

[floral meteor]

don't even need walrus

dont even need namespace

just

make a dictionary in globals called L

and use it as though it's locals

L.__setitem__('varname',value)

instead of varname := value

[quartz wave]

ye but then we'd have to do L["varname"]

[floral meteor]

¯_(ツ)_/¯

we could also define the L at the start of the lambdas if you want them to actually be local

remember also you can use mutable defaults in lambdas to hide variables

perhaps even just throw away names and index lists of objects

names are just a readability assistant

[sullen kayak]

code is just a readability assistant

program this entire shit into memory with instructions.

[floral meteor]

actually did that before

in python

made brainfuck interpreter that used native memory and mutated the code string into the output of the brainfuck

[floral meteor]

!e like this? ```py
from ctypes import*
class output:
def init(self, victim):
self.victim=id(victim);ol = len(victim)
(c_char*ol).from_address(id(victim)+48)[:]=[0]*ol
c_longlong.from_address(id(victim)+16).value=0
def lt(self, item):
_l = c_longlong.from_address(self.victim+16)
_l.value += 1
_a = c_byte.from_address(self.victim+48+_l.value)
_a.value=-item;return 0;
neg=lambda s:s
def gt(self, a):
a,i,_l=a,c_longlong.from_address(self.victim+16)
data = (c_byte_l.value).from_address(self.victim+49)
v, *data[:-1] = data;_l.value -= 1;a[i] = v;return 0;

class array:
def init(self, victim):
self.memory = "\0\0\0\0\0\0\0\0"[:]256
_ = self.memory = id(self.memory)
c_longlong.from_address(+16).value=2568-1
self.i=0; self.o = output(victim)
@property
def mem(self):return(c_byte*(256*8-1)).from_address(self._memory+48);
def inc(self):self.mem[self.i]+=1;return 0;
def dec(self):self.mem[self.i]-=1;return 0;
def left(self):self.i-=1;return 0;
def rite(self):self.i+=1;return 0;
def dot(self):return self.o <- self.mem[self.i];
def com(self):return(self.mem, self.i) <- self.o;
def loop(self):return+(not self.mem[self.i]);
def endl(self):return-bool(self.mem[self.i]);
def iter(self):return iter([self.mem])
def run(self, c):
p=t=0
try:
while 0<=p<len(c):
if t:t+=(c[p]=='[')-(c[p]==']')
else:t={
'+':self.inc,'-':self.dec,'<':self.left,'>':self.rite,
'.':self.dot,',':self.com,'[':self.loop,']':self.endl
}.get(c[p],lambda:0)()
p+=1-2(t<0)
except KeyboardInterrupt:pass
finally:return self.o

def brainfuck(victim, , _n=[0]):
clone,a = victim+'.',array(victim);a.run(clone)
_l=c_longlong.from_address(a.o.victim+16);l=_l.value;_l.value-=1
data = (c_bytel).from_address(a.o.victim+48)
_,*data[:-1]=data;import('sys')._getframe(1).f_globals[_n[0]]=a.o;_n[0]+=1

code = "++++++++++[>+++++++>+++++++++++>+++<<<-]++++++++++>++>--.>++.<.<.>-------.,..+++.>++++++++++++.,.+<<+++++++++++++++.>.+++.,.--------.>.<<<."
print(code)
brainfuck(code)
print(code)

[night quarry]

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

001 | ++++++++++[>+++++++>+++++++++++>+++<<<-]++++++++++>++>--.>++.<.<.>-------.,..+++.>++++++++++++.,.+<<+++++++++++++++.>.+++.,.--------.>.<<<.
002 | Hello, World!

[sullen kayak]

holy fucking shit

WHAT?

[floral meteor]

this sort of happened one day when i had too much red bull and ascended to a higher plane of existence

I'm like the Florida Man of programming

[quartz wave]

average #esoteric-python

[floral meteor]

random Turing Machine trivia https://www.quora.com/Is-DNA-a-Turing-machine

[restive void]

Hmm, I think your code is broken.. :P https://tinyurl.com/3f6mdpsj (sorry for the shortened link; too long for Discord)

[unreal echo]

if you were to write a key logged in ctypes, it would be so obfuscated i wouldn't even notice

[royal whale]

key logger?

[finite blaze]

Hey, I wonder if there is any way to escape string literals to access some variable?
I have a

variable = "text"
inputText = input()
print(f"{inputText}")

[quartz wave]

that's because TIO is running on python 3.7.4

[quartz wave]

it's so outdated that its latest is python 3.8 pre-release

[restive void]

(no, I was joking, it works fine; that was a quine)

[unreal echo]

what do you mean?

[vestal solstice]

it's just eval

[next flame]

or globals()[inputText] if you want to be slightly safer

[quartz wave]

locals()[inputText]

[floral meteor]

backwards compatibility was only tested as far back as 3.9

[floral meteor]

some people can read brainfuck though

although, not if I used intfuck, a brainfuck derivative that interprets one massive integer and substitutes digits for brainfuck characters

I could obfuscate it harder by calling oct on the string

then in the base ten system the integer isn't even readable

[high tree]

!e print(1)

[night quarry]

@high tree :white_check_mark: Your 3.10 eval job has completed with return code 0.

1

[floral meteor]

hmmmm

imma just make a portal to #bot-commands

[floral meteor]

https://paste.pythondiscord.com/nuviyegibi.txt?noredirect
nice quine XD

[floral meteor]

not in that case

[distant briar]

"fine", "quine", what a rhyme!

[royal patrol]

need help with generating a middle finger ascii art with the "esoteric" code in python, any help would be appreciated

ping me up if anyone interested

[floral meteor]

shortest way, or an obfuscated way?

[royal patrol]

obfuscated way would be better

but shortest way will do as well

[turbid glade]

lol

[royal patrol]

yes

this is the holy temple where people should learn python, this is the channel

[turbid glade]

what esoteric lang will u use

[royal patrol]

python in an esoteric way is the way to go

not like gonna write that shit in brainfuck

or wait

i can actually

but python would be cool

[turbid glade]

import pywhatkit

source_path = 'flower.png'
target_path = 'demo_ascii_art.text'
pywhatkit.ascii_art.image_to_ascii_art(source_path, target_path)

@royal patrol

done

[royal patrol]

-_--

not the esoteric way but thanks i can take it from here

[unreal echo]

cap 🧢

[quartz wave]

if i could learn what the characters mean i might be able to do that

[near briar]

Holy fking shit dude u are a genius

@floral meteor do u know any other pg languages?

[floral meteor]

I used to code in MATLAB

I also figured out Casio basic from playing around with a graphics calculator

[floral meteor]

You would also need to recognise patterns

Other than MATLAB and python, I tried java but coding in java feels like drowning in lemon juice after getting entire body papercutted

[languid hare]

there are small patterns in bf that you can understand locally, keeping track of a whole program gets messy though

eg, [-] is a common pattern that means "set the current cell to 0"

[floral meteor]

If they're bytes, then [+] also works

n+[>m+<o-]> means #0, #1 to 0, #1+(#0+n)*m/o

Sort of

You can do one of these with 12, 3, 1 to get to digits for decimal conversion

++++++++++[>+++++<-]> 0 50
++++. 6
+++.  9

[lyric fern]

!e

print(int(True))

[night quarry]

@lyric fern :white_check_mark: Your 3.11 eval job has completed with return code 0.

1

[floral meteor]

!e print(1==1.==True==False**False)

[night quarry]

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

True

[lyric fern]

!e
print(True/False)

[night quarry]

@lyric fern :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 | ZeroDivisionError: division by zero

[lyric fern]

!e

true = True

while true <= 10:
    true += 1
    print(true)

[night quarry]

@lyric fern :white_check_mark: Your 3.11 eval job has completed with return code 0.

001 | 2
002 | 3
003 | 4
004 | 5
005 | 6
006 | 7
007 | 8
008 | 9
009 | 10
010 | 11

[lyric fern]

!e print(True+6)

[night quarry]

@lyric fern :white_check_mark: Your 3.11 eval job has completed with return code 0.

7

[lyric fern]

!e print(1==True)

[night quarry]

@lyric fern :white_check_mark: Your 3.11 eval job has completed with return code 0.

True

[lyric fern]

Lmao

[languid hare]

!e

print({
  1: "one",
  2: "two"
}[True])

[night quarry]

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

one

[languid hare]

😉

[floral meteor]

!e 3.10

from fishhook import hook
@hook(bool)
def __str__(self):return"True or False"

print( True or False )

@hook(bool)
def __truediv__(self, other):
  if type(other)is type(self):
    return"True/False"

print( True/False )

[night quarry]

@floral meteor :white_check_mark: Your 3.10 eval job has completed with return code 0.

001 | True or False
002 | True/False

[floral meteor]

that's what chads do instead of match case

!e ```py
from random import randint
c=randint(1,9);{
c in(1,2,3):lambda:print("this"),
c in(4,5,6):lambda:print("that"),
c in(7,8,9):lambda:print("foo")
}1

[night quarry]

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

this

[floral meteor]

!e ```py
a = [1,5,7,2,8,3,9,0]
print('4 ','not '*(4 in a),'in a',sep='')

[night quarry]

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

4 not in a

[floral meteor]

I'd do anything to avoid if statements

[wheat river]

#ifBad

[royal whale]

What kind of occult sorcery is this? There are too many problems

Another match statement I guess

[floral meteor]

it's better than match

!e there's also the execute all matched cases version ```py
c = 4
[b()for a,b in(
((1,2,3),lambda:print("this")),
((4,5,6),lambda:print("that")),
((7,8,9),lambda:print("foo"))
)if c in a]

[night quarry]

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

that

[floral meteor]

!e or the reverse hierarchy ```py
c = 4
iter(b()for a,b in(
((1,2,3),lambda:print("this")),
((4,5,6),lambda:print("that")),
((7,8,9,4),lambda:print("foo"))
)if c in a).next()

[night quarry]

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

that

[floral meteor]

!e as opposed to ```py
from random import randint
c=randint(1,9);{
c in(1,2,3):lambda:print("this"),
c in(4,5,6,c):lambda:print("that"),
c in(7,8,9,c):lambda:print("foo")
}1 # foo

[night quarry]

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

foo

[floral meteor]

this whole channel is black magic

[finite blaze]

Hey, could you drop some impressive examples of code golfing?

I am trying to introduce my friend to this topic

[last locust]

This is an example of a golfed fizzbuzz program @finite blaze

[unreal echo]

this channel is essentially for brainfuck now

[sick hound]

!e
count = 0
while True:
print(int(True) * count)
count += 1

[night quarry]

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

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

Full output: too long to upload

[sick hound]

someone make this into a one-liner

!e

exec("""\ncount = 0\nwhile True:\n    print(int(True) * count)\n""")

[night quarry]

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

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

Full output: too long to upload

[sick hound]

!e
exec("""\ncount = 0\nwhile True:\n print(int(True) * count)\n count += 1\n""")

[night quarry]

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

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

Full output: too long to upload

[languid hare]

why not just print count :p

[sick hound]

because I dont wanna

you can prolly figure it out though right...

[sick hound]

make this into a one liner

[quartz wave]

!e ```py
[*map(print,import("itertools").count())]

[languid hare]

hey i wanted to do that

[sick hound]

without iter tools

[night quarry]

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

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

Full output: too long to upload

[quartz wave]

kk sure i'm just gonna pull out my dunder methods hold on

[sick hound]

Okaty

[languid hare]

!e

(count:=0,next(_ for _ in iter(int,1) if (print(count),count:=count+1,0)[-1]))

[night quarry]

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

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

Full output: too long to upload

[old socket]

!e ```py
while True:print(int(True)*(i:=globals().setdefault("i",0)+1))

[night quarry]

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

001 | 1
002 | 2
003 | 3
004 | 4
005 | 5
006 | 6
007 | 7
008 | 8
009 | 9
010 | 10
011 | 11
... (truncated - too many lines)

Full output: too long to upload

[languid hare]

kinda gross but it works

[old socket]

Oh mine starting at 1

😔

[languid hare]

you could just -1 it

[twilit grotto]

you don't need int in int(True)

[old socket]

It looks like they are asking specifically for it though

Probably not but

[sick hound]

Yeah

[old socket]

!e ```py
while True:print(int(True)*(i:=globals().setdefault("i",-1)+1))

[night quarry]

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

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

Full output: too long to upload

[sick hound]

Nice

[old socket]

You can also use iter but it'll probably be more characters

[quartz wave]

!e ```py
(:=name.class.class.base.subclasses().getitem((:=name.len().bool()).add().pow((:=.add(.add())).add()).add()))(builtins.getattribute((:=(builtins.dict).getitem)(.add().pow(.add().mul(.add())).add().mul(.add().mul(.add()))))(builtins.getattribute((.add().mul(.add()).add().mul(.add()).mul())),((:=(:=(:=(___:=(:=(:=name.class)).class.base.subclasses()).getitem((__:=.basicsize).add(.floordiv(:=(:=(:=.len(.new(_))).eq()).add(.add()))).add())).getattribute(,(:=(:=.getitem(.add().pow(.add()).add())))(.dict).getitem()).wrapped.globals).getitem(().getitem(.len().sub(.invert().neg().pow()).add()))((:=name.iter().class.name.getitem)(___:=.invert().neg().pow().add(.invert().neg())).add((.add(__))).add((.add(.add()))).add((:=.add())).add((.add(.add()))).add((.add()).mul(.add())).add(.class.doc.getitem(.add().pow(.add()).add().add(.add()))).add(_(.neg().invert())))).dict.getitem((.dict).getitem(.add().pow(.add().mul(.add())).add(_)))())))

[night quarry]

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

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

Full output: too long to upload

[quartz wave]

@sick hound i said i'd pull out my dunder methods

[sick hound]

bro what the fuck

[unreal echo]

the hell

[unreal echo]

what the fuck is a 143 SIGTERM?

[flint hollow]

if the bot notices that a python program takes too long, then the program gets sent an end request (SIGTERM), and if it refuses, gets ended involuntary (SIGKILL). SIGTERM is signal 15, and 143 is 15+128 (128=program ended)
tl;dr infinite loop got killed by snekbox

[earnest wing]

!e ```py
from signal import*
raise_signal(SIGSEGV)

[night quarry]

@earnest wing :warning: Your 3.11 eval job has completed with return code 0.

[No output]

[unreal echo]

!d signal

[night quarry]

signal

This module provides mechanisms to use signal handlers in Python.

Similar names: pyside2.signal

[unreal echo]

.bm What the actual hell

[quartz wave]

absolutely no constants
only names with underscores
everything has to be in dunder methods except for calling and :=

[verbal spear]

!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]

@verbal spear :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/nanafekodo.txt?noredirect

[proper vault]

A somewhat interesting challenge:
Write a function set_del(obj, fun) where obj is an object you are allowed to assign arbitrary attributes to, and fun is a function that should get called once the object obj gets collected by the garbage collector.

[quartz wave]

isn't that achievable by fishhook already

[proper vault]

This can be done in pure python

Though upon further consideration it is way easier than I thought

But it is possible without a custom __del__ method

And fishhook would make it a right pain to do this, since __del__ is on the class

[earnest wing]

seems simple with weakref.finalize

[earnest wing]

... hold on, isn't this function literally weakref.finalize

[proper vault]

It is possible tbh

I was honestly just trying to get people to figure out how try: finally: in generators leaks when the GC runs

[astral rover]

try: finally: leaks with generators?

[edgy lava]

bruh i jest went overlapping different graphs of mandelbrot sequence for various values of of Z(0) and now it looks like the graph is scared

[proper vault]

!e Leaks when the GC runs, doesn't leak memory.

def gen(s):
  try:
    yield 1
  finally:
    print(s)
any(q:=gen(2))
any(gen(3))

[night quarry]

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

001 | 3
002 | 2

[quartz wave]

why does any() stop collecting at the first yield?

[languid hare]

makes sense, no?

[quartz wave]

it stops at the first yield even when there are still multiple others

!e ```py
def x():
yield from [1, 0, 0]

print(any(x()))

[night quarry]

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

True

[languid hare]

that's how any works

did you confuse it with all

[quartz wave]

oh

yeah i did

[fleet bridge]

!e

def set_del(obj: 'Any', fun: 'Callable') -> None:
    cls = type(obj)
    new_type = type(cls.__name__, (cls,), dict(__del__=fun))
    obj.__class__ = new_type

import gc

class X:
    pass

x = X()
del x
gc.collect()

x = X()
set_del(x, print)
del x
gc.collect()

x = X()
set_del(x, lambda self: print(self))
del x
gc.collect()

x = X()
set_del(x, lambda self: globals().__setitem__('resurrected', self))
del x
gc.collect()

print(resurrected)

[night quarry]

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

001 | 
002 | <__main__.X object at 0x7f58a3a22f50>
003 | <__main__.X object at 0x7f58a3a230d0>

[fleet bridge]

it creates new subclass of object type with defined __del__ and changes object type

>>> set_del(object(), print)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 4, in set_del
TypeError: __class__ assignment only supported for mutable types or ModuleType subclasses

[earnest wing]

I think that was the challenge but still well done!

[proper vault]

Honestly, I was curious how many other solutions people would come up with

[finite blaze]

about challanges

I've got something for you

import dataclasses
import os
import random

FLAG = "getme"

@dataclasses.dataclass
class Message:
    message: str

    def __str__(self):
        return self.message

    __repr__ = __str__


MESSAGES = [
    Message("Thank you for using our service."),
    Message("Here is your pattern:"),
    Message("Until next time!")
]

pattern = input("pattern> ")
count = int(input("count> "))

final_pattern = pattern * count
print(f"{{message}} {final_pattern}".format(message=random.choice(MESSAGES)))

Try to get the flag

(its from a ctf from the last week)

[proper vault]

Ah, ||format injection||

[languid hare]

something like ||{message.__str__.__globals__}|| should work

[finite blaze]

Yep, that was our solution

Something like this

[restive void]

!e just because someone claimed in #python-discussion that keys of a dict need to be unique..

import random
class F:
    def __hash__(self):
        return random.randint(1, 5)
f = F()
d = {}
for i in range(100):
    d[f] = i
print(d)

[night quarry]

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

{<__main__.F object at 0x7f97c048b450>: 96, <__main__.F object at 0x7f97c048b450>: 77, <__main__.F object at 0x7f97c048b450>: 99, <__main__.F object at 0x7f97c048b450>: 95, <__main__.F object at 0x7f97c048b450>: 98}

[light grove]

@solar quail

[quartz wave]

seems like pretty normal behaviour

[unreal echo]

what did this achieve

i have lost a majority of my brain cells now

[tall meteor]

LOL

[lyric fern]

Lmao

[sick hound]

Oh

[restive void]

Yes, nothing special for this channel, but I couldn't just let that claim stand. Also, it's pretty funny what happens if you print d several times.

[royal sun]

https://www.hillelwayne.com/post/python-abc/

[languid hare]

😆

excellent

[royal whale]

Nice

How about this

!e

from abc import ABC

class TheTrueBaseObjectThatEverythingInheritsFrom(ABC): # ...but it inherits from ABC?
  @classmethod
  def __subclasshook__(*WeDontCareWhatYouPutHere):
    return True

print(isinstance(str, TheTrueBaseObjectThatEverythingInheritsFrom)) # yes
print(isinstance(int, TheTrueBaseObjectThatEverythingInheritsFrom)) # yes
print(isinstance(list, TheTrueBaseObjectThatEverythingInheritsFrom)) # yes
print(isinstance(ABC, TheTrueBaseObjectThatEverythingInheritsFrom)) # the superclass is an instance of its own subclass...
print(isinstance(object, TheTrueBaseObjectThatEverythingInheritsFrom)) # object's true base class revealed

[night quarry]

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

001 | True
002 | True
003 | True
004 | True
005 | True

[earnest wing]

abc subclasshooks are honestly kind of straightforward

glad they interact nicely with patma

[royal whale]

before that article I had no idea what either ABCs or __subclasshook__ did

Now I want to learn more 😈

[astral rover]

i wouldnt have thought you even needed the abc subclass here

oh you do?