esoteric-python

[cloud fossil]

The eleven characters programming language: Python

[lime bane]

Now that's a good slogan

[cloud fossil]

Actually thirteen if you count space and newline

Wait

!e

print(len("while1-=[]:\n "))

[night quarry]

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

13

[cloud fossil]

Yes

Thirteen

[lime bane]

Is lambda calculus also turning complete?

[frigid wharf]

according to https://en.wikipedia.org/wiki/Lambda_calculus#Explanation_and_applications it is

[lime bane]

Now for the ultimate obstructification:
You could use exec and input every character of the original code like this:

exec(chr(1+1+1+1+1+...)+chr(1+1+1+...)+...)

!e ```py
x = set(list("exec(chr(1+1))"))
print(x)
print(len(x))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

001 | {'c', 'e', 'r', ')', '+', '1', '(', 'x', 'h'}
002 | 9

[lime bane]

This is borderline cheating but still valid python code. This way you could actually write any characters and execute them.

And also this is less fun

!e py exec(chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1)+chr(1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

Hi

[fleet bridge]

yes
i can use only 1 to construct all integers
for quick scripting i used 0-9, but it can be constructed with 1s

[fleet bridge]

exchr()1+ - this is your set of chars, it is smaller than my

[lime bane]

But as I said, also more boring

It is astonishing how flexible this language is

[cloud fossil]

And it didn't mean to

[lime bane]

!e py exec(chr(111+1)+chr(111+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+11)+chr(111+1+1+1+1+1)+chr(11+11+11+1+1+1+1+1+1+1)+chr(11+11+11+1)+chr(11+11+11+11+11+11+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(111)+chr(11+11+1+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1+1)+chr(111)+chr(111+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1)+chr(11+11+11+1)+chr(11+11+11+1+1+1+1+1+1+1+1))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

Hello World

[lime bane]

I would say it is also quite fortunate that exec and chr together only have 5 different characters

!e ```py
def get_num2(n):
g_111 = n>=111
rest = n%111
whole_11s = rest//11
rest_rest = rest%11
p_111s = ['111'] * g_111
p_11s = ["11"]*whole_11s
p_1s = ["1"]*rest_rest
nums = []
nums += p_111s
nums += p_11s
nums += p_1s
#print(nums)
return "+".join(nums)

def obf(text):
numbers = [ord(x) for x in text]
chars = [f"chr({get_num2(n)})" for n in numbers]
txt = "+".join(chars)
result = f"exec({txt})"
return result

print(obf('print("Hello Discord")'))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

exec(chr(111+1)+chr(111+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+11)+chr(111+1+1+1+1+1)+chr(11+11+11+1+1+1+1+1+1+1)+chr(11+11+11+1)+chr(11+11+11+11+11+11+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(111)+chr(11+11+1+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1)+chr(111+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11)+chr(111)+chr(111+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1)+chr(11+11+11+1)+chr(11+11+11+1+1+1+1+1+1+1+1))

[lime bane]

I'd like to introduce you to exchr(+1):
Python, but worse

[earnest wing]

Have you heard of exc("%)?

[lime bane]

No

[earnest wing]

Hold on nevermind that doesn't necessarily work

Okay with =; it works

exchr(+1) and its sister language exc('%=;)

basically a bunch of assignments inside exec

[sudden osprey]

how do you get numbers (other than 0 or 1)? (or do you not need them?)

[earnest wing]

e=''=='%';ee=e=='';exec('eee=%c%%c%%%c'%ee%e%ee)```

`eee` is now 101

[sudden osprey]

aha

[earnest wing]

you can concat + modulo to get p much any number

then pass those to %c

then exec

I wonder actually

could you avoid having to use ;?

[sudden osprey]

I was thinking something like exec("%c"%((""=="")+(""=="")+(""=="")+...))

[earnest wing]

That works too

I forgot we had parens

so yeah you should be able to, right?

[lime bane]

Can you think of any way without =?

[sudden osprey]

was thinking the same thing 😄, i'm not sure

[earnest wing]

Use 1 and %?

To replace = and +

or is that not enough to reach most chars

[lime bane]

This still has 9 characters, because you brought the + back

exec('%c%c'%(111, 111))

exec('%c'%(111+1)+'%c'%(11+11))

[lime bane]

This has exc('%1,)

[earnest wing]

!e let me test:

s = set()
for i in range(1, 1000):
    x = int("1"*i)
    y = x % 1111
    if y < 256:
        s |={y}
print(len(s))

[night quarry]

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

4

[lime bane]

and this one has exc('%1+)

[earnest wing]

!e let me test:

s = set()
for i in range(1, 1000):
    for j in range(i, 1000):
        x = int("1"*i)
        y = int("1"*j)
        z = x % y
        if z < 256:
            s |={z}
print(len(s))

[night quarry]

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

4

[earnest wing]

too bad

I wonder...

[icy frigate]

How exactly does this work?

[lime bane]

Do you know what the exec function does?

[icy frigate]

Yeah

[lime bane]

Ok

[icy frigate]

But the hello world output

[lime bane]

!e py print(chr(111+1)+chr(111+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+11)+chr(111+1+1+1+1+1)+chr(11+11+11+1+1+1+1+1+1+1)+chr(11+11+11+1)+chr(11+11+11+11+11+11+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(111)+chr(11+11+1+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1+1)+chr(111)+chr(111+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+chr(11+11+11+11+11+11+11+11+11+1)+chr(11+11+11+1)+chr(11+11+11+1+1+1+1+1+1+1+1))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

print("Hello World")

[earnest wing]

!e let me test:

for q in range(1,10):
    s = set()
    for i in range(1, 1000):
        for j in range(i, 1000):
            x = int(str(q)*i)
            y = int(str(q)*j)
            z = x % y
            if z < 256:
                s |={z}
    print(q, len(s))

[night quarry]

@earnest wing :x: Your eval job timed out or ran out of memory.

001 | 1 4
002 | 2 4

[lime bane]

This is, what the text evaluates to, when you don't execute it and then I just execute it

!e ```py
exec('print("Hello World")')

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

Hello World

[lime bane]

So the thing in the parenthesis gets evaluated to code, that is then run

[lime bane]

Do you understand now?

@earnest wing Do you have a link to a list with all the %format characters? Whenever I google it, it just says use f-strings instead

[icy frigate]

where you can find what each alphabet is represented by?

[lime bane]

Do you know what ascii is?

[icy frigate]

not really but I've heard of it

[sudden osprey]

https://docs.python.org/3/library/string.html#format-specification-mini-language

[lime bane]

!e ```py
n = ord("H")
print(n)
print(chr(60))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

001 | 72
002 | <

[icy frigate]

what is ord?

!d ord

[night quarry]

ord

ord(c)```
Given a string representing one Unicode character, return an integer representing the Unicode code point of that character. For example, `ord('a')` returns the integer `97` and `ord('€')` (Euro sign) returns `8364`. This is the inverse of [`chr()`](https://docs.python.org/3/library/functions.html#chr "chr").

[icy frigate]

oh okay

[icy frigate]

so what's the relationship between the ones and this 72 in this case?

[sudden osprey]

nevermind I think it's a different thing for "%" formatting

I think it's this https://docs.python.org/3/library/stdtypes.html#old-string-formatting

[lime bane]

When I want to encode the letter "H", I need to do chr(72)

[icy frigate]

!e print(chr(72))

[night quarry]

@icy frigate :white_check_mark: Your eval job has completed with return code 0.

H

[lime bane]

But my goal is to use as few different characters as possible. That means I could do chr(1+1+1+1+1+....) until I get to 72

[icy frigate]

oh

[lime bane]

Or I use this algorithm

[icy frigate]

that seems very tedious checking what each adds up to

lol

oh nevermind

[lime bane]

That makes it shorter

[icy frigate]

how would you make it give what numbers to add up in a simpler method? That seems to be a bit complicated and can be shortnened

[lime bane]

!e ```py

def get_num(n):
return "+".join(["1"]*n)

print(get_num(72))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1+1

[lime bane]

This was the old, space inefficient way

[icy frigate]

oh that's way too long yeah

[lime bane]

But when I wanted to print Hello word the 2000 character limit of discord hit, so I had to reduce the size

[restive void]

You have e, so in some cases you can do stuff like 1e11, right? :D

[lime bane]

you mean 0x1e11

That's not a bad idea either

[lime bane]

Well actually, it is a bad idea, because then I need the 0 character

[icy frigate]

anyways, I wanted to ask what's a complicated looking one liner someone can do

i like trying to understand what it does even though I know I can't

[sudden osprey]

def conv(s):
    return 'exec('+ '+'.join('"%c"%('+ '+'.join(['(""=="")']*ord(l)) +')' for l in s) +')'

[lime bane]

exc('%1+) is probably the best character set

[sudden osprey]

a more primitive converter than @lime bane's one

!e

def conv(s):
    return 'exec('+ '+'.join('"%c"%('+ '+'.join(['(""=="")']*ord(l)) +')' for l in s) +')'

code = conv("print('Hello World!')")
print(exec(code))
print()
print(code)

[night quarry]

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

001 | Hello World!
002 | None
003 | 
004 | exec("%c"%((""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(""=="")+(
... (truncated - too long)

Full output: too long to upload

[sudden osprey]

oh too long to upload

lol
(ignore the print(exec )

[restive void]

no, I mean 1e11, as in scientific notation. Although chr() likely doesn't like floats..

Also... 1e11 is way too large, 1e1 doesn't help :D

[sudden osprey]

you might be on to something

it doesn't necessarily need to be 1, could use any digit

e.g.

!eval

print(3e33 % 33)

[night quarry]

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

12.0

[sudden osprey]

gives us a new number

[restive void]

!e

print(chr(12.0))

[night quarry]

@restive void :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 1, in <module>
003 | TypeError: 'float' object cannot be interpreted as an integer

[sudden osprey]

ah sad

[icy frigate]

what's 3e33 ?

[restive void]

3x10^33

[sudden osprey]

oh I just realised that's only because of floating point imprecision lol, so the digit used doesn't really matter, more just the size.

[lime bane]

!e ```py
def get_num(n):
whole_111s = n//111 # for ascii this is 1 or 0
rest = n%111
whole_11s = rest//11
rest_rest = rest%11
p_111s = ['111'] * whole_111s
p_11s = ["11"]*whole_11s
p_1s = ["1"]*rest_rest
nums = []
nums += p_111s
nums += p_11s
nums += p_1s
#print(nums)
return "+".join(nums)

def obf(text):
strings = []
for t in text:
if t in "exc(1+)":
strings += [f"'{t}'"]
elif t == "%":
strings += [f"'%'"]
else:
strings += [f"'%c'%({get_num(ord(t))})"]
string = "+".join(strings)
result = f"exec({string})"
return result

code = obf('print("Hello World")')
print(code)
exec(code)```

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

001 | exec('%c'%(111+1)+'%c'%(111+1+1+1)+'%c'%(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1)+'%c'%(11+11+11+11+11+11+11+11+11+11)+'%c'%(111+1+1+1+1+1)+'('+'%c'%(11+11+11+1)+'%c'%(11+11+11+11+11+11+1+1+1+1+1+1)+'e'+'%c'%(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+'%c'%(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+'%c'%(111)+'%c'%(11+11+1+1+1+1+1+1+1+1+1+1)+'%c'%(11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1+1)+'%c'%(111)+'%c'%(111+1+1+1)+'%c'%(11+11+11+11+11+11+11+11+11+1+1+1+1+1+1+1+1+1)+'%c'%(11+11+11+11+11+11+11+11+11+1)+'%c'%(11+11+11+1)+')')
002 | Hello World

[fleet bridge]

!e

assert __debug__

[night quarry]

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

[No output]

[balmy condor]

Anybody got any good ideas for text based games? Preferably arcade style games, so not a text based rpg

[fleet bridge]

text quest

[fleet bridge]

!e

def f():
    assert __debug__

def g():
    pass

import dis
dis.dis(f)
print()
dis.dis(g)

[night quarry]

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

001 |   2           0 LOAD_CONST               0 (None)
002 |               2 RETURN_VALUE
003 | 
004 |   5           0 LOAD_CONST               0 (None)
005 |               2 RETURN_VALUE

[cloud fossil]

Optimization

[balmy condor]

What is a text quest

Like Pokemon, turn based combat?

[sudden osprey]

!e

print(__debug__)

[night quarry]

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

True

[sick hound]

IlIIIII = __import__("string", globals(), locals(), ["Template"]).Template(b"\x70\x65\x61\x63\x65".decode("utf-8")); IIIIIlI = __import__("string", globals(), locals(), ["Template"]).Template(b"\x69\x6e".decode("utf-8")); IIIIIIl = __import__("string", globals(), locals(), ["Template"]).Template(b"\x6f\x75\x72".decode("utf-8")); IIlIlIII = __import__("string", globals(), locals(), ["Template"]).Template(b"\x74\x69\x6d\x65".decode("utf-8")); lllllII = " ".join([   IlIIIII.substitute(), IIIIIlI.substitute(),IIIIIIl.substitute(), IIlIlIII.substitute(),] ); __import__("sys").stdout.write(lllllII + "\n")

my first obfuscation attempt

!e ```py
IlIIIII = import("string", globals(), locals(), ["Template"]).Template(b"\x70\x65\x61\x63\x65".decode("utf-8")); IIIIIlI = import("string", globals(), locals(), ["Template"]).Template(b"\x69\x6e".decode("utf-8")); IIIIIIl = import("string", globals(), locals(), ["Template"]).Template(b"\x6f\x75\x72".decode("utf-8")); IIlIlIII = import("string", globals(), locals(), ["Template"]).Template(b"\x74\x69\x6d\x65".decode("utf-8")); lllllII = " ".join([ IlIIIII.substitute(), IIIIIlI.substitute(),IIIIIIl.substitute(), IIlIlIII.substitute(),] ); import("sys").stdout.write(lllllII + "\n")

[night quarry]

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

peace in our time

[quartz wave]

oh nice you used the Is and ls trick

[sudden osprey]

What's the Is and ls trick? Just a way of making quite indistinguishable variable names?

[quartz wave]

yes

you can add 1s for further effect

[floral meteor]

just a command line interface, or something more like a text adventure?

what you could do is write a novel, but write every possibility for every decision the protagonist makes, and make the user decide what the protagonist does.

If you don't want the scrolling terminal effect, just print(end="\x1b[H\x1bc") at the start of every frame.

a sequence better known as CSIHESCc

[gleaming hazel]

anyone got any experience with programming in qiskit? if u do ping me pls

[sick hound]

!e

b=lambda a: chr(int(('0'*(8-len(bin(ord(a))[2:]))+bin(ord(a))[2:])[4:]+('0'*(8-len(bin(ord(a))[2:]))+bin(ord(a))[2:])[::-1][4:][::-1],2));exec(''.join([b(z)for z in "\x07'\x96æG\x82r\x84VÆÆö\x12\x02ä\x966V\x02Gö\x02ÖVVG\x02\x97öWÂ\x02\x94ÅrÖ\x02ö&fW76\x16GVF\x12r\x92"])) 

[night quarry]

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

Hello! Nice to meet you, I'm obfuscated!

[sick hound]

https://tenor.com/view/floppa-big-floppa-chad-big-floppa-chad-floppa-chad-gif-23500776

[split salmon]

Lmao

[sick hound]

!e

exec(''.join([chr(round(int(a,2)/10240)) for a in '_-__-__@-@__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__-__-__-__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__@-@__@-@__-__-__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__-__@-@__@-@__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__-__@-@__@-@__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__-__@-@__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__-__@-@__@-@__@-@__@-@__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__-__-__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__-__-__-__-__-__@-@__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__@-@__-__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__@-@__-__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__-__@-@__-__@-@__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__-__@-@__-__-__@-@__@-@__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__-__@-@__-__@-@__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__-__-__-__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__@-@__@-@__@-@__@-@__-__-__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__-__-__-__-__@-@__-__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@-@__@~@__-__-__@-@__@-@__@'.replace('_-_', '1').replace('_@-@_', '0').split('_@~@_')]))```

[night quarry]

@sick hound :x: Your 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 <listcomp>
004 | ValueError: invalid literal for int() with base 2: '1100_@'

[floral meteor]

I think vids get filtered out by a bot in this server

plus even nitro has its upload limits and videos are chonky

[errant crescent]

?

!e

import __hello__, this

[night quarry]

@errant crescent :white_check_mark: Your eval job has completed with return code 0.

001 | Hello world!
002 | The Zen of Python, by Tim Peters
003 | 
004 | Beautiful is better than ugly.
005 | Explicit is better than implicit.
006 | Simple is better than complex.
007 | Complex is better than complicated.
008 | Flat is better than nested.
009 | Sparse is better than dense.
010 | Readability counts.
011 | Special cases aren't special enough to break the rules.
... (truncated - too many lines)

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

[errant crescent]

[turbid dragon]

what

whats __hello__ and this?

[errant crescent]

They're a part of python's builtin modules.

[turbid dragon]

why does it print that when you import them?

the zen of python

[errant crescent]

Thats what it does...

!e
import this

[night quarry]

@errant crescent :white_check_mark: Your eval job has completed with return code 0.

001 | The Zen of Python, by Tim Peters
002 | 
003 | Beautiful is better than ugly.
004 | Explicit is better than implicit.
005 | Simple is better than complex.
006 | Complex is better than complicated.
007 | Flat is better than nested.
008 | Sparse is better than dense.
009 | Readability counts.
010 | Special cases aren't special enough to break the rules.
011 | Although practicality beats purity.
... (truncated - too many lines)

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

[errant crescent]

!e

import __hello__

[night quarry]

@errant crescent :white_check_mark: Your eval job has completed with return code 0.

Hello world!

[turbid dragon]

i see

[errant crescent]

They're little easter eggs that were added by the devs iirc.

[turbid dragon]

makes sense

thanks

[floral meteor]

wrapping my brain around writing this felt like my brain was physically being wrapped around something

gameend=property(lambda self:not any([
      not all([self[a]for a in self]), # is there blank squares?
      any([a==b for(i,j,k,l),(w,x,y,z)in zip(self,filter(
        lambda n:n in range(4),map(sum,product(self,filter(lambda x:sum(map(abs,x))==1,product(range(-1,2),repeat=4))))
      )for a,b in map(int,[self[i,j,k,l],self[w,x,y,z]])) # are any adjacent cells equal?
    ]))

__setitem__=lambda self,item,value:(lambda i,j,k,l:self.matrix[l][k][j].__setitem__(i,[Cell,BaseCell][not value](*item,*[[value],()][not value])if type(value)==int else value))(*item)

[floral meteor]

a visual snippet of code from the 4D 2048 game

two more interface methods left and I'm done.
one is the output visualisation of the 4D board, the other is the event loop.

ok that's the event loop done ```py
def run(self):
self.eng and print('\x1b[H\x11b[c')
while not self.eng.gameend:c=self.input.kbhit()and self.input.getch().decode('ANSI')or'\1';self.next_frame({c in'wasd':lambda:self.eng.swipe(str.maketrans('wasd','\0\3\1\2')[ord(c)]),c=='\xe0':lambda:(lambda _c:self.eng.swipe(str.maketrans('KHMP','\7\4\6\5'))if chr(_c)in'KHMP'else self.quit())(ord(self.input.getch().decode('ANSI'))),c=='\x1b':self.quit}.get(1>0,lambda:'')())

makes async look so easy, doesn't it?

[finite jay]

this channel terrifies me

But I also think I've incurred Stockholm Syndrome for it

[floral meteor]

I think the frame display could be fit in one line if I soft code it enough. here is the untested complete soft coded board:

def next_frame(s,message=''):print(f"""\x1b[H\x1bc\x1b[0m{message}\x1b[90m^{"^=^".join(['^~^'.join(['^'.join(['@'.join(['|'.join([f"{self.eng[i,j,k,l]:{m}}"for i in range(4)])for k in range(4)])for m in map(str,range(3))])for j in range()])for l in range(4)])}\x1b[m\x1b[A""".replace('~','@'.join(['+'.join(4*['-'*7])]*4)).replace('=','@'*127).replace('@','\x1b[100m \x1b[49m').replace('^','\n'))

I left out the outermost borders to shave off a few characters of board size

[floral meteor]

and I made it under my goal of 44 lines

[icy frigate]

how long have you been programming using python? just curious

[sick hound]

!e

print(lambda: "esoteric python")()

[night quarry]

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

001 | <function <lambda> at 0x7fc17a7dfd90>
002 | Traceback (most recent call last):
003 |   File "<string>", line 1, in <module>
004 | TypeError: 'NoneType' object is not callable

[knotty delta]

wrap ur func in tuple then call it inside the print statement

!e
print((lambda: "esoteric python")())

[night quarry]

@knotty delta :white_check_mark: Your eval job has completed with return code 0.

esoteric python

[sick hound]

https://tenor.com/view/cat-cat-jumping-cat-excited-excited-dance-gif-19354605

[floral meteor]

yes.

alright, at 50 lines, the abomination is now at least playable.

you see the two 32s on the left?
you press s that would make them 64;
you then press the down arrow, and it combines with another 64 to make 128

s

then down arrow

[lusty sinew]

thats not a tuple

[knotty delta]

whups

[earnest wing]

What's that trick you can use with == to mutate the content of an immutable mappingproxy?

(And furthermore, is it cpython specific?)

[fleet bridge]

yeah, it is cpython bug

[fleet bridge]

https://bugs.python.org/issue43838

!e

from types import MappingProxyType

orig = {1: 2}
proxy = MappingProxyType(orig)

class X:
    def __eq__(self, other):
        other[1] = 3

assert proxy[1] == 2
proxy == X()
assert proxy[1] == 3
assert orig[1] == 3

[night quarry]

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

[No output]

[earnest wing]

Thank you

[floral meteor]

4 dimensional 2048 is coming along nicely

[rapid sparrow]

wow that's looking great

[steady lily]

great work! what IDE are you using?

and do you program 'readably' before you scramble it?

[wise bone]

how do people not apply decorators in source files when running tests? i'm currently patching google.cloud.firestore.transactional before importing my source files to no avail

[earnest wing]

Hi guys 🙂

I made UDLs in Python

(also supports decorator syntax)

(unregistering literals + context managers soontm)

[wise bone]

@earnest wing that's really cool!

[earnest wing]

Thank you! I'm making it a challenge to make this as "production-friendly" as possible within the silly premise (monkeypatching dynamic attributes to builtins lol) lol

[steady siren]

what's that?

[severe canyon]

i hate this with a passion

[languid hare]

is custom_literals your own module? nice job if it is

[earnest wing]

User-defined literals, so basically custom suffixes for literals
Based on the c++ model

[earnest wing]

yes, and I'm planning on pushing it to pypi soon lmao, thank you!

[languid hare]

[steady siren]

i dont understand a single thing you guys are talking about

google time

[earnest wing]

sometimes even i don't understand what i'm talking about

[steady siren]

mood

[steady siren]

explain to me like im 5 what on earth that is

[earnest wing]

in C++, you can use templates (basically macros) to define custom code that runs when you write e.g. 100_kg in code, by registering kg as a "literal suffix" for integers.

this is that system using pure Python, which wouldn't usually be possible (since you can't add attributes to builtin types) but thanks to some cpython implementation details is doable!

The syntax is a little different, of course ("blah".x vs "blah"_x), but otherwise it's the same idea

[steady siren]

i think all this is way beyond my skillset because i dont understand anything you're explaining 😂 ty for the explanation tho

i only know python and I'm still like kinda intermediate at it

[severe canyon]

what theme is this? i like the colors a lot :)

[earnest wing]

That's Ayu Mirage Bordered for vsc 😄

[fair quartz]

how can I

def foo(ref_x):
  ref_x += 1

a = 1
foo(a)
print(a)
```desired output```>>> 2```

where X is immutable

ping on reply pls

or dm :D

[full timber]

hi

[upbeat sonnet]

just add a = ref_x to the foo func

!e ```py
def foo(ref_x):
ref_x += 1
a = ref_x

a = 1
foo(a)
print(a)```

hmm

inchesting

!e
def foo(ref_x):
a = ref_x + 1

a = 1
foo(a)
print(a)

maybe if we make a var global?

!e ```py
def foo(ref_x):
global a
a = ref_x + 1

a = 1
foo(a)
print(a)```

[night quarry]

@upbeat sonnet :white_check_mark: Your eval job has completed with return code 0.

2

[upbeat sonnet]

ah

[upbeat sonnet]

@fair quartz could probably be cleaner but ye ^

wait

let me see

!e ```py
def foo(ref_x):
global a
a = ref_x + 1

a = 1
foo(a)
foo(a)
print(a)```

[night quarry]

@upbeat sonnet :white_check_mark: Your eval job has completed with return code 0.

3

[upbeat sonnet]

ahhhh

nice

[burnt pasture]

you can't do that.

[upbeat sonnet]

um.

you cant

?

[burnt pasture]

not the way you did it at least 🙂

[upbeat sonnet]

x is immutable hmm

[burnt pasture]

they want foo(a) to mean that a is incremented, and presumably that foo(b) would increment b.

[upbeat sonnet]

so thats not possible then

[burnt pasture]

no, because the function gets the value of the variable, not a reference to the variable.

[viscid nymph]

you can possibly just guess which one it is from the last frames locals, wouldn't work if multiple variables have the same value though

[burnt pasture]

yeah, ick

[viscid nymph]

or check the last frame's code

[burnt pasture]

i know it's kind of the point of this channel to do crazy things, so I guess go ahead.

[rugged sparrow]

I'm prob gonna write a proof of concept that'll examine the calling co_code to get a name and then loop up frames to find it and replace it

[errant crescent]

Looks awesome!

[errant crescent]

For more dynamic printing perhaps allow for the passing to take upon the length of the longest element + an offset, to prevent small lists being too far apart.

e.g. if an entire column contains only 2 elements per list, there would be no need to justify it with 20 spaces; more like 12 spaces.

Although, thats just a quality of life feature, and may be unnecessary for your current use.

[quartz wave]

!e ```py
from ctypes import *
def foo(ref_x):
res = ref_x + 1
if eval(f"ref_x is (_ := {ref_x})"):
print(f"warning: {ref_x} will be turned into {res}")
POINTER(c_uint).from_address(id(ref_x)+int.basicsize).contents = POINTER(c_uint).from_address(id(res)+int.basicsize).contents
c_ssize_t.from_address(id(ref_x)+int.basicsize - sizeof(c_void_p)).value = c_ssize_t.from_address(id(res)+int.basicsize - sizeof(c_void_p)).value

a = 1233
foo(a)
print(a)
b = 2
foo(b)
print(b)

[night quarry]

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

001 | 1234
002 | warning: 2 will be turned into 3
003 | 3

[quartz wave]

that's assigning by id

[floral meteor]

notepad++, and no.
If I need to debug a line, I "expand" it temporarily.

imma just casually play 2048 in 4 dimensions

is 16 spawns per swipe too much? I feel like it's a bit chaotic.

also aesthetics, should I make the vertical hyperborders thicker?

[floral meteor]

Actually I have to they're currently the same width as the subborders

Sort of

Cell separation width is the same

[fair quartz]

ofc i can't, i wanna tho

[fair quartz]

ok what ifi

do this:

!e

def foo(x):
    a = globals()[x.__name__]
    a += 1
    globals()[x.__name__] = a

a = 10
print(a)
foo(a)
print(a)

[night quarry]

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

001 | 10
002 | Traceback (most recent call last):
003 |   File "<string>", line 8, in <module>
004 |   File "<string>", line 2, in foo
005 | AttributeError: 'int' object has no attribute '__name__'. Did you mean: '__ne__'?

[fair quartz]

!e a = 4; print(*globals())

[night quarry]

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

__name__ __doc__ __package__ __loader__ __spec__ __annotations__ __builtins__ a

[fair quartz]

!e a = 5; print(*dir(a))

[night quarry]

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

__abs__ __add__ __and__ __bool__ __ceil__ __class__ __delattr__ __dir__ __divmod__ __doc__ __eq__ __float__ __floor__ __floordiv__ __format__ __ge__ __getattribute__ __getnewargs__ __gt__ __hash__ __index__ __init__ __init_subclass__ __int__ __invert__ __le__ __lshift__ __lt__ __mod__ __mul__ __ne__ __neg__ __new__ __or__ __pos__ __pow__ __radd__ __rand__ __rdivmod__ __reduce__ __reduce_ex__ __repr__ __rfloordiv__ __rlshift__ __rmod__ __rmul__ __ror__ __round__ __rpow__ __rrshift__ __rshift__ __rsub__ __rtruediv__ __rxor__ __setattr__ __sizeof__ __str__ __sub__ __subclasshook__ __truediv__ __trunc__ __xor__ as_integer_ratio bit_count bit_length conjugate denominator from_bytes imag numerator real to_bytes

[fair quartz]

ok then

[quartz wave]

@fair quartz

[fair quartz]

!e

def foo(x):
    xid = id(x)
    xname = None
    for glob in globals():
        if id(glob) == xid:
            a = globals()[glob]
            xname = glob

    a += 1
    globals()[xname] = a

a = 10
print(a)
foo(a)
print(a)

[night quarry]

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

001 | 10
002 | Traceback (most recent call last):
003 |   File "<string>", line 14, in <module>
004 |   File "<string>", line 9, in foo
005 | UnboundLocalError: local variable 'a' referenced before assignment

[fair quartz]

lemme look at that

[quartz wave]

warning: any number in range(-5, 257) will be overwritten with itself plus one

[fair quartz]

tooo we can do that?

[quartz wave]

so like i still have to make something that's better than what i just did

[fair quartz]

yeah could you tell me how to get a var with ctypes if we know the id?

@quartz wave sorry for the ping

so do we just do

ctypes.POINTER(ctypes.c_uint).from_address() ?

and im guessing .contents is a property which supports getting and setting (and delling) but im wondering why need all this whacky stuff to change ints? can't we reassign it to the value of res?

[quartz wave]

to the last question in this message, no

def foo(ref_x):
    res = ref_x + 1 # the result
    if eval(f"ref_x is (_ := {ref_x})"): # if the integer is cached...
        print(f"warning: {ref_x} will be turned into {res}") # warning, there's no turning back from converting a cached integer
    pointer(POINTER(c_uint).from_address(id(ref_x)+int.__basicsize__)).contents = pointer(POINTER(c_uint).from_address(id(res)+int.__basicsize__)).contents
    # ^ pointer to the digits of ref_x                                              ^ pointer to the digits of res
    # so it basically changes the digits array of ref_x, and may leak the old one
    c_ssize_t.from_address(id(ref_x)+int.__basicsize__ - sizeof(c_void_p)).value = c_ssize_t.from_address(id(res)+int.__basicsize__ - sizeof(c_void_p)).value
    # ^ size of ref_x                                                                ^ size of res
    # so then it changes the internal size of ref_x

[fair quartz]

ouu

damn this is beyond low

level

python

can i do it with globals()?

[quartz wave]

no

it's just impossible to know what variable has the exact value as ref_x

because there are cached integers

[quartz wave]

in this range

so is won't work

[fair quartz]

ahh

i see

[quartz wave]

equivalent code in C: ```c
#define PY_SSIZE_T_CLEAN
#include "Python.h"

int foo(PyLongObject *ref_x) {
PyLongObject *res;
PyObject *string;
Py_ssize_t ref_x_size;
digit *new_ob_digit;
int status;

res = PyNumber_InPlaceAdd(ref_x, _PyLong_GetOne());
if (res == NULL) {
    return -1;
}
ref_x_size = Py_SIZE(ref_x);
if (((size_t)ref_x_size) + 1U < 3U) {
    if (ref_x_size) {
        digit first_digit = ref_x->ob_digit[0] * ref_x_size;
        if (first_digit + _PY_N_SMALLNEGINTS < 0 || first_digit >= _PY_N_SMALLPOSINTS) {
            goto change;
        }
    }
    string = PyUnicode_FromFormat("warning: %u will be turned into %U", first_digit, res);
    if (string == NULL) {
        goto error;
    }
    status = PyObject_Print(string, stdout, Py_PRINT_RAW);
    Py_DECREF(string);
    if (status < 0) {
        goto error;
    }
}

change:
// it's just hard to not have a memory leak here
ref_x->ob_digit = res->ob_digit;
Py_SET_SIZE(ref_x, Py_SIZE(res));
return 0;
error:
Py_DECREF(res);
return -1;
}

[floral meteor]

That's just the command line interface level.
It's around the same width as the "graphical" user interface. I put graphical in quotation marks because it does not use a gui

[floral meteor]

this is in the terminal

only ascii characters are printed.

and that does the entire display.

that's how games were coded in the early days of computers, but they only used 8 colours

how the code looks in vscode. the document outline on the right is rather pointless.

[floral meteor]

!e ```py
@type.call
class foo:
def getattribute(self, name):
if type(globals()[name])is int:
globals()[name]+=1;return globals()[name]
else:return super().getattribute(name)

a = 2 + 2
foo. a
print(a)

[night quarry]

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

5

[fair quartz]

foo . a?

😳

[floral meteor]

I coulda named it dystopia

a = 2 + 2
dystopia. a
print("2 + 2 =", a)

[fair quartz]

isnt that same as foo.a

[floral meteor]

yeah, but the space puts emphasis on its relation to the global name a

[fair quartz]

ohh

getattribute

[floral meteor]

esoteric code is an artform

[fair quartz]

of class

it really is

lol

[floral meteor]

@type.__call__

don't forget that

[jovial monolith]

upload to a github repo xD

!e

class PythonPrivateClass:
  def __getattribute__(self, name):
    raise ValueError("this class is private")
priv_class = PythonPrivateClass()
priv_class.a```

[night quarry]

@jovial monolith :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 5, in <module>
003 |   File "<string>", line 3, in __getattribute__
004 | ValueError: this class is private

[jovial monolith]

!e

class PythonPrivateClass:
  def __getattribute__(self, name):
    raise ValueError("this class is private")
priv_class = PythonPrivateClass()
priv_class.super().a```

[night quarry]

@jovial monolith :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 5, in <module>
003 |   File "<string>", line 3, in __getattribute__
004 | ValueError: this class is private

[fair quartz]

!e

@type.__call__
class Ugh:
  def __getattribute__(self, name):
    globals()[name] = "ughh"

Ugh.a
print(a)

[night quarry]

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

ughh

[fair quartz]

lol

[fleet bridge]

use Cells

>>> from types import CellType as C
>>> def f(ref):
...   ref.cell_contents += 1
...
>>> a = 1
>>> c = C(a)
>>> f(c)
>>> a = c.cell_contents
>>> a
2

!e

from types import CellType as C
def f(ref):
    ref.cell_contents += 1

a = 1
c = C(a)
f(c)
a = c.cell_contents
print(a)

[night quarry]

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

2

[fleet bridge]

also you can use one-element list

!e

def f(ref):
    ref[0] += 1
    return ref

a = 1
ref = [a]
f(ref)
a = ref[0]
print(a)

[night quarry]

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

2

[left thicket]

jump_hurdle()

[wise bone]

what do i do if my class method is wrapped by another class and thus is called with the wrong self?

[floral meteor]

use the letter s as the secondary self

Or use an underscore as a prefix

[vale tangle]

!e

import pickle as rick
# lovin this joke haha

[night quarry]

@vale tangle :warning: Your eval job has completed with return code 0.

[No output]

[turbid dragon]

very esoteric

[severe canyon]

!e

print(*(''.join(chr(34+len((z:=0j)or[(z:=z*z+x/64*2.67-2.1+y/32*2.5j-1.25j)for _ in[0]*61if abs(z)<=2]))for x in range(64))+'\n'for y in range(32)),sep='')

[night quarry]

@severe canyon :white_check_mark: Your eval job has completed with return code 0.

Code block escape attempt detected; will not output result

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

[severe canyon]

ah man :(

!e

print(*(''.join(chr(34+len((z:=0j)or[(z:=z*z+x/64*2.67-2.1+y/32*2.5j-1.25j)for _ in[0]*61if abs(z)<=2]))for x in range(64))+'\n'for y in range(32)),sep='')

[night quarry]

@severe canyon :white_check_mark: Your eval job has completed with return code 0.

001 | #############$$$$$$$$$$$$$$%%%%%%%%%%%%%%%%%%%%%%%%$$$$$$$$$$$$$
002 | ############$$$$$$$$$$%%%%%%%%%%%%%%%%%%%%%%&&&&&%%%%%%%$$$$$$$$
003 | ###########$$$$$$$%%%%%%%%%%%%%%%%%%%%&&&&&&')-)''&&&&%%%%%$$$$$
004 | ##########$$$$$$%%%%%%%%%%%%%%%%%%%&&&&&&&'''(-8*)-'&&&&&%%%%%$$
005 | ########$$$$$%%%%%%%%%%%%%%%%%%%&&&&&&&&&'''()*-3,)(''&&&&&%%%%%
006 | ########$$$%%%%%%%%%%%%%%%%%%%&&&&&&&&&'''().04T_0-;('''&&&&%%%%
007 | #######$$$%%%%%%%%%%%%%%%%%%&&&&&&&&''(((()*5A____L,)((''''&&&%%
008 | ######$$%%%%%%%%%%%%%%%%%%&&&&&&&''()*+***+,.E____2,+*)((()*'&&%
009 | #####$$%%%%%%%%%%%%%%%%&&&&&&''''(()4LPF.__N________<_7,,2,]+'&&
010 | #####$%%%%%%%%%%%%%%%&&&'''''''((()*,:_____________________.*('&
011 | ####$%%%%%%%%%%%%&&&'''''''''((()*453_____________________0+)('&
... (truncated - too many lines)

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

[severe canyon]

made my mandlebrot golf into ascii output ¯_(ツ)_/¯

[floral meteor]

that is beautiful!

can I has an ANSI Sequence version?

yesterday I was discussing in #python-discussion why mathematics is essential to programming.

[floral meteor]

you said golf?

r=range;print(*(''.join(chr(34+len((z:=0j)or[(z:=z*z+x/64*2.67-2.1+y/32*2.5j-1.25j)for _ in[0]*61if abs(z)<=2]))for x in r(64))+'\n'for y in r(32)),sep='')

actually it might be the same size

range vs r=;rr

5170 characters, 45 lines, and it is harder to lose than win;
4-dimensional 2048 is now complete and fully debugged.

that took me 4 days.

[severe canyon]

The previous file output version was golfed

This is just a conversion

No golfing there

[floral meteor]

yeah i realised that after XD

I'm getting rusty

[severe canyon]

Replace 34 with 31, and 61 with 64, then it looks better

But python bot wouldn't let me :(

[floral meteor]

why not?

[severe canyon]

The escaping thingy

Trying to break out of code block

[floral meteor]

lmao

[severe canyon]

.

[floral meteor]

it must've outputted `

[severe canyon]

Indeed it did

[floral meteor]

optemisashun

a=[a()for _ in range(16)];{[a[n+_].move(*r())for n in range(16-_)]and()or()for _ in iter(lambda:len({*a}),16)}

a is the list of coordinates for the 16 cells the game starts on.

[floral meteor]

I just realised I could use random.choice instead

[floral meteor]

I put a comment at the eof to make it a nice, even 5000 characters

[floral meteor]

It took soooooo long, but I finally played far enough to test a feature: numbers more than 7 digits long

[vague cairn]

the android version I've been playing stops when you get to 7 digits. :(

[floral meteor]

I'm smarter than the android devs

Mine goes up to 21 digits

after that, it will keep going, but the board will be distorted

[vague cairn]

Right, these people could just lower the font size, but no?

So clarify, yours is 4d?

[floral meteor]

yes

[vague cairn]

Nice

[floral meteor]

there is up, down, left, right, in, out, forwards and backwards

if you play with just wasd, it's like playing 16 games of 2d 2048 at once

[vague cairn]

Also, terminal based, which is based in its own dimension.

[floral meteor]

yes

it is in the terminal, and only ascii characters are printed

[bronze merlin]

how do we split between each statesment?

i'm not sure 🤔

@split salmon

[split salmon]

using the semi-colon (;)

@bronze merlin

[bronze merlin]

wtf

dude i know it used to end line

it's an statesment\

i wanna split them

[languid hare]

can you elaborate

whats the original code

and what properties do you want in the modified code

[bronze merlin]

well

here

while True:
  print("yes")
print("no")
print("nah")

i wanna split 2 print out of the While statesment

anyway to do that?

[languid hare]

can you use proper grammar, cant really understand sry

[languid hare]

this specifically

[bronze merlin]

easily explain

i'll take print("no") and print("nah") out of while statesment

but how

it's for one line code

[languid hare]

what do you mean

take out of while statement

[bronze merlin]

take it out

o0f while statesment

in 1 line

please understand i don't know how to explain it out

[quartz wave]

they mean take it out in a one-line statement

[bronze merlin]

while True: print("yes") #HELP ME SPLIT THEM BOTH OUT

[bronze merlin]

yes!

[quartz wave]

like you can't do this ```py
while True: print("yes"); "have beyond outside of the while loop"; print("no"); print("nah")

[bronze merlin]

yeah we can't just do that

it will stay in while statesment

giving it a tuple caused syntax error

[languid hare]

so the no and nah are inside the while block

you dont want them in there

[bronze merlin]

qyes

yes

[languid hare]

well i dont believe it's possible, it expects a newline to delimit the "block". another option is to use a comprehension over an infinite iterable

[bronze merlin]

it's possible

[languid hare]

how do you know

[bronze merlin]

well

[sly root]

you just want one-liner while loop?

[bronze merlin]

probably?

there must be way to escape that line

[sly root]

with basic while its not possible

[bronze merlin]

anyway to do?

[sly root]

you can do this:

[some_function for(_)in map(sum,enumerate(iter(int,1)))];another_function_here```

[languid hare]

there must be a way to delimit the while. that way is the newline. with your constraint, it's not possible.

[sly root]

map(sum,enumerate(iter(int,1))) is an alternative to while True

[bronze merlin]

jesus christ

you turned me into christian for a second

[languid hare]

^

[bronze merlin]

[ was not closed keep terrifying me all along

dammit

[earnest wing]

while cond: thing
[thing for _ in iter(lambda:not cond,1)]

[bronze merlin]

alr thanks for help

[bronze merlin]

nah just saying that ( and [ was not closed is the worst thing happens without even realizing it

[sly root]

@bronze merlin Так ты русский?

[bronze merlin]

da

[sly root]

впервые вижу тут русского лол

[bronze merlin]

i don't have my keyboard layout rn i couldn't write

[split salmon]

!e
b=lambda a: chr(int(('0'(8-len(bin(ord(a))[2:]))+bin(ord(a))[2:])[4:]+('0'(8-len(bin(ord(a))[2:]))+bin(ord(a))[2:])[::-1][4:][::-1],2));exec(''.join([b(z)for z in "\x07'\x96æG\x82r\x84VÆÆö\x12\x02ä\x966V\x02Gö\x02ÖVVG\x02\x97öWÂ\x02\x94ÅrÖ\x02ö&fW76\x16GVF\x12r\x92"]))

[night quarry]

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

Hello! Nice to meet you, I'm obfuscated!

[bronze merlin]

fr?

[bronze merlin]

yes

obfuscasted is in it's name

yow static cast char

[split salmon]

I coded it myself

[bronze merlin]

can you explain this part

[sly root]

?

[bronze merlin]

map(sum,enumerate(iter(int,1)))

[sly root]

basically you're iterating over int infinite times

[sly root]

map(sum,..) isn't even needed here but it works anyway

[bronze merlin]

what it basically does

[sly root]

iterating over a type

with iter(int,1) you're just making a iterator object, and enumerating it using enumerate(...)

[south totem]

full list of magic methods?

[knotty delta]

!e print(dir(object))

[night quarry]

@knotty delta :white_check_mark: Your eval job has completed with return code 0.

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

[knotty delta]

possibly

oh nvm that's not all

http://docs.python.org/reference/datamodel.html#special-method-names

[azure meteor]

!e print("hello world")

[night quarry]

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

hello world

[azure meteor]

! print(globals())

!e print(globals())

[night quarry]

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

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

[restive void]

#bot-commands

[floral meteor]

@azure meteor ❌ Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 1, in <module>
003 | WrongChannelError: please use #bot-commands

[sly root]

@floral meteor do you implemented the functions and classes in bf snippet?

if not, i think it will be better with this design:

FUNCTION(NAME)[
  function body or lambda
]

or

NAME: FN(..argument list..)[function body]```

[steady lily]

why?

the pow functions give different answers and it isnt equal?

[earnest wing]

pow returns a float meaning it has limited precision

[quartz wave]

1. as olivia said, pow returns a float which has only about 15 digits of precision
2. totally not the right channel to ask for help, see #❓｜how-to-get-help

[rugged sparrow]

https://github.com/chilaxan/pysnippets/blob/main/byref.py i ended up writing a conclusive implementation

@fleet bridge ^ you might find this interesting

@burnt pasture I ended up writing an implementation, its awful

[burnt pasture]

i thought you were saying that I ended up writing an implementation 🙂

[frozen hazel]

https://media.discordapp.net/attachments/833965088372424745/927311793209110528/unknown.png

[rugged sparrow]

had to work around something that i think is a bug in STORE_GLOBAL opcode

https://github.com/python/cpython/blob/main/Python/ceval.c#L2869-L2878

[night quarry]

Python/ceval.c lines 2869 to 2878

TARGET(STORE_GLOBAL) {
    PyObject *name = GETITEM(names, oparg);
    PyObject *v = POP();
    int err;
    err = PyDict_SetItem(GLOBALS(), name, v);
    Py_DECREF(v);
    if (err != 0)
        goto error;
    DISPATCH();
}```

[rugged sparrow]

it uses PyDict_SetItem which assumes GLOBALS() is a dict (not a subclass)

but you can set GLOBALS() to a dict subclass by passing a dict subclass into the function constructor

LOAD_GLOBAL has checks (and uses the less specialized path for dict subclasses)

[quartz wave]

goddamn what

157 lines of code for just that?

[rugged sparrow]

yea theres a lot needed for it to even start to cover all the edge cases

and i prob missed many

im proud of this function py def find_segments(co_code, idx, end_size): # take in set of opcodes and final stack size # determine segments of code for each final stack item stack_size = 0 isjump = [*dis.hasjabs, *dis.hasjrel].__contains__ segments = [] seg = bytearray() seg_marker = 0 while isjump(co_code[idx]) or stack_size != end_size: opcode = co_code[idx] oparg = co_code[idx + 1] seg[:] = bytes([opcode, oparg]) + seg jump = False if isjump(opcode) else None change = dis.stack_effect(opcode, oparg if opcode >= dis.HAVE_ARGUMENT else None, jump=jump) stack_size += change seg_marker += change idx -= 2 if seg_marker == 1 and not isjump(co_code[idx]): segments.append(seg) seg = bytearray() seg_marker = 0 return list(reversed(segments)) which i use to split the arguments of a function call into sets of opcodes

lets me look for direct variable loads in the function call at runtime (to find the names in the upper frame)

[rugged sparrow]

doesnt even include the asm_hook and f_locals libs that are on that repo as well

[quartz wave]

so i have a simpler solution that involves a snippet you posted in #esoteric-python some time ago

!e ```py
def getframe(level=0):
try:raise
except Exception as e:
frame = e.traceback.tb_frame.f_back
for () in [()] * level:
frame = frame.f_back
return frame

def foo(val, frame=0):
frame = getframe(frame + 1)
name = frame.f_code.co_code[frame.f_lasti - 1]
load_inst = frame.f_code.co_code[frame.f_lasti - 2]
if load_inst in {101, 116}:
frame.f_locals[frame.f_code.co_names[name]] += 1
elif load_inst in {124, 135, 136, 148}:
frame.f_locals[(frame.f_code.co_varnames + frame.f_code.co_cellvars + frame.f_code.co_freevars)[name]] += 1
else:
raise Exception("passed value is a constant")

a = 5
foo(a)
print(a)
b = 1233
foo(b)
print(b)

[night quarry]

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

001 | 6
002 | 1234

[quartz wave]

it's only now i realized i could've just used frame.f_locals instead of complex opcode replacing stuff

[rugged sparrow]

Yea that works but mine is more generalized so it works in almost any situation

And thru a decorator so you can wrap arbitrary functions

[quartz wave]

like ```py
a = "a"
@byref
def foo(b: byref):
b *= 2

foo(a)
print(a)
``` would output aa?

[rugged sparrow]

yup

[rugged sparrow]

that works. I've noticed a few edge cases surrounding kwargs and how they are done with CALL_FUNCTION_EX and CALL_FUNCTION_KW that ill try to fix tomorrow, but as of now most calling conventions work (or are explicitly forbidden)

i had to forbid stuff like foo(*a, ref_var, *b) because it wasn't possible to know how a would expand

i could also theoretically hook into function construction to make the decorator apply itself automatically but ill do that once i have those edge cases fixed

if you want to mess with it, i wrote it specifically for 3.10 (no idea if it works on other versions) and you can just clone it from the github link

[quartz wave]

would be nice if it wasn't just limited to globals

but damn there's gonna be a lot of complexity

[quartz wave]

it definitely won't work in 3.11

there they changed a lot of stuff

[floral meteor]

lol

[quartz wave]

well it's not windows-compatible so i can't improve it or something

[floral meteor]

3.11 sounds like a nightmare already

[quartz wave]

yes
new opcodes

[floral meteor]

3.10 was already a bummer.

match case where instead of otherwise or default it's case _ which makes it confusing if I wanna use _ as a variable

[quartz wave]

>>> dis("a(b)")
              0 RESUME                   0

  1           2 LOAD_NAME                0 (a)
              4 LOAD_NAME                1 (b)
              6 PRECALL_FUNCTION         1
              8 CALL                     0
             10 RETURN_VALUE
``` 3.11 `a(b)`

!e ```py
from dis import dis
dis("a(b)")

[night quarry]

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

001 |   1           0 LOAD_NAME                0 (a)
002 |               2 LOAD_NAME                1 (b)
003 |               4 CALL_FUNCTION            1
004 |               6 RETURN_VALUE

[quartz wave]

3.10 a(b)

[floral meteor]

and I already have {a:lambda...,b:lambda:...}.get(value, d)() for switching

so I'm gonna stay in 3.9

maybe 3.9.7, although I'm still on 3.9.2 cos Im lazy

speaking of lazy I made an autosolver for 4D 2048, and it's reached the quarter-billion tile already

top left black cell

[quartz wave]

that's not lazy

or i guess it's hardworking for laziness

[floral meteor]

the autosolver is really lazy

[quartz wave]

at least you made one

[quartz wave]

!e 3.10 keyword call ```py
from dis import dis
dis("a(b=3)")

[floral meteor]

def run(interface):
  print('\x1bc');from random import randint as r
  while not interface.eng.gameend:
      interface.next_frame(interface.eng.swipe(r(0,7)))

[night quarry]

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

001 |   1           0 LOAD_NAME                0 (a)
002 |               2 LOAD_CONST               0 (3)
003 |               4 LOAD_CONST               1 (('b',))
004 |               6 CALL_FUNCTION_KW         1
005 |               8 RETURN_VALUE

[quartz wave]

then 3.11 keyword call ```py

dis("a(b=3)")
0 RESUME 0

1 2 LOAD_NAME 0 (a)
4 LOAD_CONST 0 (3)
6 PRECALL_FUNCTION 1
8 KW_NAMES 1
10 CALL 1
12 RETURN_VALUE

[floral meteor]

precall function whaaaaat

[quartz wave]

        TARGET(KW_NAMES) {
            assert(oparg < PyTuple_GET_SIZE(consts));
            call_shape.kwnames = GETITEM(consts, oparg);
            DISPATCH();
        }

        TARGET(LOAD_CONST) {
            PREDICTED(LOAD_CONST);
            PyObject *value = GETITEM(consts, oparg);
            Py_INCREF(value);
            PUSH(value);
            DISPATCH();
        }

like the only difference between these two is that the latter pushes to stack

        TARGET(PRECALL_FUNCTION) {
            /* Move ownership of reference from stack to call_shape */
            call_shape.callable = PEEK(oparg + 1);
            call_shape.postcall_shrink = 1;

            call_shape.total_args = oparg;
            assert(call_shape.kwnames == NULL);
#ifdef Py_STATS
            extern int _PySpecialization_ClassifyCallable(PyObject *);
            SpecializationStats *stats =
                &_py_stats.opcode_stats[PRECALL_FUNCTION].specialization;
            stats->failure++;
            int kind = _PySpecialization_ClassifyCallable(call_shape.callable);
            stats->failure_kinds[kind]++;
#endif
            DISPATCH();
        }

PRECALL_FUNCTION

>>> dis("a + 3")
              0 RESUME                   0

  1           2 LOAD_NAME                0 (a)
              4 LOAD_CONST               0 (3)
              6 BINARY_OP                0 (+)
              8 RETURN_VALUE
``` also `BINARY_ADD` is now `BINARY_OP 0`

[floral meteor]

I've been calling this an AI all day just because it looks like one and it's doing better at 4D 2048 then me.

[quartz wave]

nice

[floral meteor]

I managed to screenshot it in the middle of printing.

and there's the evidence that this is in the terminal.

[floral meteor]

this is smarter than me. D:

[quartz wave]

how did you create it then 🤔

i mean it's just made for a single purpose

[floral meteor]

and it does that single purpose better than me

although it's been running for about 36 hours

to get the quarter-billion tile

[potent comet]

The rationale behind the new opcodes is here:
https://github.com/faster-cpython/ideas/issues/267#issuecomment-1042928991
By breaking up call into two opcodes, you can have 1 that deals with the type of callable (Python, method, builtin, c function, etc), and then one that deals with the sort of arguments passed (positional, positional+keyword, complex unpacking). Each of those can then be swapped to specialised opcodes independently.

[near gust]

code golf challenge: print all of the lyrics to Never Gonna Give You Up by Rick Astley in the least amount of characters without using web requests or files

[floral meteor]

someone's gonna think up of some crazy formula to generate the text.

most people will just apply lossless compression to the string

I'm not gonna participate cos there's a massive spider in my bathroom

I made eye contact with it

well, two of its 8 eyes

[near gust]

You can do both at the same time :)

[floral meteor]

it's looking at me funny

[near gust]

you can use compression and weird magic

[floral meteor]

mathematical magic

[near gust]

So far my record in my own programming language is 57,370ish characters

[floral meteor]

I got 4 dimensional 2048 in 5000 characters exactly

[near gust]

because I used a python program to generate it

Not even math can save you here https://esolangs.org/wiki/Brainfuckn't

[floral meteor]

and the last few characters were a comment that said #5000 chars

[near gust]

Hello, World! with a trailing newline in 104 characters is the best I could do with this crappy language

++++++~>~>>~>~<<<<.@^>^>>^<<<.@>>>>^..<<<<@|.-@>>>>>.[>]<.+^<.@.[>]<^<<<.>>^<<<<<.>>&<<.[>-]~++~++<.--@.

It's hard to do math magic when you don't have arithmetic

To add 2 8-bit numbers, I believe it took me around 300ish characters

[floral meteor]

"bit fuckery"

[near gust]

"bit fuckery"

basically, standard formulas for generating output don't work here

and don't look at my shitty C++ code that I used to implement the language

I'm working on a rewrite in Rust and maybe I could do a second rewrite in C++

however, it uses 8x more memory to work on bytes instead of bits to allow vectorization of certain operations

rn, a single ^ is like 110 assembly instructions, but there's a bunch of vectorization

[floral meteor]

wait, you're the one who wrote that page?

noice

[sly root]

try to codegolf this lexer

tried my best

def lex(_0,_1,_2=0,_3=[]):
  while _2<len(_0):
    _4=None
    for(_5)in _1:
      if(_4:=re.compile(_5[0]).match(_0,_2)):
        _3+=([],[(_4.group(0),_5[1])])[len(_5)>1and _5[1]is not None];break
    if not _4:sys.stderr.write('Illegal character: %s on line %d'%(_0[_2],_2))
    else:_2=_4.end(0)
  return _3```

input:

const x: str = "hello";```

output:

('const', 'RESERVED')                                   ('x', 'ID')                                             (':', 'RESERVED')                                       ('str', 'ID')                                           ('=', 'RESERVED')                                       ('"hello"', 'STR')                                      (';', 'RESERVED')```

tokens:

RESERVED = 'RESERVED'
INT      = 'INT'
STR      = 'STR'
ID       = 'ID'
tokens = [
    (r'[ \n\t]+', None),
    (r'#[^\n]*', None),
    (r'\=', RESERVED),
    (r'\:\=', RESERVED),
    (r'\(', RESERVED),
    (r'\)', RESERVED),
    (r'\{', RESERVED),
    (r'\}', RESERVED),
    (r'\[', RESERVED),
    (r'\]', RESERVED),
    (r'\:', RESERVED),
    (r'\;', RESERVED),
    (r'\+', RESERVED),
    (r'\-', RESERVED),
    (r'\*', RESERVED),
    (r'\/', RESERVED),
    (r'\<\=', RESERVED),
    (r'\<\=\=', RESERVED),
    (r'\<', RESERVED),
    (r'\<\<', RESERVED),
    (r'\>\=', RESERVED),
    (r'\>\=\=', RESERVED),
    (r'\>', RESERVED),
    (r'\>\>', RESERVED),
    (r'\=\=', RESERVED),
    (r'\=\=\=', RESERVED),
    (r'\!\=', RESERVED),
    (r'\!\=\=', RESERVED),
    (r'and', RESERVED),
    (r'or', RESERVED),
    (r'not', RESERVED),
    (r'if', RESERVED),
    (r'elif', RESERVED),
    (r'else', RESERVED),
    (r'while', RESERVED),
    (r'let', RESERVED),
    (r'const', RESERVED),
    (r'[0-9]+', INT),
    (r'\".*\"', STR),
    (r'[A-Za-z][A-Za-z0-9_]*', ID),
]```

[sly root]

_0 is input text and _1 is token list

[near gust]

I was the one who wrote the page and the language :)

the documentation is pretty bad lmao

it doesn't exactly say how to put a character into the tape (like are the most significant bits placed first or ???)

I'm also going to try to brute force a shorter hello world

but for that, I need to make a safe version of the interpreter that can take a file as an argument

and also doesn't segfault

and I need to figure out how to make sure I generate a program that halts

[floral meteor]

segfault go brrrrrrr

[frozen hazel]

seems like an alternative of brainfuck

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

Hello World! in bf, pretty similar

[near gust]

not really similar at all in how it works. They look similar because brainfuckn't IS a superset of a brainfuck with binary cells

and I decided to use similar notation

[frozen hazel]

Hmm, alright.

[near gust]

the + increases the pointer size, not the pointer value

[formal sandal]

Any ideas for a concurrent/parallel version of Brainfuck?

I had an idea like this: extend the language with a single ! character. This character sets the current I/O port to the value on the cursor.

By default the I/O port is 0, which means standard input & standard output.

But if you set it to something other than 0, it reads to/writes from a different "thread" with the given port number

what I don't know is to how to spawn these threads from within the language.

Maybe there could be another pair of brackets, like { and }, which sets the program on the thread with the current port?

[near gust]

There is systemf which allows syscalls

[jagged osprey]

you could implement time sharing using brainfuck code itself

[formal sandal]

I guess, it is Turing complete after all

I was wondering about this in the context of making an HTTP client in brainfuck

[jagged osprey]

it is easier if you make it async, instead of context-switching at any point, like threads do

[formal sandal]

I think I can make it work by only context switching on . and ,

or even only on ,

Also, I imagined that each thread would have an entirely isolated memory. So it shouldn't matter when they context-switch, right?

!e

def bftopy(bf):
    lines = ["async def _bf_fun(io):", "    p,x,m=0,0,{}"]
    indent = 1
    addline = lambda line: lines.append(" " * indent * 4 + line)
    for char in bf:
        if char == "+":
            addline("m[p]=m.get(p,0)+1")
        elif char == "+":
            addline("m[p]=m.get(p,0)-1")
        elif char == ">":
            addline("p+=1")
        elif char == "<":
            addline("p-=1")
        elif char == "[":
            addline("while m.get(p):")
            indent += 1
        elif char == "]":
            indent -= 1
        elif char == "!":
            addline("x=m.get(p,0)")
        elif char == ".":
            addline("await io.write(x,m.get(p,0))")
        elif char == ",":
            addline("m[p]=await io.read(x)")
    return "\n".join(lines)

print(bftopy(", >+!< .>>.<< >+!< .>>.<<"))

[night quarry]

@formal sandal :white_check_mark: Your eval job has completed with return code 0.

001 | async def _bf_fun(io):
002 |     p,x,m=0,0,{}
003 |     m[p]=await io.read(x)
004 |     p+=1
005 |     m[p]=m.get(p,0)+1
006 |     x=m.get(p,0)
007 |     p-=1
008 |     await io.write(x,m.get(p,0))
009 |     p+=1
010 |     p+=1
011 |     await io.write(x,m.get(p,0))
... (truncated - too many lines)

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

[formal sandal]

@jagged osprey something like this^

[jagged osprey]

if they are isolated I don't see the point

I guess it doesn't need a point 😄

[formal sandal]

the point of what 🤔

the idea is that they can communicate with messages, using . and ,

!e

import asyncio

def bftopy(bf):
    lines, indent = ["async def _bf_fun(io):", "    p,x,m=0,0,{}"], 1
    addline = lambda line: lines.append(" " * indent * 4 + line)
    for char in bf:
        if char == "+": addline("m[p]=m.get(p,0)+1")
        elif char == "+": addline("m[p]=m.get(p,0)-1")
        elif char == ">": addline("p+=1")
        elif char == "<": addline("p-=1")
        elif char == "[": addline((indent := indent + 1) or "while m.get(p):")
        elif char == "]": indent -= 1
        elif char == "!": addline("x=m.get(p,0)")
        elif char == ".": addline("await io.write(x,m.get(p,0))")
        elif char == ",": addline("m[p]=await io.read(x)")
    exec("\n".join(lines), (g := {}))
    return g["_bf_fun"]

class DummyIO:
    def __init__(self, inputs):
        self._inputs = list(reversed(inputs))

    async def read(self, _port):
        if not self._inputs:
            await asyncio.Future()
        return self._inputs.pop()

    async def write(self, port, data):
        print(f"{data} on port {port}")

fn = bftopy(", >+!< .>>.<< >+!< .>>.<<")
asyncio.run(fn(DummyIO([5, 6])))

[night quarry]

@formal sandal :white_check_mark: Your eval job has completed with return code 0.

001 | 5 on port 1
002 | 0 on port 1
003 | 5 on port 2
004 | 0 on port 2

[sly root]

lexer update

#[some_decorator]
trait SomeTrait {
  field: type;
  another_field: int;
  some_field: Some<Generic, Expression>;
}

to

('#[some_decorator]', 'DecoratorExpression', (1, (0, 17)))                                                      ('trait', 'ReservedKeyword', (2, (0, 5)))               ('SomeTrait', 'Identifier', (2, (6, 15)))               ('{', 'ReservedLiteral', (2, (16, 17)))                 ('field', 'Identifier', (3, (2, 7)))                    (':', 'ReservedLiteral', (3, (7, 8)))                   ('type', 'Identifier', (3, (9, 13)))                    ('another_field', 'Identifier', (4, (2, 15)))           (':', 'ReservedLiteral', (4, (15, 16)))                 ('int', 'Identifier', (4, (17, 20)))                    ('some_field', 'Identifier', (5, (2, 12)))              (':', 'ReservedLiteral', (5, (12, 13)))                 ('Some<Generic, Expression>', 'GenericExpression', (5, (14, 39)))                                               ('}', 'ReservedLiteral', (6, (0, 1)))```

[sly root]

lexer error handling

some_field: Some<Generic, Expression>;                ————————————————∆ ---- illegal character (<, 5:18)```

[floral meteor]

!e ```py
def _(q):
for l in range(3):
y=q[l:]
while not y[0]and any(y[1:]):q+=[q.pop(l)];y+=[y.pop(0)]
for l in range(3):
if not[*{*q[l:l+2]}][1:]:q[l]+=q.pop(l+1);q+=[0]

___=[0,0,0,1],[0,1,1,1],[1,1,1,1],[2,1,2,2]

[_(__)for __ in ]
print()

[night quarry]

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

([1, 0, 0, 0], [2, 1, 0, 0], [2, 2, 0, 0], [2, 1, 4, 0])

[floral meteor]

what even is this field of mathematics called, that I can do these things to these arrays?

[upbeat sonnet]

dis negrozma is writing to a py file 💀

[upbeat sonnet]

dont think its mathematics at this point

[long fulcrum]

r,p=range,print
l=x=0
h=9
k=[[0for i in r(10)]for j in r(10)]
for v in r(5):
    for i in r(l,h+1):k[v][i]=x;x+=1
    for i in r(l+1,h+1):k[i][h]=x;x+=1
    for i in r(h-1,l-1,-1):k[h][i]=x;x+=1
    for i in r(h-1,l,-1):k[i][l]=x;x+=1
    l+=1;h-=1
for i in r(10):
    for j in r(10):k[0][j]=" "+str(k[0][j]);p(k[i][j],end=" ")
    p()```

how could I golf this

k[0][j]=" "+str(k[0][j])
I think this part can definitely be reworked

[upbeat sonnet]

!e print(help)

[night quarry]

@upbeat sonnet :x: Your eval job has completed with return code 1.

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

[upbeat sonnet]

i mean

!e py from pydocs import help print(help)

ok?

!e py from pydoc import help print(help)

!e py from pydoc import help print(help)

[night quarry]

@upbeat sonnet :white_check_mark: Your eval job has completed with return code 0.

<pydoc.Helper instance>

[upbeat sonnet]

ic

!e py import pydoc print(pydoc.__getattr__)

[night quarry]

@upbeat sonnet :x: Your eval job has completed with return code 1.

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

[upbeat sonnet]

hmmmmmm

!e py import pydoc print(pydoc.__init__)

[night quarry]

@upbeat sonnet :white_check_mark: Your eval job has completed with return code 0.

<method-wrapper '__init__' of module object at 0x7fd133c8bd30>

[bronze merlin]

!e

print("yes") if  1 == 1; print("no") elif 0 == 0;

[night quarry]

@bronze merlin :x: Your eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     print("yes") if  1 == 1; print("no") elif 0 == 0;
003 |     ^^^^^^^^^^^^^^^^^^^^^^^
004 | SyntaxError: expected 'else' after 'if' expression

[bronze merlin]

hm

!e

print("yes") if  1 == 1; print("no") if 0 == 0; else None;

[night quarry]

@bronze merlin :x: Your eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     print("yes") if  1 == 1; print("no") if 0 == 0; else None;
003 |     ^^^^^^^^^^^^^^^^^^^^^^^
004 | SyntaxError: expected 'else' after 'if' expression

[quartz wave]

what are you trying to do

[bronze merlin]

testing some syntax

[quartz wave]

thanks to you i'm just about to make that syntax possible except for the else part

[bronze merlin]

ye

!e

print("yes") if 1 == 1 else None; print("no") if 0 == 0 else None

oops

[night quarry]

@bronze merlin :white_check_mark: Your eval job has completed with return code 0.

001 | yes
002 | no

[bronze merlin]

there we go

[near gust]

never thought I would be legitimately doing this, but here we go

[jovial monolith]

tuple

[jovial monolith]

wtf

pointers in python 🤯

[quartz wave]

!e ```py
#ifndef __cplusplus
#include <stdio.h>
#define and ?
#define or :
#define def int
#define main() main() { a
#define None 0; }
#define print(x) puts(x);
#/*
"""#"/
int
#/
"""#"*/
C = 0;

def main():
print(++C and "C" or "python")
return None

#if 0
main ()
#endif
#else
#include <iostream>
#/*
"""#"/
int main() {
puts("C++");
}
#/
"""#"*/
#endif

[night quarry]

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

python

[quartz wave]

prints C in C, C++ in C++, and python in python

[jovial monolith]

woah how does this work?

[quartz wave]

not sure how to explain

[near gust]

making a randomizer for a specific pokemon romhack because the universal randomizer doesn't work

[jovial monolith]

rom

[quartz wave]

uses C preprocessor directives and takes advantage of the fact that python uses # as the starting symbol of a comment

[jovial monolith]

yeah i can tell, but ++C, is that valid python syntax

[quartz wave]

yes

two unary adds

which do absolutely nothing in python

[jovial monolith]

ah

[quartz wave]

!e ```py
a = int(pow(2, 31, 2**31-1))
b = 1
print(a is b, a, b)

[night quarry]

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

False 1 1

[quartz wave]

non-cached integers

[knotty delta]

wat does da third argyument do

[quartz wave]

mod

[knotty delta]

ooh

[quartz wave]

probably will be fixed in python 3.11

[knotty delta]

!e

custom_block = ">:)"
def test() -> lambda:\
    customblock:
        print("hi")
test()

[night quarry]

@knotty delta :white_check_mark: Your eval job has completed with return code 0.

hi

[wary shoal]

wait what

Confusion

oh damn

I now see it

[knotty delta]

custom_block is a part of the lambda i think, and it does nothing

then the : is just the def's

[wary shoal]

That's amazing

[quartz wave]

you know code is good when it confuses even people who have studied python grammar for months

[knotty delta]

lmfao

[languid hare]

!e ```py
from fishhook import hook

@hook(type(lambda: 0))
def rrshift(self, other):
return self(other)

@hook(type(print))
def rrshift(self, other):
return self(other)

@hook(type)
def rrshift(self, other):
return self(other)

(
123
>> (lambda x: x*2)
>> str
>> (lambda x: x[::-1])
>> print
)

[night quarry]

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

642

[languid hare]

@maiden blaze

:)

[maiden blaze]

very nice

hmm i think i had something similar

[languid hare]

did you ever figure out a _ replacement

the math operators are easy enough to deal with _, but function calls die

[maiden blaze]

ah never mind i had cursed method chaining, not piping
#bot-commands message

[languid hare]

[maiden blaze]

hmm can you rewrite ASTs in a decorator

i know you can change the bytecode

[languid hare]

hmm

well numba (probably) reads the ast

[maiden blaze]

oh wait i could get the original ast with inspect, change it, compile the ast to get its bytecode and rewrite the bytecode to that

so rewrite the function call nodes which have an argument named _ to be lambda _: f(_) instead

[languid hare]

https://numba.pydata.org/numba-doc/latest/user/5minguide.html#how-does-numba-work

Numba reads the Python bytecode for a decorated function
seems about right, so we just gotta read the ast, edit, compile a new function

[maiden blaze]

hm

[earnest wing]

If you could find a way to intercept function calls from the object being passed as argument... I wonder whether audit events or GC hooks could be useful

[maiden blaze]

@languid hare https://paste.pythondiscord.com/edosisaser

[languid hare]

hot

[maiden blaze]

oh i forgot to remove the has_underscore function

wtv

[languid hare]

a good use of the match statement too

[maiden blaze]

probably not foolproof since it'll change it if there's any _ at all
might be some way to break it with sthn like for x, _ in ...

[floral meteor]

!e this could be used instead of modulus for modulus operations! ```py
a = 34
print(pow(a,1,5))
print(34%5)

[night quarry]

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

001 | 4
002 | 4

[floral meteor]

obfuscation 100

[quartz wave]

i can't believe i didn't spot this bug when i added long_mod to Objects/longobject.c

[rapid sparrow]

wow

[hard epoch]

im so confused 😭 i started python not long ago and i wanted to see if i could earn anything in this server but everything is so complicaated

[quartz wave]

probably because you're in this channel?

[hard epoch]

is this a bad channel?

[quartz wave]

for beginners yes

you should stay away from this channel if you don't wanna be confused

[floral meteor]

this confusing is channel

[languid hare]

this channel is more for doing over the top things with python. dont worry too much.

[floral meteor]

did I make a new mathematics?

[floral meteor]

It... tastes... like discrete mathematics, but the only field I recognised here is set theory where I used it to obfuscate determining equality of q[l] == q[l+1]

[quartz wave]

me just casually making this code windows-compatible using a modified cpython

it works

noice

[quartz wave]

hey what if i made a pull request for this

[hard epoch]

whats the non confusing channel?

and any tips for where i can learn more python

they all look confusing

[quartz wave]

!resources

[night quarry]

Resources

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

[hard epoch]

oh thx!

[knotty delta]

any tips where I can learn more #esoteric-python

[quartz wave]

probably not
you can look at IFcoltransG's guide but he's kinda missing for 2 months right now and may not have updated his esoteric python guide

[knotty delta]

🙆‍♀️

[quartz wave]

today i just learned that you can use iterator.__setstate__ to reset an iterator

[languid hare]

you what

[rugged sparrow]

not all iterators support __setstate__ its part of how pickle is implemented

!e py i = iter(range(2000)) for _ in range(1000): next(i) print(i.__reduce__())

[night quarry]

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

(<built-in function iter>, (range(0, 2000),), 1000)

[rugged sparrow]

!e ```py
i = iter(range(0, 2000))
i.setstate(1000)
print(next(i))

[night quarry]

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

1000

[solemn laurel]

is this the old advanced python channel?

[sick hound]

someone's probably already done this but
why not make a brainfuck compiler by simply running it's compiler for c in python

[solemn laurel]

OT but your MAL is amazing

[sick hound]

ty

[solemn laurel]

[solemn laurel]

does anyone have experience with this error? SystemError: no locals when loading <NULL>

here's the code i was playing around with when i encountered this error:

def print_members(obj):
    for attr in dir(obj):
        if attr[:2] == '__':
            continue

        val = getattr(obj, attr)

        print(f'{attr}: {val}')

class abc:
    def __init__(self):
        pass

print_members(abc.__init__.__code__)

a = abc()

src = 'self.x = 123\nself.stu = "hola"'
code_obj = compile(src, __file__, 'exec')

code_obj = code_obj.replace(
    co_argcount = 1,
    co_posonlyargcount = code_obj.co_posonlyargcount,
    co_kwonlyargcount = code_obj.co_kwonlyargcount,
    co_nlocals = code_obj.co_nlocals,
    co_stacksize = code_obj.co_stacksize,
    co_flags = 0x43,
    co_firstlineno = code_obj.co_firstlineno,
    co_code = code_obj.co_code,
    co_consts = code_obj.co_consts,
    co_names = ('x', 'stu'),
    co_varnames = ('self',),
    co_freevars = code_obj.co_freevars,
    co_cellvars = code_obj.co_cellvars,
    co_filename = code_obj.co_filename,
    co_name = '__init__',
    co_lnotab = code_obj.co_lnotab
)

print()
print_members(code_obj)

abc.__init__.__code__ = code_obj
b = abc()

print(dir(a))
print(dir(b))

the full traceback:

Traceback (most recent call last):
  File "/home/anthy/Work/py/dex-tools/pydex/test.py", line 53, in <module>
    b = abc()
  File "/home/anthy/Work/py/dex-tools/pydex/test.py", line 1, in __init__
    def print_members(obj):
SystemError: no locals when loading 'x'

[quartz wave]

found the error:

1. co_names doesn't have 3 elements (you need to add some sort of padding, or replace some LOAD_NAME opargs with the correct numbers)
2. co_nlocals, co_argcount, and co_posonlyargcount must be 1
3. the code still thinks self is in co_names, so you have to do py co_code = code_obj.co_code.replace(bytes((LOAD_NAME, 0)), bytes((LOAD_FAST, 0))), where ```py
   LOAD_FAST = import('dis').opmap['LOAD_FAST']
   LOAD_NAME = import('dis').opmap['LOAD_NAME']

4. this is not so much of an error but more like a tip: if you're just gonna do ```py
    co_X = code_obj.co_X,
``` then just don't do it at all, the code will still be the same since you're doing `code_obj.replace`

!e ```py

class abc:
def init(self):
pass

a = abc()

src = 'self.x = 123\nself.stu = "hola"'
code_obj = compile(src, 'main.py', 'exec')

LOAD_FAST = import('dis').opmap['LOAD_FAST']
LOAD_NAME = import('dis').opmap['LOAD_NAME']

code_obj = code_obj.replace(
co_argcount = 1,
co_posonlyargcount = 1,
co_nlocals = 1,
co_flags = 0x43,
co_code = code_obj.co_code.replace(bytes((LOAD_NAME, 0)), bytes((LOAD_FAST, 0))),
co_names = ('NAME FOR PADDING ONLY', 'x', 'stu'),
co_varnames = ('self',),
co_name = 'init',
)
abc.init.code = code_obj
b = abc()

print(dir(a))
print(dir(b))

[night quarry]

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

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

[quartz wave]

so it seems to work now

[solemn laurel]

noice, i might have some questions still but let me read through and analyze this

why does co_posonlyargcount have to be 1? in the class function __init__, it's set to 0. does this work differently for class functions than regular functions?

[night quarry]

Hey @steady perch!

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

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

[quartz wave]

it's because the self argument is positional-only i think

[restive void]

No, that's #internals-and-peps

[floral meteor]

We swallowed most of their subtopics and left them with only internals and peps to talk about

[knotty delta]

lmao

[lime bane]

!e ```py
class A:
def init(self):
self.b = 15
self.c = self.b, 8

print((A.init.code.co_stacksize))

[night quarry]

@lime bane :white_check_mark: Your eval job has completed with return code 0.

2

[floral meteor]

remember I made 4 dimensional 2048?
well it's now functional as a command line interface

The engine can load a game from a 4x4x4x4 array, which can be generated from eval(str(engine))

with all this implemented, I still gotta make it a nice round number by adding a comment at the end # this source code for 4D 2048 is exactly 5000 characters, even with this extremely lengthy comment at the end of the file.

[sleek flume]

!e python locals().update({f"vector{i}d": tuple[(float,)*i] for i in range(2, 10)}) print(vector2d) print(vector4d)

[night quarry]

@sleek flume :white_check_mark: Your eval job has completed with return code 0.

001 | tuple[float, float]
002 | tuple[float, float, float, float]

[night quarry]

@craggy bear :warning: Your eval job has completed with return code 0.

[No output]

@craggy bear :x: Your eval job has completed with return code 1.

001 |   File "<string>", line 1
002 |     print(u fuck)
003 |           ^^^^^^
004 | SyntaxError: invalid syntax. Perhaps you forgot a comma?

[craggy bear]

!e
print('"u fuck"')

[night quarry]

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

"u fuck"

[fringe fractal]

e!

[turbid tendon]

Can someone help me i want to translate python to lua but python has functions that lua does not have

like inits and import and classes how would i even use these in lua?

[bronze cloak]

aight so i got a spicy one, anyone know why my code runs faster when i introduce an artificial delay into it? when its commented out im running ~40fps and with it in the code i get ~60. Programs intended purpose is live video analysis

[sick hound]

!e

def sqrt(n):
    if n < 0: return 0

    fn = len(str(n)) * 100
    fp = 0

    while fn != fp: 
        fp = fn
        fn = 0.5 * (fp + n / fn)

    return fn

print(sqrt(634))

[night quarry]

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

25.179356624028344

[sick hound]

!e

def sqrt(n):
    if n < 0: return 0

    fn = len(str(n)) * 100
    fp = 0

    while fn != fp: 
        fp = fn
        fn = 0.5 * (fp + n / fn)

    return fn

print(sqrt(625))

[night quarry]

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

25.0

[sick hound]

god it works 🙌

[sage panther]

Does simulating an Operating System in Python (Terminal) counts as esoteric?

[quartz wave]

any python weirdness counts as esoteric
like modifying bytecode, obfuscation, code gore, python vm languages, golfing, etc.

[finite blaze]

Hey, if anyone like me is looking for small golfing challanges, maybe you could try making roman numerals to decimal converter

Good Luck

[severe canyon]

!e

v={'I':1,'V':5,'X':10,'L':50,'C':100,'D':500,'M':1000}
f=lambda n:(s:=0)or[(s:=s+(a:=v[n[i]])-(a>(b:=v[n[i-1]])and s)*2*b)for i in range(len(n))][-1]

print(f('MMMM'))
print(f('MMXXII'))
print(f('IV'))

[night quarry]

@severe canyon :white_check_mark: Your eval job has completed with return code 0.

001 | 4000
002 | 2022
003 | 4

[severe canyon]

v={'I':1,'V':5,'X':10,'L':50,'C':100,'D':500,'M':1000}
f=lambda n:(s:=0)or[(s:=s+v[c]-(v[c]>(a:=v[n[i-1]])and s)*2*a)for i,c in enumerate(n)][-1]

shorter

[prisma coral]

Slightly shorter

v=dict(zip("IVXLCDM",[1,5,10,50,100,500,1000]))
f=lambda n,s=0:[(s:=s+v[c]-(v[c]>(a:=v[n[i-1]])and s)*2*a)for i,c in enumerate(n)][-1]

[severe canyon]

nice

[knotty delta]

is this a roman numeral to integer golf 👀

oh nvm my golf was int to roman numeral

e.e

[knotty delta]

esopy isn't really a channel you should get help in 👀

#❓｜how-to-get-help

[ruby prawn]

oo ok

[floral meteor]

@ruby prawn ❌ Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 1, in <module>
003 | WrongChannelError: unable to get help from this channel. See how-to-get-help for more information

[sly root]

!e ```py
import importlib
def _0imp(_0,_1=None,_2=None,_3=(),_4=0):
try:_5=_1['name']if _1 else'main'
except KeyError:_5='main'
if'os'==_0:
if any((_xx:=_x)in['remove','rmdir','unlink']for(_x)in _3 or[]):raise ImportError(f"'{_xx}' is restricted.")
elif'sys'==_0:
if any((_xx:=_x)in['exit']for(_x)in _3 or[]):raise ImportError(f"'{_xx}' is restricted.")
elif'importlib'==_0:raise ImportError(f"module '{_0}' is restricted.")
return importlib.import(_0,_1,_2,_3,_4)
builtins.dict['import']=_0imp

from os import rmdir

[night quarry]

@sly root :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 13, in <module>
003 |   File "<string>", line 6, in _0imp
004 | ImportError: 'rmdir' is restricted.

[finite blaze]

Great job @severe canyon @prisma coral

My answer was slightly longer :p

[quartz wave]

noice

anyone wanna add anything to cpython

[sly root]

!e ```py
import importlib,types,dis,sys
def _0hook(_0,_1):
if'exec'==_0 and _1:
if _1 and isinstance(_1[0],types.CodeType):
_2,_3=_1[0],None;_4=[dis.opname[_x]for(_x)in _2.co_code]
for(_i,_n)in enumerate(_4):
if'STORE_NAME'==_n:_3=_2.co_code[_i+1];break
if'LOAD_NAME'==_n:return
if _2 is not None and _3 is not None:
if(_5:=_2.co_names[_3])in['_0hook','_0imp']:
raise Exception("cannot re-assign variable")
def _0imp(_0,_1=None,_2=None,_3=(),_4=0):
try:_5=_1['name']if _1 else'main'
except KeyError:_5='main'
if'os'==_0 and any((_xx:=_x)in['remove','rmdir','unlink']for(_x)in _3 or[]):raise ImportError(f"'{_xx}' is restricted.")
elif'sys'==_0 and any((_xx:=_x)in['exit']for(_x)in _3 or[]):raise ImportError(f"'{_xx}' is restricted.")
elif'importlib'==_0:raise ImportError(f"module '{_0}' is restricted.")
return importlib.import(_0,_1,_2,_3,_4)
builtins.dict['import']=_0imp
import('sys').addaudithook(_0hook)

from os import rmdir

_0imp=None

[night quarry]

@sly root :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 24, in <module>
003 |   File "<string>", line 16, in _0imp
004 | ImportError: 'rmdir' is restricted.

[sly root]

what? it should give an error

>>> _0imp=None                                                                                Traceback (most recent call last):                     File "restrict.py", line 13, in _0hook        try:_5=_1['__name__']if _1 else'__main__'        Exception: cannot re-assign variable
>>>```

[prisma coral]

Realised you can shorten the first line by doing v=dict(I=1,V=5,X=…)

[sly root]

!e ```py
import importlib,types,dis,sys
def _0hook(_0,_1):
if'exec'==_0 and _1:
if isinstance(_1[0],types.CodeType):
_2,_3=_1[0],None;_4=[dis.opname[_x]for(_x)in _2.co_code]
for(_i,_n)in enumerate(_4):
if'STORE'in _n:_3=_2.co_code[_i+1];break
if _2 is not None and _3 is not None:
if(_5:=_2.co_names[_3])in['builtins','_0hook','_0imp']:
raise Exception("Cannot re-assign variable")
def _0imp(_0,_1=None,_2=None,_3=(),_4=0):
try:_5=_1['name']if _1 else'main'
except KeyError:_5='main'
if'os'==_0 and any((_xx:=_x)in['remove','rmdir','unlink']for(_x)in _3 or[]):raise ImportError(f"'{_xx}' is restricted.")
elif'sys'==_0 and any((_xx:=_x)in['exit']for(_x)in _3 or[]):raise ImportError(f"'{_xx}' is restricted.")
elif'importlib'==_0:raise ImportError(f"module '{_0}' is restricted.")
return importlib.import(_0,_1,_2,_3,_4)
builtins.dict['import']=_0imp
import('sys').addaudithook(_0hook)

from os import rmdir # this should give an error that will day that this function is restricted

_0imp = None # gives error while executing normally or in repl
builtins.dict['import'] = None # should give an error too```

[night quarry]

@sly root :x: Your eval job has completed with return code 1.

001 | Traceback (most recent call last):
002 |   File "<string>", line 21, in <module>
003 |   File "<string>", line 14, in _0imp
004 | ImportError: 'rmdir' is restricted.

[sly root]

anyways, it works outside of snekbox:

>>> __builtins__.__dict__['__import__']=8            Traceback (most recent call last):                     File "restrict.py", line 10, in _0hook        raise Exception("Cannot re-assign variable")     Exception: Cannot re-assign variable                 >>> _0imp=8                                          Traceback (most recent call last):                     File "restrict.py", line 10, in _0hook        raise Exception("Cannot re-assign variable")     Exception: Cannot re-assign variable                 >>> ```

[quartz wave]

i am here to tell you that you can still bypass this hook
thank you for listening and have a good time

[sly root]

how?

have a good time too

[quartz wave]

simple enough
dict assignment

nice use of hooks though

[sly root]

like this?

>>> __builtins__ = {'__import__': 3}                 Traceback (most recent call last):                     File "restrict.py", line 10, in _0hook        raise Exception("Cannot re-assign variable")     Exception: Cannot re-assign variable```

[quartz wave]

no not like that

like this globals()['_0imp']=None

[sly root]

oh

sec

added ```
print(_2.co_names[_3])

globals()['_0imp']=None globals >>>```

i think there's a way to handle that

[quartz wave]

code is really messed up on pc

[sly root]

 print(_2.co_names[_3],"\n",
      [(_x.opname, _x.argval)for(_x)in dis.Bytecode(_2)]
    )

>>> globals()['_0imp']=None                          globals                                               [('LOAD_CONST', None), ('LOAD_NAME', 'globals'), ('CALL_FUNCTION', 0), ('LOAD_CONST', '_0imp'), ('STORE_SUBSCR', None), ('LOAD_CONST', None), ('RETURN_VALUE', None)]                                              >>>```

so the only thing is needed is to check names

[sick hound]

!e def cat(): try: eeee except: cat() cat()

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "<string>", line 3, in cat
003 | NameError: name 'eeee' is not defined
004 | 
005 | During handling of the above exception, another exception occurred:
006 | 
007 | Traceback (most recent call last):
008 |   File "<string>", line 3, in cat
009 | NameError: name 'eeee' is not defined
010 | 
011 | During handling of the above exception, another exception occurred:
... (truncated - too many lines)

Full output: too long to upload

[sly root]

@quartz wave

if(_5:=_2.co_names[_3])in(_z:=['__builtins__','_0hook','_0imp']):
     raise Exception("Cannot re-assign variable")
    elif'globals'==_5:
     if any(_xx in _z for(_xx)in[_x.argval for(_x)in dis.Bytecode(_2)]):
      raise Exception("Cannot re-assign variable")

>>> globals()['_0imp']=None                          Traceback (most recent call last):                     File "restrict.py", line 16, in _0hook        raise Exception("Cannot re-assign variable")     Exception: Cannot re-assign variable                 >>>```

[quartz wave]

cool

[sick hound]

how would one mess up so badly that errors begin occuring during the process of handling the error

[sly root]

if(_5:=_2.co_names[_3])in(_z:=['globals','__builtins__','_0hook','_0imp']):
 if'globals'==_5:if any(_xx in _z[1:]for(_xx)in[_x.argval for(_x)in dis.Bytecode(_2)]):raise _e
 else:raise _e```

[sick hound]

what does this do

[severe canyon]

!e

class String:
    def __init__(s,chars):
        s.string = chars;
    def __pow__(s,a):
        if isinstance(a,int):
            if a == 2:
                ords = [*map(ord,s.string)]
                return '\n'.join(''.join(chr(a*b)for b in ords)for a in ords)
            else:
                raise ValueError("not squared :(")
        else:
            raise ValueError("not an integer")
            
a = String("abc")
print(a**2)

[night quarry]

@severe canyon :white_check_mark: Your eval job has completed with return code 0.

001 | Ⓛ┢▃
002 | ┢▄◦
003 | ▃◦♉

[severe canyon]

[languid hare]

!e

from fishhook import hook
@hook(str)
def __pow__(s,a):
    if isinstance(a,int):
        if a == 2:
            ords = [*map(ord,s)]
            return '\n'.join(''.join(chr(a*b)for b in ords)for a in ords)
        else:
            raise ValueError("not squared :(")
    else:
        raise ValueError("not an integer")
print("abc"**2)

[night quarry]

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

001 | Ⓛ┢▃
002 | ┢▄◦
003 | ▃◦♉

[quartz wave]

finally i can do this locally with my modified cpython

[severe canyon]

i love it

[floral meteor]

"Hello World!"**2

[languid hare]

baseball 门

[knotty delta]

what is this

😨

[floral meteor]

!e ```py
from fishhook import hook;from functools import reduce;from itertools import product
@hook(str)
def pow(s,a):
if isinstance(a,int):
o=*map(ord,s),
return''.join(chr(reduce(int.mul,b))for b in product(o,repeat=a))
else:raise TypeError(str(type(a)))
print("abc"**2, "abc"**3)
print("Hello World!"**2)

[night quarry]

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

001 | Ⓛ┢▃┢▄◦▃◦♉ 󞴡󡇢󣚣󡇢󣜄󥰦󣚣󥰦󨆩󡇢󣜄󥰦󣜄󥲈󨈌󥰦󨈌󪟲󣚣󥰦󨆩󥰦󨈌󪟲󨆩󪟲󬸻
002 | ᑀᱨṠṠἸऀᡸἸ‐Ṡᰠैᱨ⟙⪜⪜⯋ಠ≓⯋⳺⪜❴അṠ⪜ⶐⶐ⻔඀⒴⻔〘ⶐ⨰෬Ṡ⪜ⶐⶐ⻔඀⒴⻔〘ⶐ⨰෬Ἰ⯋⻔⻔〡෠▹〡ㅮ⻔⭜๏ऀಠ඀඀෠Ѐૠ෠เ඀ಀРᡸ≓⒴⒴▹ૠᶑ▹⚾⒴⇼ଷἸ⯋⻔⻔〡෠▹〡ㅮ⻔⭜๏‐⳺〘〘ㅮเ⚾ㅮ㋄〘ⲈາṠ⪜ⶐⶐ⻔඀⒴⻔〘ⶐ⨰෬ᰠ❴⨰⨰⭜ಀ⇼⭜Ⲉ⨰✐೤ैഅ෬෬๏Рଷ๏າ෬೤с

[floral meteor]

now I see why only squared is implemented

[frozen hazel]

tf

[severe canyon]

:)

i thought of the others... never came that far for reasons

btw, the str * str operation, should yield an appropriate matrix

someone implement that

[floral meteor]

but should it really be newline separated?

that's just regularly inserting the character ascii value 10, which makes it a different string to the one acheived from multiplying: data is added from somewhere

where are you getting these chr(10)s from?

[quartz wave]

what's the shortest factorial program you guys can think of

[rugged sparrow]

f=lambda n:n*f(n-1)if n else 1``` this is a decent one

you could prob make smaller

[quartz wave]

f=lambda n:n<2or n*f(n-1)

yeah this is the smallest i can think of

[languid hare]

reduce is a dead end and not even in builtins anymore :p

[languid hare]

hmm does this count? it returns a bool for n=1 but true is technically 1

[quartz wave]

yes

[languid hare]

alright

you posed the question anyway

f=lambda n:1>>n or n*f(n-1)

(not shorter but the true makes me uncomfortable)

[knotty delta]

f=__import__("math").factorial

oh nvm probably longer

but

[earnest wing]

if your criterion is "f: int -> int" then yes it is valid

[gritty mesa]

!e

@lambda _:_()
class _:
    def __format__(self, spec):        
        def test(_):
            header=spec.split("(")[1].split(")")[0].split(";");header[0]=header[0].split("let ")[1];body="\n".join(x.strip() for x in spec.split("[")[1].strip().split("]")[0].strip().split("\n"));globals()[header[0]]=0
            while eval(header[1]):exec(body);globals()[header[0]] += 1
        self.__class__._ = property(test)
        return ""
    


def function() -> f'''
{_:
    for (let i; i < 2; i++)[
        print(i)
        print("Hello world")
    ]
}


''':_._

function()

[night quarry]

@gritty mesa :white_check_mark: Your eval job has completed with return code 0.

001 | 0
002 | Hello world
003 | 1
004 | Hello world

[gritty mesa]

Yeah the syntax highlighter had fun with that one

[severe canyon]

i prefer 0**n instead of 1>>n :P

[knotty delta]

interesting, i wonder how is there no syntax error at all

[languid hare]

half is due to the annotation, half due to the format

[knotty delta]

ohh

[languid hare]

poor syntax highlighter couldn't figure it out though

[alpine peak]

!e

def a(arg=[]):
    arg.append(1)
    print(arg)
    
a()
a()
a.__defaults__[0].clear()
a()
a()

[night quarry]

@alpine peak :white_check_mark: Your eval job has completed with return code 0.

001 | [1]
002 | [1, 1]
003 | [1]
004 | [1, 1]

[fleet bridge]

!e

class X:
    def __iпit__(self):
        print('init called')
        self.x = 123

x = X()
print(x.x)

[night quarry]

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

001 | Traceback (most recent call last):
002 |   File "<string>", line 7, in <module>
003 | AttributeError: 'X' object has no attribute 'x'

[quartz wave]

python with the cyrillic characters again

[knotty delta]

init not called confirmed

[fleet bridge]

__in1t__

[severe canyon]

__innit_bruv__

[knotty delta]

hax

!charinfo iпit

[night quarry]

You are not allowed to use that command here. Please use the #bot-commands channel instead.

[knotty delta]

ok

[earnest wing]

"ipenit"?

[fleet bridge]

"ipit"

[long hamlet]

Foiled lol

[floral meteor]

It rool mysterious, innit?

[finite blaze]

Hey, whats the shortest way to import libs?

from a,b,c,d import *

trying to do something like this

[royal coral]

ironically i use this method

globals().update({i: __import__(i) for i in ["random", "string", "math"]})

[finite blaze]

it is really long tho

interesting nonetheless

[potent comet]

__import__() isn't really intended for user use, you should probably use importlib.import_module() which has a more sane API.

[floral meteor]

Most of the shiny things we play with here aren't intended for user use

[earnest wing]

lambda*a:exec("from %s import*;"*len(a)%a)

[earnest wing]

just pass strings and it should work, I haven't tested though

[sleek flume]

Import goes inside a function (closure), so imported stuff is not in global scope.

[quartz wave]

simple fix ```py
lambdaa:exec("from %s import;"*len(a)%a,globals())

[sleek flume]

Yep, now it works.

[fleet bridge]

lambda*a,scope=globals():{scope.update({k:v for k,v in __import__(m).__dict__.items() if not k.startswith('_')}) for m in a}

[icy wing]

Is there a more elegant way to do this in one line?

f=lambda w:len([None for i in range(len(w)) for j in range(i + 1, len(w)) if w[i]+w[j] in w])

I'm being forced to take an intro to python class so I'm trying to do all functions in one line. I'm currently using the length of an array as a sort of counter and using a list constructor to fill it. Feels stupid and runs slower than another 6-line solution I came up with

[restive void]

Be prepared to fight; some instructors will not like you trying to outsmart them and give you a bad grade for this kind of thing.

[icy wing]

Yeah, I've put the functions in comments above my more standard looking solutions

My instructor seems chill tho hopefully

def count_weights(w):
    count = 0
    for i in range(len(w)): # sets up the 2D iteration space, which includes all values of i and only the values of j greater than i. This is because all of the previous indexes become restircted sequentially, so I don't need to check the uniqueness manually.
        for j in range(i+1, len(w)):
            if w[i]+w[j] in w: # add one to the count if the sum in in the list
                count += 1
    return count
``` the full solution if it helps to see it laid out

[restive void]

I did something similar in highschool, when we had to submit a website with links to different parts and pictures and everything and I handed in a single file (images via base64-URIs, links via hashes and javascripty hiding of irrelevant parts)

[icy wing]

I like the idea of esoteric coding so I've been trying to dip my toes into one-liners

What are some of the best features that aid compact code?

[dark wharf]

Constructing a set would make this massively faster

!e

f=lambda w:(s:=set(w),sum(1 for i in range(len(w)) for j in range(i + 1, len(w)) if w[i]+w[j] in s))[1]
import random, time
random.seed(420)
l=[random.randint(0,1000) for _ in range(500)]
t=time.time()
print(f(l))
print(time.time()-t)
```this takes 0.95 on your original solution

[night quarry]

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

001 | 24429
002 | 0.024204730987548828

[icy wing]

ah, I see

[dark wharf]

Also changed the len to sum of a generator for a little extra speed

[icy wing]

Yeah

[dark wharf]

No need to make a massive list

[icy wing]

yeah, summing the 1s is smart

[dark wharf]

Was debating on swapping out your w[i] thing for an enumerate but idk if it'll speed things up much if at all

[icy wing]

Thanks, I'll try to wrap my head around the specifics

Ah, yeah

I got it all

[dark wharf]

!e

f=lambda w:(s:=set(w),l:=len(w),sum(1 for i,v in enumerate(w) for j in range(i + 1, l) if v+w[j] in s))[-1]
import random, time
random.seed(420)
l=[random.randint(0,1000) for _ in range(500)]
t=time.time()
print(f(l))
print(time.time()-t)
```here's a potentially faster one but I'm on mobile and not bothered to run proper tests

[night quarry]

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

001 | 24429
002 | 0.016857147216796875

[dark wharf]

This uses enumerate and avoids calling len a bunch of times though that's a very marginal gain

[icy wing]

yeah, avoiding the double len call is a smart move

and enumerate could be faster, I'm not sure on that though

It does avoid another array index

but not really if you think about it

ah, yeah it does.

Thanks @dark wharf

Ah, one question