K might as well be out of the sandbox
#esoteric-python
vast wave
!e exec(type((lambda:0).__code__)(0, 0, 0, 0, 1, 0, b'\x97\x00t\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00d\x02\xa6\x01\x00\x00\xab\x01\x00\x00\x00\x00\x00\x00\x00\x00\x01\x00d\x00S\x00', ("hi chat",), ("print",), (), "", "", "", 0, b"", b""))
night quarryBOT
@vast wave :warning: Your 3.11 eval job has completed with return code 139 (SIGSEGV).
[No output]
hm
weak imp
!e `exec(type((lambda:0).__code__)(0, 0, 0, 0, 1, 0, b'\x97\x00t\x01\x00\x00\x00...
You will get it 
vast wave
You will get it <:esnoobCuteS:924452098878357574>
i'm just fucking around at this point, i know what i did wrong in the first place
rough moat
i found this online ๐ญ
versed eagle
fun
low lynx
btw the aoc code golfs are available again for a short period of time
so if any of you guys want to archive certain golfs, now's the time to go grab them
low lynx
fun
sick hound
Golfing brainf looks fun
vast wave
Golfing brainf looks fun
i would rather walk all the way down to africa and back than to golf brainfuck again
sick hound
?
meager zinc
Golfing brainf looks fun
!e This can likely be shorter```py
p="++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++."
s=[i:=0];c=''
for o in p:c+=' '*i+[f'i+=1\n{" "i}if i>=len(s):s+=[0]','i-=1','s[i]+=1','s[i]-=1','print(end=chr(s[i]))','s[i]=ord(import("sys").stdin.read(1))','while s[i]:','s','']['><+-.,['.find(o)]+'\n';i+=(92-ord(o))(o in'][')
exec(c)
@meager zinc :white_check_mark: Your 3.11 eval job has completed with return code 0.
Hello World!
That one has an unlimited stack size, which is atypical so it can be shortened by removing that feature too
versed eagle
Is there any pythonfuck?
this is the shortest "alphabet" that i could come up with which could still do everything that normal python could do
.__()nameldgtribusc
got a shorter one
.__()excmhrlnasd
shorter one
()exlchrbo+~
versed eagle
That one has an unlimited stack size, which is atypical so it can be shortened b...
iirc, brainfuck officially is meant to have a stack size of 30000
sly ibex
hey guys
open(0).readlines()
is just for esoteric codes or is common to see them in normal codes i hope you can get my question
versed eagle
that reads from stdin until it reads an EOF
agile meteor
python isnt bad
ruby idol
Lol i made a script that draws trees for you wherever you click
versed eagle
Lol i made a script that draws trees for you wherever you click
i think you're a bit lost
this is not the channel you're looking for
versed eagle
this is for esoteric python
meager zinc
iirc, brainfuck officially is meant to have a stack size of 30000
Yes, I know, but I was playing with it and I didn't like it for certain things (recursive stuff in particular). Here is a better optimized version:
p="++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++."
s=[i:=0]*30000;c=''
for o in p:c+=' '*i+['i+=1','i-=1','s[i]+=1','s[i]-=1','print(end=chr(s[i]))','s[i]=ord(open(0).read(1))','while s[i]:','s','']['><+-.,['.find(o)]+'\n';i+=(92-ord(o))*(o in'][')
exec(c)
rewriting it as a join might be shorter
I need to try it
here is that:
p="++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++."
s=[i:=0]*30000
exec(''.join(' '*i+['i+=1','i-=1','s[i]+=1','s[i]-=1','print(end=chr(s[i]))','s[i]=ord(open(0).read(1))','while s[i]:','s','']['><+-.,['.find(o)]+(0*(i:=i+(92-ord(o))*(o in']['))or'\n')for o in p))
prob longer
@quartz wave might want to take a look at this
rugged sparrow
Yes, I know, but I was playing with it and I didn't like it for certain things (...
https://github.com/brain-lang/brainfuck/blob/master/brainfuck.md#:~:text=single%2C unsigned 8-bit number does yours follow this spec?
rugged sparrow
and each cell is an unsigned 8bit digit
meager zinc
https://github.com/brain-lang/brainfuck/blob/master/brainfuck.md#:~:text=single%...
no but I could fix it pretty easy by appending ;i=i%256
and stopping negative numbers is more tricky
i=(i or 0)%256
rugged sparrow
!e py print(-1%256)
night quarryBOT
@rugged sparrow :white_check_mark: Your 3.11 eval job has completed with return code 0.
255
Oh yeah I don't need to worry about it
rugged sparrow
ya
meager zinc
let me go back to the one that has an unlimited stack
p="++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++."
s=[i:=0];c=''
for o in p:c+=' '*i+[f'i+=1\n{" "*i}if i>=len(s):s+=[0]','i-=1','s[i]+=1','s[i]-=1','print(end=chr(s[i]))','s[i]=ord(open(0).read(1))','while s[i]:','s','']['><+-.,['.find(o)]+';s[i]%=256\n';i+=(92-ord(o))*(o in'][')
exec(c)
should work as intended
meager zinc
thats wayyyy too verbsoe
versed eagle
thats wayyyy too verbsoe
yeah it's fairly verbose
but the idea is more how they do it, not how short it is
it's easy up shorten once you've gotten a good idea working
meager zinc
```py
p="++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>...
afaik this is working and follows the spec
versed eagle
anyway im gonna go back to being a sleeping gremlin since im not supposed to be awake rn
gn
meager zinc
gn
rugged sparrow
Noice
rugged sparrow
Now if you can make it autogenerate structure definitions that would be cool
solid basalt
Is there any specific reason that range implements a count method? Won't the result always be 0 or 1?
| count(...)
| rangeobject.count(value) -> integer -- return number of occurrences of value
maiden blaze
i think its part of the Sequence abc
solid basalt
that checks out
!e
from collections.abc import Sequence
print(Sequence.count.__doc__)
@solid basalt :white_check_mark: Your 3.11 eval job has completed with return code 0.
S.count(value) -> integer -- return number of occurrences of value
dry mirage
Is there any specific reason that `range` implements a count method? Won't the r...
also no
count is equality, not identity. That optimization can't be made
!e
class Foo:
def __eq__(self, other):
return True
r = range(100)
print(r.count(Foo()))
@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.
100
there is actually a fast path for bool / int exact items though, in that case it returns 1 or 0 without iterating the whole thing
sick hound
What other stuff are there that would be fun to golf?
low lynx
lots of things you could golf
for example a function that calculates the nth digit of pi in decimal
dry mirage
yeah I had an idea before to optimize (float) in range() comparison before
since it took a very long time to return False, but apparently you can't just optimize it because someone may have a custom object that is equal to one of the range elements and therefore in is true
quartz wave
<@310263589913100288> might want to take a look at this
at what?
quartz wave
cereal has a collection
i didn't know i had that
quartz wave
!e
```py
from collections.abc import Sequence
print(Sequence.count.__doc__)
```
why is Sequence a keyword
fleet bridge
discord has very bad highlighting rules
sick hound
discord has very bad highlighting rules
Don't blame Discord directly, blame highlight.js developers
They don't fix issues with individual lexers
golden finch
import ctypes
globals_type=ctypes.py_object.from_address(id(globals())+8)
class new_globals(dict):
__slots__=()
def __getitem__(self,name,bound1=dict,globals_type=globals_type):
try:
globals_type.value = bound1
combinators = {
'S': lambda x:lambda y:lambda z:x(z)(y(z)),
'K': lambda x:lambda y:x,
}
if name[0] in locals()['self'] or name[0] in combinators:
result=combinators[name[0]]
for combinator in name:
if combinator in combinators:
result=result(combinators[combinator])
elif combinator in locals()['self']:
result=result(locals()['self'][combinator])
else:
return self[name]
return result
return self[name]
finally:
globals_type.value=__class__
globals_type.value = new_globals
I = SKK
U = SII
U(print) # <built-in function print>
print(I(1)) # 1
SKI combinator calculus
(heavy inspiration from this channel)
quartz wave
```py
import ctypes
globals_type=ctypes.py_object.from_address(id(globals())+8)...
why locals()['self']?
earnest wing
name lookup is builtins -> locals -> globals, so it intuitively should work
but I haven't tested myself
last locust
name lookup is builtins -> locals -> globals, so it intuitively *should* work
Not quite
It's locals, enclosed, globals, builtins
Builtins is the last thing that's checked, not the first
night quarryBOT
earnest wing
yes i'm aware
fleet bridge
Why __class__ variable is not compiled to LOAD_CONST <class 'Foo'> ?
Because class must be places in co_consts tuple at method creation time, and at this moment class doesn't exist
Thank you
lucid lake
What is esosteric python? Can someone give a summary?
earnest wing
I would describe it as "anything that goes deliberately and shamelessly against the grain of mainstream python programming"
also the channel topic I guess
dry mirage
Why `__class__` variable is not compiled to `LOAD_CONST <class 'Foo'>` ?
well it's not really a constant
it's sort of like a cell / closure variable
night quarryBOT
Python/compile.c lines 1745 to 1754
/* Cook up an implicit __class__ cell. */
int res;
assert(u->u_scope_type == COMPILER_SCOPE_CLASS);
assert(PyDict_GET_SIZE(u->u_cellvars) == 0);
res = PyDict_SetItem(u->u_cellvars, &_Py_ID(__class__),
_PyLong_GetZero());
if (res < 0) {
compiler_unit_free(u);
return 0;
}```i know, but i dont know any case where you want to change value __class__ variable
dry mirage
it just can't be known at the time when constants are compiled
astral rover
it cant be known ever can it?
dry mirage
you sort of build the function code first, then you can find out if it is in a class
astral rover
cause class in subclasses is the subclass
dry mirage
cause __class__ in subclasses is the subclass
I don't think so
!e
class Foo:
def abc(self):
print(__class__)
class SubFoo(Foo):
pass
SubFoo().abc()
@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.
<class '__main__.Foo'>
astral rover
cause the whole rationale behind not allowing Self in staticmethods atleast for me was that you couldnt ever have it always return the enclosing class
maybe im just misremembering though
dry mirage
cause the whole rationale behind not allowing Self in staticmethods atleast for ...
you can very easily, it's just not statically inferable
when a class method returns a cls() instance the type checkers can infer the Self type
astral rover
Yeah but I thought I ran this test for staticmethods and just normal methods and concluded that since there was no way to access the current enclosing class in a staticmethod, there was no way that Self would ever work (inheritance wise)
I'm trying to load the thread I had for the pep but discord is being bad and is just hanging
well i have no idea ig its just lost to time, i dont particularly want to go through the comment history of https://docs.google.com/document/d/1ujuSMXDmSIOJpiZyV7mvBEC8P-y55AgSzXcvhrZciuI/edit#
quartz wave
well i have no idea ig its just lost to time, i dont particularly want to go thr...
why is there a big ร in the middle of the docs file
astral rover
idk people must have just messed with it after finding it on typing-sig
turbid glade
@gleaming linden have u ever heard of clash of code?
turbid glade
but I haven't tried it before
well here is coode golf game mode
it is cool
it is kinda like leet code
questions
but who can do it in least amount of code
join up
๐
gleaming linden
done
gleaming linden
t,s=map(int,open(0))
print(t*s//60)
print(int((int(input()) / 60)*int(input())))
how could you use for inputs?
oh
didn't know that
that's smart
@gleaming linden wanna go again
gleaming linden
sure
gleaming linden
I need to learn pyth
turbid glade
@gleaming linden how do u do it
turbid glade
print(sum((n*3, n*5)[n%2] for n in [int(input()) for _ in range(int(input()))]))
[1:] to get rid of the first one
gleaming linden
I came up with a 56 in the first minute lol
turbid glade
๐ฆ
gleaming linden
something like ```py
_,*n=map(int,open(0))
print(...)
22c lol
gleaming linden
it would be like 2-3c in pyth
gleaming linden
yup same
turbid glade
@gleaming linden samething again
gleaming linden
yup
turbid glade
print(1000-int(input())*2)
gleaming linden
can't do 1e3 because floats
turbid glade
ya
gleaming linden
are there any harder ones?
gleaming linden
ah, sad
gleaming linden
oof
gleaming linden
how about a normal one then?
versed eagle
https://www.codingame.com/clashofcode/clash/29710208b56d1bcb53d09880dfd00d2ec114...
what's this
turbid glade
what's this
clash of code
versed eagle
why does co_consts contain None when None isn't used in the function?
seems like a waste of space
!e e.g.
f = lambda: "not None"
print(f.__code__.co_consts)
@versed eagle :white_check_mark: Your 3.11 eval job has completed with return code 0.
(None, 'not None')
I suspect so the compiler doesn't need to figure out all paths where you implicitly return
versed eagle
ah
fair
however, for a lambda that shouldnt be necessary?
since afaik lambdas have no way of implicitly returning
dry mirage
just not super worth it in general, extra None reference is just 8 bytes
calm loom
!e
def test():
return "hi"
print(test.__code__.co_consts)
@calm loom :white_check_mark: Your 3.11 eval job has completed with return code 0.
(None, 'hi')
versed eagle
just not super worth it in general, extra None reference is just 8 bytes
for a single function, yeah
calm loom
hm python always inserts None as the default return value whether u explicitly provide a return or not
thats cool
versed eagle
it shouldnt be necessary for lambda functions though
calm loom
yeh
dry mirage
could be possible to remove it for lambdas, yeah
vast wave
why does `co_consts` contain `None` when `None` isn't used in the function?
funny reason
wait nvm im thinking of the wrong null
this is why
im sure none finds a way into lambdas as well
wait nevermind lambdas don't have a none anywhere
you're right
weird
dry mirage
!e
from einspect import view
f = lambda: "not none"
view(f.__code__.co_consts).remove(None)
print(f())
print(f.__code__.co_consts)
@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | not none
002 | ('not none',)
apparently still works if you remove the None from the tuple
dry mirage
hm?
vast wave
!e ```py
from einspect import view
f = lambda: "not none"
view(f.code.co_consts).remove(None)
print(eval(f.code))
print(f.code.co_consts)
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | not none
002 | ('not none',)
the fuck
yeah no that shouldnt work
!e ```py
from einspect import view
f = lambda: "not none"
target = f.code.replace(co_consts=("not none",))
print(target.co_consts)
print(eval(target))
@vast wave :x: Your 3.11 eval job has completed with return code 139 (SIGSEGV).
('not none',)
yeah
pretty sure that the fields on a code object are just for show, the actual value is cached
.replace() makes a new code object with the given params tho, that's why the previous one didn't segfault but this one did
dry mirage
pretty sure that the fields on a code object are just for show, the actual value...
oh wait no
the None did get removed, it still works because it's accessing the memory at index 1 which isn't freed
I guess there is no check for the size of co_const and they directly access the memory at index 1
so if we try to access None in that function it will be the string at index 0
earnest wing
yeah no that shouldnt work
undefined behavior 
dry mirage
!e
from einspect import view
f = lambda: "not none" + None
view(f.__code__.co_consts).remove(None)
print(f())
print(f.__code__.co_consts)
@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | not nonenot none
002 | ('not none',)
might just be accessing the memory that used to be there
vast wave
might just be accessing the memory that used to be there
i guess
using einspect might what fucks this one up
directly killing the pointer to None but not killing the reference to "not none" still being where it was before
dry mirage
using einspect might what fucks this one up
well, I follow what list does, when you pop items it doesn't deallocate the memory
so you can technically unpop a list by reverting ob_size back
!e
from einspect import view
ls = [1, 2, 3]
ls.pop()
ls.pop()
print(ls)
with view(ls).unsafe() as v:
v.size += 2
print(ls)
@dry mirage :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | [1]
002 | [1, 2, 3]
!e ```py
import base64 as b
import zlib, dis
exec(zlib.decompress(b.a85decode(b'GhRjND/\,^&H88.@I[c#g,'7UKdK\B9p>8Yp5L6=4s-o]O@6?^;"ltC^2@u>m8>1kND9D(SAsPCO=*&dQJm'Y1jFiNQJJE=3U%mZu"Ci:?nII@+$fL8M,4)X=Ii5@AB%?)"iqga95;[4R/>D4lF).3IC2#75AstE59.#=+N"[DC\\]e&3[>!\\IZA%9f!k,)Q#g4J@AYBB&*Zs3:RT1i9>V"=LThTt)"e_hnH1No:8\'3rSEbc^8eGP*\'/S=(/D4g=1="&l;-a%QQK8:LJ^5.Qh:^F:Z=Vds/Mdi7oBGpm0nh,A+5bu=>7^BK'Psr7hL0PQ0,>"m&^Q"pIPB6@k-^%9eg!<8O]]>jBMeW^jPdaa".5ZG^CfUBPX1%b6<#$2a[3C+AkWS8R2nr%bf-p#b4A;Gi]kZUr#5#X#s^"Ja$Xh[I;Nn4EEra5,M2T"]!.Wn^H+:23'l-0um&Pmu"7AhC/bNBMP/g?1TGH2ho0hneV5$&Co<JR$TLal-W&e\U+sHV<6n=hqW;8KHOMlRB\R9!b;;p2<RuY@6?::usu@CN9F-3R]n'8P2\/:,XQKk6.MLdt@9DZU;V);8Tlhn8>/bE+Ra;B9et_)@hFTuK1RP;bG0fp,pD_[6l(rU;Li1QA.UB%d!i;!0(nHkDuJ1,LR(r\8KVQ[Lg?9Y=:0t::,P"PJMVQik:=7\1D;@.tn3:C^nP20@YCUsTD1gB$e0Qd7gW4I\LrHJkPD'@]-MSOkbT'eNKh4G#U@[XT%GjDNh7r[n#tDXa!n3ojE'^HDF&gQNHs=rH3K1Qle3+F.6/is=8U(i@R&t[(E8HUEDVjQT6+!-_Ci1-Un;#Z/ALfB=RFnQ>"&Hh&7@=5raeikW/&61!hi0=TLZX8&W)c`),')A(#VF(o5ku:n)FHV&q"Z)f]V;B_>24^U"I%7#"16n8eXj+b2?M(^T&iJ5-@*c"@HS4i8U%kS".pqkjjK@.66JL-K\q-V4')))
a = Assembler()
nc = a.consts_create_or_get
i = a.insn
i("RESUME",0)
i("LOAD_CONST", nc("not null"))
i("RETURN_VALUE")
co = a.pack_code_object()
print(co.co_consts)
dis.dis(co)
print(eval(co))
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | ('not null',)
002 | 0 RESUME 0
003 | 2 LOAD_CONST 0 ('not null')
004 | 4 RETURN_VALUE
005 | not null
its possible to not have none in it
dry mirage
well yeah if you manually provide the bytecode
but it might negatively affect speed to have that work in the compiler
vast wave
at that point just accept that its 1 byte
8 bits
no wait im stupid
fuck i need to retake comp sci
dry mirage
I suppose it could be 8 bits if we have 8 bit python ๐
sudden osprey
why does `co_consts` contain `None` when `None` isn't used in the function?
IIRC the first value in co_consts is the docstring
!e
def x():
"""test"""
print(x.__code__.co_consts)
@sudden osprey :white_check_mark: Your 3.11 eval job has completed with return code 0.
('test', None)
Not sure
versed eagle
if it's None, it just shouldnt be present
sudden osprey
if it's `None`, it just shouldnt be present
It seems like the function object relies on it being there to set it's __doc__ https://github.com/python/cpython/blob/0a22aa0528a4ff590854996b8854e9a79120987a/Objects/funcobject.c#L157
night quarryBOT
Objects/funcobject.c line 157
PyObject *consts = code_obj->co_consts;```Sort of makes sense, and I guess a slight optimization over a separate field on the code object since it can be shared if the func uses none
sudden osprey
that shouldn't be needed for lambdas though
I guess, yeah, although afaict lambda functions aren't really special cased so the actual objects are the same as normal function/code objects.
Is there even any way of telling the difference other than the name 
versed eagle
I guess, yeah, although afaict lambda functions aren't really special cased so t...
yeah you're right
quartz wave
if it's `None`, it just shouldnt be present
that's fixed in some version in 3.12
versed eagle
oh nice
versed eagle
it doesn't actually remove the first thing
The compiler now removes all unused constants from code objects (except the first one, which may be a docstring).
quartz wave
ok maybe not
i'm gonna go test
yep
i'm gonna add a path in my modified cpython for that
fixed ```diff
-
// The first constant may be docstring; keep it always. -
index_map[0] = 0;
-
// If it's a docstring, keep it. -
index_map[0] = PyUnicode_Check(PyList_GET_ITEM(consts, 0)) ? 0 : -1;
>>> from dis import dis
>>> dis("lambda:2")
0 0 RESUME 0
1 2 LOAD_CONST 0 (<code object <lambda> at 0x000001C93A4F7220, file "<dis>", line 1>)
4 MAKE_FUNCTION 0
6 RETURN_VALUE
Disassembly of <code object <lambda> at 0x000001C93A4F7220, file "<dis>", line 1>:
1 0 RESUME 0
2 LOAD_CONST 0 (2)
4 RETURN_VALUE
Won't that break docstrings? When the function object is created how does it know if the first value is the docstring or not?
versed eagle
PyUnicode_Check checks if it's a string
quartz wave
Won't that break docstrings? When the function object is created how does it kno...
doesn't seem to break ```pycon
def abc():
... """
... doc
... """
...
abc.code.co_consts
('\n\tdoc\n\t', None)
If there's no docstring, but you return a string
Won't that string be set as the docstring?
kindred stone
getting started on esoteric, its pretty fun ๐
!e
f = lambda c, n: (lambda np: (exec(compile(c, "<string>", "exec"), np), np[n]))({})[1]
code = """
def fn():
print("hello, world")
"""
fn = f(code, "fn")
fn()```@kindred stone :white_check_mark: Your 3.11 eval job has completed with return code 0.
hello, world
@sick hound :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | Hello world lol
002 | 12345
gusty dust
a = 2
b = 3
c = 4
a, b, c *= 2 # a => 4, b => 6, c => 8
Cell In[4], line 1
a, b, c *= 2 # a => 4, b => 6, c => 8
^
SyntaxError: 'tuple' is an illegal expression for augmented assignment
the fact that this fails makes me sad
i get why but
limpid spruce
???!
languid hare
```py
a = 2
b = 3
c = 4
a, b, c *= 2 # a => 4, b => 6, c => 8
Cell In[4], lin...
0 size augmented assignment 
!e
, *= 2
@languid hare :x: Your 3.11 eval job has completed with return code 1.
001 | File "/home/main.py", line 1
002 | , *= 2
003 | ^
004 | SyntaxError: invalid syntax
@languid hare :x: Your 3.11 eval job has completed with return code 1.
001 | File "/home/main.py", line 1
002 | (,) *= 2
003 | ^
004 | SyntaxError: invalid syntax
night quarryBOT
@languid hare :x: Your 3.11 eval job has completed with return code 1.
001 | File "/home/main.py", line 1
002 | () *= 2
003 | ^^
004 | SyntaxError: 'tuple' is an illegal expression for augmented assignment
ok fair enough
vast wave
It seems like the function object relies on it being there to set it's `__doc__`...
!e so that means ```py
from einspect import view
a = lambda: "big boobs"
v = view(a.code.co_consts)
del v[:]
v.extend(["the cat", "big boobs"])
print(a.code.co_consts)
print(a.doc)
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
None
no it doesn't
sadge
!e ```py
from einspect import view
a = lambda: "big boobs"
v = view(a.code.co_consts)
del v[:]
v.extend(["the cat", "big boobs"])
print(a.code.co_consts)
print(a.doc)
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | ('the cat', 'big boobs')
002 | None
.__doc__ is not dynamic
fleet bridge
Yes. It is set once at function creation time, and you can set it to other value after function is created
kindred stone
def create_function(code_string: str, var_to_extract: str):
code = compile(code_string, "<string>", "exec")
namespace = {}
exec(code, namespace)
return namespace[var_to_extract]
its just a different way of executing code, similar to eval
sick hound
Hi there it's my first time posting on here but I was wondering if someone could help me with Pytorch specifically
lunar marsh
Esoteric Python request:
Find a way to override the assignment operator (=) by any means neccesary. Be it modification of the locals dictionary or some runtime hooking. Preferably portable (i.e work when passed between function calls and stack frames). By overriding I mean detecting an assignment and even prevent it, so for example
x=5
# override
x=4
print(x) # -> 5
!e ```py
x = 4
ั
= 5
print(x)
@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.
4
easy
lunar marsh
sneaky boi
meager zinc
!e ```py
x = 4
ั
= 5
print(x)
```
(legit)
versed eagle
Esoteric Python request:
Find a way to override the assignment operator (=) by a...
alr give me like an hour
ill be able to get on my computer then
i don't wanna type on mobile
meager zinc
from types import CodeType
from sys import _getframe, exit
import dis
orig = _getframe().f_code
code = orig.co_code
payload = list(code)
allow_assign = True
for idx, op in enumerate(code):
if op == dis.opmap["STORE_NAME"] and orig.co_names[code[idx+1]] == "x":
if allow_assign:
allow_assign = False
else:
payload[idx] = dis.opmap["POP_TOP"]
exec(orig.replace(co_code=bytes(payload[176:])))
exit(0)
# Everything after this is the input code
x = 5
x = 4
print(x)
Works on my machine
@lunar marsh
lunar marsh
```py
from types import CodeType
from sys import _getframe, exit
import dis
ori...
That works, could adapt it into a tracer function to make it work between frames I suppose.
dry mirage
Esoteric Python request:
Find a way to override the assignment operator (=) by a...
!e
import ast
import inspect
import sys
from functools import cache
class Final:
def __new__(cls, obj):
stack = inspect.stack(1)
c_stack = stack[1:]
c_frame_info = c_stack[0]
c_frame = c_frame_info.frame
def _trace(frame, _event, _arg= None):
info = inspect.getframeinfo(frame)
node = ast.parse(info.code_context[0].strip())
if isinstance(node.body[0], ast.Assign):
raise SyntaxError(f"cannot assign to final variable {obj!r}")
return _trace
cls._enable_trace()
c_frame.f_trace = _trace
return obj
@cache
@staticmethod
def _enable_trace() -> None:
if not sys.gettrace():
sys.settrace(lambda *args: None)
x = Final(100)
print(x)
x = 5
print(x)
@dry mirage :x: Your 3.11 eval job has completed with return code 1.
001 | 100
002 | Traceback (most recent call last):
003 | File "/home/main.py", line 32, in <module>
004 | x = 5
005 | ^
006 | File "/home/main.py", line 32, in <module>
007 | x = 5
008 | ^
009 | File "/home/main.py", line 17, in _trace
010 | raise SyntaxError(f"cannot assign to final variable {obj!r}")
011 | SyntaxError: cannot assign to final variable 100
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/otigivotuw.txt?noredirect
lunar marsh
```py
from types import CodeType
from sys import _getframe, exit
import dis
ori...
(Ignoring the disregard for instruction opcodes and arguments >:( )
lunar marsh
!e
```py
import ast
import inspect
import sys
from functools import cache
class...
I like it, have not seen code_context before
dry mirage
you also have access to the caller dictionary
so you can check if x is bound to anything else via eval exec globals() and error as well
lunar marsh
!e
```py
import ast
import inspect
import sys
from functools import cache
class...
Though this will fail when ran in the REPL right?
meager zinc
the f_trace usage is pretty smart
I've seen it before in my testing but never really explored it
dry mirage
Though this will fail when ran in the REPL right?
yeah, none of the source inspect works in repl
rugged sparrow
!e ```py
import sys
class Final:
slots = ['value']
def init(self, value):
self.value = value
frames = {}
def tracefunc(frame, event, arg):
for dct_name in ['f_globals', 'f_locals']:
mp = getattr(frame, dct_name)
for key, value in mp.items():
if isinstance(value, Final) and hasattr(value, 'value'):
frames.setdefault(frame, {}).setdefault(dct_name, {})[key] = value.value
for key, value in frames.get(frame, {}).get(dct_name, {}).items():
mp[key] = value
return tracefunc
sys.settrace(tracefunc)
def foo():
x = Final(100)
print(x, type(x))
x = 1
print(x, type(x))
foo()
@rugged sparrow :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 100 <class 'int'>
002 | 100 <class 'int'>
doesn't seem to want to work for global scope but it works for locals
dry mirage
doesn't seem to want to work for global scope but it works for locals
you still set f_locals for the module frame iirc
radiant anchor
fizzbuzz = lambda x: [
div3 := x % 3 == 0,
div5 := x % 5 == 0,
result := "",
div3 and [
result := result + "fizz",
],
div5 and [
result := result + "buzz",
],
result or [
result := str(x),
],
result,
][-1]
for i in range(1, 101):
print(fizzbuzz(i))
proof that python supports braces instead of indentation, and that lambdas allow multiple lines :P
(you can also use set notation for the "if statements", as long your results are all hashable)
night quarryBOT
@fleet bridge :x: Your 3.11 eval job has completed with return code 1.
001 | File "/home/main.py", line 1
002 | from __future__ import braces
003 | ^
004 | SyntaxError: not a chance
vast wave
of python supporting brace syntax
!e ```py
if 1 == 1: {
print("wrong")
}
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
wrong
!e
if 1 == 1: {
right := {},
print(right)
}
@radiant anchor :x: Your 3.11 eval job has completed with return code 1.
001 | {}
002 | Traceback (most recent call last):
003 | File "/home/main.py", line 1, in <module>
004 | if 1 == 1: {
005 | ^
006 | TypeError: unhashable type: 'dict'
!e ```py
if 1 == 1: {
right = {}
print(right)
}
@vast wave :x: Your 3.11 eval job has completed with return code 1.
001 | File "/home/main.py", line 2
002 | right = {}
003 | ^^^^^^^^^^
004 | SyntaxError: invalid syntax. Maybe you meant '==' or ':=' instead of '='?
oh right
radiant anchor
(you can also use set notation for the "if statements", as long your results are...
.
quartz wave
```python
fizzbuzz = lambda x: [
div3 := x % 3 == 0,
div5 := x % 5 == 0,...
!e ```py
fizzbuzz = lambda x: {
div3 := x % 3 == 0,
div5 := x % 5 == 0,
result := "",
div3 and (
result := result + "fizz",
),
div5 and (
result := result + "buzz",
),
result or (
result := str(x),
),
result,
}.pop()
for i in range(1, 101):
print(fizzbuzz(i))
@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | False
002 | False
003 | False
004 | False
005 | False
006 | False
007 | False
008 | False
009 | False
010 | False
011 | False
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/katoxoliwa.txt?noredirect
versed eagle
you could use a dict
radiant anchor
you could have a dict with "line numbers"
versed eagle
hashable_thingy:
something,
hashable_thingy:
something_else,
}```something like that
could be done
as long as you keep even numbers of stuff
and have every other line be hashable
shouldnt be too hard, yeah?
vast wave
!e ```py
@lambda :()
class a:pass
if 1 == 1: {
a:"something",
a:print("something else")
}
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
something else
signal swallow
!e
radiant anchor
!e ```py
@lambda _:_()
class a:pass
if 1 == 1: {
a:"something",
a:print("som...
!e
if 1 == 1: {
...:"something",
...:print("something else")
}
@radiant anchor :white_check_mark: Your 3.11 eval job has completed with return code 0.
something else
mystic hull
i1 = [False, True, False, False, True, True, False, True, False, True, True, True, False, False, True, False, False, False, True, False, True, True, True, False, False, False, True, False, False, False, False, False, False, True, False, True, False, True, True, True, False, True, True, False, True, False, False, True, False, True, True, True, False, False, False, False, False, True, True, True, False, False, False, False]
i111 = ""
for i in i1: i2 = int(i); i111 += str(i2)
i11 = int(i111, 2);i1111 = i11.bit_length() + 7 // 8;i111111 = i11.to_bytes(i1111, "big");i11111 = i111111.decode();print(i11111)
!e
i1 = [False, True, False, False, True, True, False, True, False, True, True, True, False, False, True, False, False, False, True, False, True, True, True, False, False, False, True, False, False, False, False, False, False, True, False, True, False, True, True, True, False, True, True, False, True, False, False, True, False, True, True, True, False, False, False, False, False, True, True, True, False, False, False, False]
i111 = ""
for i in i1: i2 = int(i); i111 += str(i2)
i11 = int(i111, 2);i1111 = i11.bit_length() + 7 // 8;i111111 = i11.to_bytes(i1111, "big");i11111 = i111111.decode();print(i11111)
@mystic hull :white_check_mark: Your 3.11 eval job has completed with return code 0.
๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝ๏ฟฝMr. Wipp
lmao finally
night quarryBOT
:incoming_envelope: :ok_hand: applied mute to @sick hound until <t:1679246074:f> (10 minutes) (reason: duplicates rule: sent 4 duplicated messages in 10s).
The <@&831776746206265384> have been alerted for review.
calm loom
Ayyy
heavy ermine
!e import math; import numpy; sine = lambda from_, to_, amp_, step_: (math.sin(s * amp_) for s in numpy.arange(from_, to_, step_)); square = lambda from_, to_, amp_, step_: ((-1 if c <= 0 else 1) for c in sine(from_, to_, amp_, step_)); print(list(square(0, 4, 1, 0.1)));
night quarryBOT
@heavy ermine :white_check_mark: Your 3.11 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]
python one-liner folks
eager ore
(
lambda math, numpy: ...
)(
__import__("math"),
__import__("numpy")
)
@heavy ermine
heavy ermine
```py
(
lambda math, numpy: ...
)(
__import__("math"),
__import__("n...
what this for
eager ore
to avoid using semicolons
heavy ermine
uhhhh why lambda math, numpy: ...
eager ore
the math and numpy arguments are passed to your code, the ... part
then you can truly make it one line
heavy ermine
why just not use __import__
eager ore
it would save characters if you were using it a lot
heavy ermine
ok????? how so
eager ore
because it's easier to type math instead of typing __import__("math") each time
you can shorten it to just m
heavy ermine
import math as m;import numpy as n;...
that's not one line though
heavy ermine
technically it is. there are no \n in this code
eager ore
that's cheating
heavy ermine
its not, technically
eager ore
in esoteric python it is
versed eagle
because it's easier to type `math` instead of typing `__import__("math")` each t...
or, here me out,
(np:=__import__("numpy"),math:=__import__("math"))
somewhere in the beginning
of the function
low lynx
just have them as default arguments in the lambda
!e
f=lambda a, np=__import__('numpy'): np.array(a).T
print(f([[0, 1, 2], [3, 4, 5]]))
@low lynx :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | [[0 3]
002 | [1 4]
003 | [2 5]]
bronze agate
!e
sum(print("test") or 0for _ in range(3))```@bronze agate :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | /home/main.py:1: SyntaxWarning: invalid decimal literal
002 | sum(print("test") or 0for _ in range(3))
003 | test
004 | test
005 | test
!e py foo = lambda n:print(sum(print(x,end='+='[i==len(n)-1])or x for i,x in enumerate(n))) foo([1,2,3]) foo([4,5,6])
night quarryBOT
@last locust :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 1+2+3=6
002 | 4+5+6=15
Heh, neat
quartz wave
!e ```py
foo = lambda n:print('+'.join(map(str,n))+'=%d'%sum(n))
foo([1,2,3])
foo([4,5,6])
@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 1+2+3=6
002 | 4+5+6=15
ok
last locust
!e ```py
foo = lambda n:print('+'.join(map(str,n))+'=%d'%sum(n))
foo([1,2,3])
fo...
I was expanding on the one above though
!e py foo = lambda n:print(sum(print(x,end='=+'[i<len(n)-1])or x for i,x in enumerate(n))) foo([1,2,3]) foo([4,5,6])
night quarryBOT
@last locust :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 1+2+3=6
002 | 4+5+6=15
That's 1 char shorter by swapping the += to =+ and then i<len(n)-1 instead of i==len(n)-1
quartz wave
ok
fleet bridge
!e ```py
foo = lambda n:print('+'.join(map(str,n))+'=%d'%sum(n))
foo([1,2,3])
fo...
I believe using fstring at the end would be 1 char shorter
versed eagle
you can save 2 chars per invocation by using *args rather than taking a list
versed eagle
I believe using fstring at the end would be 1 char shorter
same number of chars
+"=%d"%sum(n)
+f"={sum(n)}"
!e
f = lambda n: print(f"{'+'.join(map(str, n))}={sum(n)}")
f([1,2,3])
f([4,5,6])
@sick hound :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 1+2+3=6
002 | 4+5+6=15
pure dew
!e ```py
f = lambda n:print(eval(f"f'{{{'+'.join(map(str,n))}=}}'"))
f([4,5,6])
f([1,2,3,4,5,6,7,8,9])
@pure dew :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 4+5+6=15
002 | 1+2+3+4+5+6+7+8+9=45
a lil bit longer but way cooler
past plank
!e you guys are way overcomplicating it ```py
def f(n):print('+'.join(map(str,n))+f'={sum(n)}')
f([4,5,6])
f([1,2,3,4,5,6,7,8,9])```
night quarryBOT
@past plank :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 4+5+6=15
002 | 1+2+3+4+5+6+7+8+9=45
f=lambda n:print(eval(f"f'{{{'+'.join(map(str,n))}=}}'")) # 57
f=lambda n:print(f"{'+'.join(map(str, n))}={sum(n)}") # 53
def f(n):print('+'.join(map(str,n))+f'={sum(n)}') # 49```it's pointless to use a lambda if you don't need to return anything py f=lambda a,b: def f(a,b):
aside from the function definition though, this improvement is obvious py print(f"{'+'.join(map(str, n))}={sum(n)}") print('+'.join(map(str,n))+f'={sum(n)}')
versed eagle
aside from the function definition though, this improvement is obvious ```py
pri...
we've already had something of similar length for a while now?
print(f"{'+'.join(map(str, n))}={sum(n)}")
print('+'.join(map(str,n))+f'={sum(n)}')
print('+'.join(map(str,n))+'=%d'%sum(n))
versed eagle
!e ```py
foo = lambda n:print('+'.join(map(str,n))+'=%d'%sum(n))
foo([1,2,3])
fo...
the bottom one is from this
you're reinventing the wheel :P
meager zinc
Adds two spaces, but removes two chars:
print('+'.join(map(str,n))+f'={sum(n)}')
print('+'.join(map(str,n)),'=',sum(n))
e.g.
1+1+1=3
1+1+1 = 3
n=[1,2,3]
print(*n,sep='+',end=f'={sum(n)}')
!e ```py
n=[1,2,3]
print(*n,sep='+',end=f'={sum(n)}')
@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.
1+2+3=6
Is it shorter?
versed eagle
mhm
night quarryBOT
@sick hound :white_check_mark: Your 3.11 eval job has completed with return code 0.
1+2+3=6
finite blaze
Hey,
I have a little golfing challenge.
You have a file with n lines and each line contains one word of random length with random letters.
Distance between letters A and B is 1, between C and F its 3 etc etc.
List words where distance between all the letters in the word is less than 10.
Example of a word that matches this condition is AHBINK
low lynx
119
import itertools as i
p=i.pairwise
for w in open(0).read().split():all(-10<ord(i)-ord(j)<10for i,j in p(w))and print(w)
92
for w in open(0).read().split():all(-10<ord(i)-ord(j)<10for i,j in zip(w,w[1:]))and print(w)
does this work? ```py
for w in open(0):p,*a,_=w;10<max((abs(ord(p)-ord(c)),p:=c)[0]for c in a)or print(w)
yes
both 76
for w in open(0):p,*a,_=w;all(-10<ord(p)-ord(p:=c)<10for c in a)and print(w)
for w in open(0):p,*a,_=w;10<max(abs(ord(p)-ord(p:=c))for c in a)or print(w)
actually no
your's is wrong
which is a good thing
because it saves a char
75
for w in open(0):p,*a,_=w;9<max(abs(ord(p)-ord(p:=c))for c in a)or print(w)
tied 75
for w in open(0):p,*a,_=w;max(abs(ord(p)-ord(p:=c))for c in a)<10==print(w)
@sly ibex :white_check_mark: Your 3.11 eval job has completed with return code 0.
-10 0 10 20 30
wouldn't they have the same lengthdef a(a) be shorter than lambda a ?
you'd also remove the b =
kinda got offtopic here sorry```hs
ghci> t = [123, 123, 893]
ghci> (\x -> sum x)(t)
1139
```py
>>> t = [123, 123, 893]
>>> (lambda x: sum(x))(t)
1139
??
maiden blaze
75
```py
for w in open(0):p,*a,_=w;9<max(abs(ord(p)-ord(p:=c))for c in a)or prin...
foo or print(w) -> [foo,print(w)]?
kindred stone
~~wouldn't `def a(a)` be shorter than `lambda a` ?~~ they have the same length
you can't call the function on the same line as it was defined without semicolons
past plank
tied 75
```py
for w in open(0):p,*a,_=w;max(abs(ord(p)-ord(p:=c))for c in a)<10=...
Not shorter, but thought I'd share my entry anyway
79
for w in open(0):any(abs(ord(a)-ord(b))>9for a,b in zip(w,w[1:-1]))or print(w)```foo is boolean, the short circuiting is necessary
quartz wave
tied 75
```py
for w in open(0):p,*a,_=w;max(abs(ord(p)-ord(p:=c))for c in a)<10=...
i have 2 ways to make this shorter but i'm currently testing one
nvm they both don't work
ancient saddle
ok this is kind of a crapshoot, but is there a semi-reasonable way to find where a certain string is being stored in what seems to be an absurdly convoluted and obfuscated data structure?
I've gone about trying to manually set a certain configuration for a certain class and it's apparently cached elsewhere but I can find absolutely nothing helpful in this descent into madness that some might generously call a module
quartz wave
ok this is kind of a crapshoot, but is there a semi-reasonable way to find where...
um yes
search the memory
if it's interned you might have better luck with gc.get_referrers()
ancient saddle
I mean, I can def find its physical memory location with scanner tools but I've no clue how to get from the memory location to which object actually owns that data
quartz wave
I mean, I can def find its physical memory location with scanner tools but I've ...
find the physical memory location, subtract a specific amount of bytes, use ctypes to get the str object, then gc.get_referrers() if the string is interned
that is if it's a str
for 64-bit systems, subtract 48 bytes for ASCII strings and 72 for non-ASCII strings in 3.10(?)+
ancient saddle
seems like I'm gettin a lot of Nones back, hmm
cunning stream
not python but i think you guys will like this language
versed eagle
reminds me of perl-
kind bronze
not python but i think you guys will like this language
bfasm ?
arctic skiff
its like brainfuck but slightly more usable
it should be named brainfuckasactualprogramminglanguage
bfaapl
sick hound
!e def function(c):
translates = {49:11036,48:11035}
for i in c:
fill = []
for j in c:
fill.append(len(bin(ord(j))[1:]))
fill = max(fill)
i = bin(ord(i))[2:]
i = i.zfill(fill)
i = i.translate(translates)
print(i)
_('<~ร รพรพ~ff', 2)
function('<~ร รพรพ~ff')
night quarryBOT
@sick hound :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
002 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
003 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
004 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
005 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
006 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
007 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
008 | โฌโฌโฌโฌโฌโฌโฌโฌโฌ
@thorny tapir :warning: Your 3.11 eval job has completed with return code 0.
[No output]
@thorny tapir :warning: Your 3.11 eval job has completed with return code 0.
[No output]
!e ```py
import sys
f = open("1.txt", 'w')
sys.stdout = f
print("Hello world")
@gleaming linden :white_check_mark: Your 3.11 eval job has completed with return code 0.
1.txt
Hello world
!e or this: ```py
from contextlib import redirect_stdout
f = open("1.txt", 'w')
with redirect_stdout(f):
print("Hello")
print("World")
@gleaming linden :white_check_mark: Your 3.11 eval job has completed with return code 0.
World
1.txt
Hello
old socket
!e ```py
import sys
with open("1.txt", "w") as sys.stdout:
print("Hello")
print("World")
@old socket :x: Your 3.11 eval job has completed with return code 1.
001 | Traceback (most recent call last):
002 | File "/home/main.py", line 5, in <module>
003 | print("World")
004 | ValueError: I/O operation on closed file.
1.txt
Hello
oh I thought it worked before
Unless it's just cus of snekbox
Oh wait it did write but there is an error
gleaming linden
it works, but you can't print anything afterward because the file is closed when you exit the with block
gleaming linden
!e ```py
from contextlib import redirect_stdout
with open('1.txt', 'w') as file, redirect_stdout(file):
print("Hello")
print("World")
@gleaming linden :white_check_mark: Your 3.11 eval job has completed with return code 0.
World
1.txt
Hello
oh, this works too
old socket
I know you can unpack with ctx manager but I thought the enter specifically needed to return something unpackable
!e ```py
with open("1.txt", "w") as foo, bar:
print(foo, bar)
@old socket :x: Your 3.11 eval job has completed with return code 1.
001 | Traceback (most recent call last):
002 | File "/home/main.py", line 1, in <module>
003 | with open("1.txt", "w") as foo, bar:
004 | ^^^
005 | NameError: name 'bar' is not defined
1.txt
[Empty]
it's doing with (open("1.txt", "w") as foo), bar:
old socket
Ahh okay
gleaming linden
!e ```py
with open('1.txt', 'w') as file:
file.write("foo\nbar\n")
with open('1.txt') as (a, b):
print(a, b)
@gleaming linden :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | foo
002 | bar
1.txt
foo
bar
cool
old socket
I wonder how secure the file stuff is on snekbox
I know there prob isn't any huge issue off the bat but I can't help but feel there is something
versed eagle
!e
open(1, "w").write("hello, world")
@versed eagle :white_check_mark: Your 3.11 eval job has completed with return code 0.
hello, world
1 is the file descriptor for stdout
its not that they could use it as stdout, it's that it is stdout
pure dew
!e ```py
open("/dev/stdout", "w").write("dlrow ,olleh")
@pure dew :x: Your 3.11 eval job has completed with return code 1.
001 | Traceback (most recent call last):
002 | File "/home/main.py", line 1, in <module>
003 | open("/dev/stdout", "w").write("dlrow ,olleh")
004 | ^^^^^^^^^^^^^^^^^^^^^^^^
005 | FileNotFoundError: [Errno 2] No such file or directory: '/dev/stdout'
dry mirage
!e ```py
open("/dev/stdout", "w").write("dlrow ,olleh")
```
/dev is not mounted in snekbox
fleet bridge
<@726368249666732074> :white_check_mark: Your 3.11 eval job has completed with r...
Since when this feature does exist?
dry mirage
Since when this feature does exist?
around 2 weeks ago snekbox#159
glass drumBOT
fleet bridge
That's awesome
arctic skiff
around 2 weeks ago snekbox#159
Why there was not announcement about this new feature?
dry mirage
Why there was not announcement about this new feature?
there's this one? 
#changelog message
arctic skiff
there's this one? <:pithink:652247559909277706> https://canary.discord.com/chan...
Totally forgot about that
sonic birch
!e ```py
with open('1.txt', 'w') as file:
file.write("foo\nbar\n")
with ope...
!e open('file.txt', 'w').write('1\n2\n3\n4')
arr=[0]*10
with open('file.txt') as (arr[0], *arr[5:7], arr[-1]):
print(arr)
night quarryBOT
@sonic birch :white_check_mark: Your 3.11 eval job has completed with return code 0.
['1\n', 0, 0, 0, 0, '2\n', '3\n', 0, 0, '4']
file.txt
1
2
3
4
smh i'll get it
quartz wave
!e ```py
open('file.txt', 'w').write('1\n2\n3\n4')
arr=[0]*10
with open('file.txt') as (arr[0],*arr[5:7], arr[-1]):
print(arr)
@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
['1\n', 0, 0, 0, 0, '2\n', '3\n', 0, 0, '4']
file.txt
1
2
3
4
!e open('file.txt', 'w').write('1\n2\n3\n4')
arr=[0]
with open('file.txt') as (arr[0],*arr[5:7], arr[-1]):
print(arr)
night quarryBOT
@pure dew :white_check_mark: Your 3.11 eval job has completed with return code 0.
['1\n', '2\n', '4']
file.txt
1
2
3
4
!e open('file.txt', 'w').write('1\n2\n3\n4')
arr=[]
with open('file.txt') as (arr[0],*arr[1:3], arr[-1]):
print(arr)
night quarryBOT
@pure dew :x: Your 3.11 eval job has completed with return code 1.
001 | Traceback (most recent call last):
002 | File "/home/main.py", line 3, in <module>
003 | with open('file.txt') as (arr[0],*arr[1:3], arr[-1]):
004 | ~~~^^^
005 | IndexError: list assignment index out of range
file.txt
1
2
3
4
ah
cloud fossil
๐ฅด
I'm assuming the first one is for lambdas, but the second one lies to me as I can get 2 (so there is something). The last one is what I would expect from both except the NOP which I didn't know is used
The last one is kind of what happen if I throw these into an actual function
(minus the NOP in the first two cases)
versed eagle
the second one is the same as the third one, except for the NOP
which makes sense, considering that they do the same thing
the NOP i assume is an artifact of the separate lines
cloud fossil
But why does the byte code tell me there is no constant loading when I can load 2
I mean
_ becomes 2
Is it just a REPL thing
versed eagle
iirc that's done by sys.displayhook
versed eagle
Is it just a REPL thing
yeah
fleet bridge
But why does the byte code tell me there is no constant loading when I can load ...
Code in repl is compiled in its special mode
!e ```py
from dis import dis
c = compile('1;2', '', mode='single')
dis(c)
@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 0 0 RESUME 0
002 |
003 | 1 2 LOAD_CONST 0 (1)
004 | 4 PRINT_EXPR
005 | 6 LOAD_CONST 1 (2)
006 | 8 PRINT_EXPR
007 | 10 LOAD_CONST 2 (None)
008 | 12 RETURN_VALUE
Cool!
And on another note, can someone explain me how the reference counting system works in Python?
The website talks about strong references and borrowed references
fleet bridge
You can read developers guide about that
fleet bridge
The website talks about strong references and borrowed references
There is some introduction to capi where these words are explained
cloud fossil
Strong references are references that if are gone, delete the object
Borrowed just means the object's strong reference is borrowed? What count is being incremented/decremented then?
versed eagle
the objects ob_refcnt
fleet bridge
Strong references are references that if are gone, delete the object
Thats not about destroying object. It is about when you should do PyINCREF(x) and when you shouldn't
night quarryBOT
Include/object.h lines 102 to 106
struct _object {
_PyObject_HEAD_EXTRA
Py_ssize_t ob_refcnt;
PyTypeObject *ob_type;
};```At which point does the virtual machine stop to take a look at objects' ob_refcnt?
versed eagle
that's the garbage collectors job
cloud fossil
When does it collect garbage
fleet bridge
When does it collect garbage
When .ob_type hits 0
When GC is triggered
versed eagle
you could try reading the source to figure that out
cloud fossil
i know it does it when leaving scopes
What if I have something like
for _ in range(1_000_000):
f()
It is not. GC isn't doing anything at this moment
cloud fossil
When .ob_type hits 0
When GC is triggered
But objects begin with a reference count that is greater than zero?
fleet bridge
What if I have something like
```py
for _ in range(1_000_000):
f()
```
There is no problem.
Assignment to _ increments RC for one object and decrements for another
versed eagle
It is not. GC isn't doing anything at this moment
sorry- i meant the ob_refcnt gets decremented when going out of scope
which means the gc can collect the object if the ob_refcnt is 0
fleet bridge
But objects begin with a reference count that is greater than zero?
Yes, because otherwise it would be weird
Leaving scope is equivalent to deleting all local variables using del x
versed eagle
yeah
cloud fossil
When in something like
print()
print()
does the garbage collector know to dereference the None from the return values
fleet bridge
sorry- i meant the ob_refcnt gets decremented when going out of scope
which mean...
If i am not mistaken, RC==0 is none of GC business.
GC is doing graph analysis, so it is not related to deleting object with RC==0.
This behaviour is inlined (i guess) into Py_DECREF macro that checks if RC==0.
cloud fossil
There is no problem.
Assignment to _ increments RC for one object and decrements...
Care to elaborate? Which object's reference count is incremented and which object's is decremented?
fleet bridge
What if I had functions which returned my own objects
Look at bytecode and realise that by yourself
fleet bridge
Care to elaborate? Which object's reference count is incremented and which objec...
Olds is decremented, news is incremented. Because there is one less reference to old object and one more ref to new object
versed eagle
If i am not mistaken, RC==0 is none of GC business.
GC is doing graph analysis,...
is it? it used to be the gc that handled that
i guess my knowledge about that is a bit behind since i dont really mess around with the gc much
well ty for teaching me that then
cloud fossil
Olds is decremented, news is incremented. Because there is one less reference to...
Which objects are considered old? The ones in the body of the for loop after an iteration?
fleet bridge
Well. It is also possible to think that there are 2 different GCs:
- stupid one: checks RC==0 and deletes object
- smart one: detects dead graphs, destroys references in this graphs, so every RC eventually hits 0, so stupid GC can collect every garbage object
fleet bridge
Which objects are considered old? The ones in the body of the `for` loop after a...
a = old
a = new # here rc of new is incremented and rc of old is decremented
cloud fossil
```py
a = old
a = new # here rc of new is incremented and rc of old is decrement...
Understood
cloud fossil
https://devguide.python.org/internals/garbage-collector/#id1
Thanks
fleet bridge
For loop can be destructured to simpler statements, and there would be an assignment
cloud fossil
Yeah, STORE_NAME ... (_) in my case
fleet bridge
Interesting fact: you can use in for loop anything if you can use it before =
for a[b] in c:... is good, for example
It is called "assignment target", it means you can assign value to it and it wouldn't raise SyntaxError
Assignment target is essentially any expression with some exceptions (you cant assign to function calls, literals, ...)
fleet bridge
You can also `del a[b]`
In this case it is even stricter: you can delete only variable, expression.attr and expression[item]
cloud fossil
!e
from string import ascii_uppercase, ascii_lowercase
a = list(ascii_uppercase)
c = list(ascii_lowercase)
b = 0
print(f"{a = }\n{c = }")
for a[b] in c:
b += 1
print(f"{a = }\n{c = }")
@cloud fossil :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | a = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z']
002 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z']
003 | a = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z']
004 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z']
So cool
versed eagle
you can also do slices
cloud fossil
!e
from string import ascii_uppercase, ascii_lowercase
a = list(ascii_uppercase[:10])
c = list(ascii_lowercase[:10])
b = 0
print(f"{a = }\n{c = }")
for a[b::2] in c:
b += 2
print(f"{a = }\n{c = }")
@cloud fossil :x: Your 3.11 eval job has completed with return code 1.
001 | a = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J']
002 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j']
003 | Traceback (most recent call last):
004 | File "/home/main.py", line 6, in <module>
005 | for a[b::2] in c:
006 | ~^^^^^^
007 | ValueError: attempt to assign sequence of size 1 to extended slice of size 5
a[b] and a[<slice>] are the same, index type is handled inside list.getitem
cloud fossil
Wait do I need to slice c with b as well
versed eagle
since people usually forget
cloud fossil
!e
from string import ascii_uppercase, ascii_lowercase
a = list(ascii_uppercase[:10])
c = list(ascii_lowercase[:10])
b = 0
print(f"{a = }\n{c = }")
for a[b::2] in c[b::2]:
b += 2
print(f"{a = }\n{c = }")
@cloud fossil :x: Your 3.11 eval job has completed with return code 1.
001 | a = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J']
002 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j']
003 | Traceback (most recent call last):
004 | File "/home/main.py", line 6, in <module>
005 | for a[b::2] in c[b::2]:
006 | ~^^^^^^
007 | ValueError: attempt to assign sequence of size 1 to extended slice of size 5
slices are one of the nicest features and yet people forget about them so often
cloud fossil
How can I do this with slices
versed eagle
it's the same as a tuple unpacking assignment-
the amount of items needs to match
cloud fossil
!e
from string import ascii_uppercase, ascii_lowercase
a = list(ascii_uppercase[:10])
c = list(ascii_lowercase[:10])
b = 0
print(f"{a = }\n{c = }")
for a[b:b+2:2] in c:
b += 2
print(f"{a = }\n{c = }")
@cloud fossil :x: Your 3.11 eval job has completed with return code 1.
001 | a = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J']
002 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j']
003 | Traceback (most recent call last):
004 | File "/home/main.py", line 6, in <module>
005 | for a[b:b+2:2] in c:
006 | ~^^^^^^^^^
007 | ValueError: attempt to assign sequence of size 1 to extended slice of size 0
They are also very non-obvious if you are doing assignment to slice or deleting slice.
For example, if you assign short sequence to slice, list will mutate and become shorter. It is tricky (and not efficient) if you are assigning to slice that is not on the end of the list
cloud fossil
!e
from string import ascii_uppercase, ascii_lowercase
a = list(ascii_uppercase[:10])
c = list(ascii_lowercase[:10])
b = 0
print(f"{a = }\n{c = }")
for a[b], a[b+1] in c[::2]:
b += 2
print(f"{a = }\n{c = }")
@cloud fossil :x: Your 3.11 eval job has completed with return code 1.
001 | a = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J']
002 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j']
003 | Traceback (most recent call last):
004 | File "/home/main.py", line 6, in <module>
005 | for a[b], a[b+1] in c[::2]:
006 | ^^^^^^^^^^^^
007 | ValueError: not enough values to unpack (expected 2, got 1)
!e
from string import ascii_uppercase, ascii_lowercase
a = list(ascii_uppercase[:10])
c = list(ascii_lowercase[:10])
b = 0
print(f"{a = }\n{c = }")
for a[b], a[b+1] in c[b:b+2:2]:
b += 2
print(f"{a = }\n{c = }")
@cloud fossil :x: Your 3.11 eval job has completed with return code 1.
001 | a = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'I', 'J']
002 | c = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j']
003 | Traceback (most recent call last):
004 | File "/home/main.py", line 6, in <module>
005 | for a[b], a[b+1] in c[b:b+2:2]:
006 | ^^^^^^^^^^^^
007 | ValueError: not enough values to unpack (expected 2, got 1)
I'm totally not understanding how this should work
fleet bridge
They are also very non-obvious if you are doing assignment to slice or deleting ...
By "tricky" i mean "it is easy to make off-by-one errors"
versed eagle
mhm
fleet bridge
!e weird way to copy a list (inspired by today's conversation) ```py
a = [*'abcdefgh']
print(a)
c = [' '] * len(a)
for i, c[i] in enumerate(a):
print(c)
assert a == c
@fleet bridge :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h']
002 | ['a', ' ', ' ', ' ', ' ', ' ', ' ', ' ']
003 | ['a', 'b', ' ', ' ', ' ', ' ', ' ', ' ']
004 | ['a', 'b', 'c', ' ', ' ', ' ', ' ', ' ']
005 | ['a', 'b', 'c', 'd', ' ', ' ', ' ', ' ']
006 | ['a', 'b', 'c', 'd', 'e', ' ', ' ', ' ']
007 | ['a', 'b', 'c', 'd', 'e', 'f', ' ', ' ']
008 | ['a', 'b', 'c', 'd', 'e', 'f', 'g', ' ']
009 | ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h']
fun implementation detail: i, c[i] = value works even if i is not previously defined because:
iis assigned tovalue[0]iis loaded to stack (there is no error becauseiwas defined by previous step)c[TOS]is assigned tovalue[1]
night quarryBOT
@languid hare :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 2 0 RESUME 0
002 |
003 | 3 2 LOAD_GLOBAL 0 (tup)
004 | 14 UNPACK_SEQUENCE 2
005 | 18 STORE_FAST 0 (i)
006 | 20 LOAD_GLOBAL 2 (arr)
007 | 32 LOAD_FAST 0 (i)
008 | 34 STORE_SUBSCR
009 | 38 LOAD_CONST 0 (None)
010 | 40 RETURN_VALUE
011 | None
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/ohapehimef.txt?noredirect
vast wave
!e ```py
t = (0, "abc")
arr = [1]
i, arr[i] = t
print(i, arr)
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
0 ['abc']
interesting
low lynx
huh
quartz wave
!e ```py
arr = arr[0] = [0]
print(arr)
@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
[[...]]
arr=arr [0]=[0] is pretty funny
vast wave
!e ```py
arr = arr[0] = [0]
print(arr)
```
[0] gets assigned to arr[0], which gets assigned to arr?
vast wave
!e ```py
import("dis").dis(compile("a = a[0] = [1]", "", "exec"))
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 0 0 RESUME 0
002 |
003 | 1 2 LOAD_CONST 0 (1)
004 | 4 BUILD_LIST 1
005 | 6 COPY 1
006 | 8 STORE_NAME 0 (a)
007 | 10 LOAD_NAME 0 (a)
008 | 12 LOAD_CONST 1 (0)
009 | 14 STORE_SUBSCR
010 | 18 LOAD_CONST 2 (None)
011 | 20 RETURN_VALUE
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/ucowuriluh.txt?noredirect
vast wave
god that is funky
versed eagle
for thing = thing2 = thing3 = value it goes thing = value, then thing2 = thing, and so on
vast wave
what the fuck?
!e ```py
import ast
print(ast.dump("a = b = c = d = 1", indent=2))
!e ```py
import ast
print(ast.dump(ast.parse("a = b = c = d = 1"), indent=2))
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | Module(
002 | body=[
003 | Assign(
004 | targets=[
005 | Name(id='a', ctx=Store()),
006 | Name(id='b', ctx=Store()),
007 | Name(id='c', ctx=Store()),
008 | Name(id='d', ctx=Store())],
009 | value=Constant(value=1))],
010 | type_ignores=[])
yep
python moment
versed eagle
it would make a little more sense imo if it went from right to left here, like this ```py
thing = (thing2 = (thing3 = value))
unreal echo
https://github.com/HRLO77/forums-website/blob/main/db_funcs.py#L239
Most of the code in this does not follow pep8
night quarryBOT
db_funcs.py line 239
p = map((lambda x: (x[0], (*x[1:5], *x[6:-2], len(x[-1])-len(x[-2])))), filter((lambda x: x[0] in ids), await get_posts()))```db_funcs.py line 400
[id, post[1] if title is None else title, post[2] if content is None else content, post[3] if date is None else date, None if fp is None and post[4] is not None else (lambda: post[4] if fp is '' else fp)(), post[5] if ip is None else ip, post[6] if pin is None else pin, f'{post[7]}' if upvotes is None else f'{upvotes}', f'{post[8]}' if downvotes is None else f'{add_downvote}', id],```db_funcs.py lines 43 to 52
if not (await os.path.isfile(file)):file = glob.glob(f'./**/{file}', recursive=True)[0]
async with aiofiles.open(file) as f:code = await f.read()
fg = ''.join(random.sample(string.ascii_letters, k=52)) + '.py'
async with aiofiles.open(fg, 'w') as f:await f.write(code)
await os.remove(fg)
g = {i:v for i,v in globals().items()}
g['DATA'] = data
if threaded:await asyncio.to_thread(exec, [code, g, {i:v for i,v in locals().items()}])
else:exec(code, g, {i:v for i,v in locals().items()})
print(f'Flow {flow} executed {file} given type data {t} on {datetime.datetime.utcnow()} UTC.')```!epy print('\n'.join(['abcdefghijklmnopqrstuvwxyz'[:_]for _ in range(1,26)]))
night quarryBOT
@arctic skiff :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | a
002 | ab
003 | abc
004 | abcd
005 | abcde
006 | abcdef
007 | abcdefg
008 | abcdefgh
009 | abcdefghi
010 | abcdefghij
011 | abcdefghijk
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/ohaloqoqum.txt?noredirect
last locust
!e```py
print('\n'.join(['abcdefghijklmnopqrstuvwxyz'[:_]for _ in range(1,26)]))...
!epy print('\n'.join('abcdefghijklmnopqrstuvwxyz'[:_]for _ in range(1,26)))
night quarryBOT
@last locust :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | a
002 | ab
003 | abc
004 | abcd
005 | abcde
006 | abcdef
007 | abcdefg
008 | abcdefgh
009 | abcdefghi
010 | abcdefghij
011 | abcdefghijk
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/zeyukumabe.txt?noredirect
last locust
You don't need the list 
arctic skiff
You don't need the list <:lemon_wink:901978998400229376>
Ok
Forgot that it also takes *args
versed eagle
it's a generator
arctic skiff
!e```py
print('\n'.join('abcdefghijklmnopqrstuvwxyz'[:_]for _ in range(1,26)))``...
!e```pg
print('\n'.join(import('string').ascii_lowercase[:_]for _ in range(1,26)))
You need the .ascii_lowercase (which makes it longer)
night quarryBOT
@arctic skiff :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | a
002 | ab
003 | abc
004 | abcd
005 | abcde
006 | abcdef
007 | abcdefg
008 | abcdefgh
009 | abcdefghi
010 | abcdefghij
011 | abcdefghijk
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/gihahedoqi.txt?noredirect
arctic skiff
You need the `.ascii_lowercase` (which makes it longer)
Haven't used string in a while
last locust
Haven't used string in a while
!e py print("\n".join(map(chr,range(97,123))))there's also this, which is even shorter
night quarryBOT
@last locust :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | a
002 | b
003 | c
004 | d
005 | e
006 | f
007 | g
008 | h
009 | i
010 | j
011 | k
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/ijocekoniy.txt?noredirect
last locust
You'd just have to add the slice logic (which thinking about it would probably make it longer)
finite blaze
I was wondering if there is a way to code in python without using letters?
Like with javascript
last locust
I was wondering if there is a way to code in python without using letters?
What do you define as letters?
finite blaze
<@442244135840382978> :white_check_mark: Your 3.11 eval job has completed with r...
^
arctic skiff
What do you define as letters?
Ig he is talking about jsfuck
finite blaze
yup
last locust
If you mean solely latin a-z then you can just use other scripts
versed eagle
You'd just have to add the slice logic (which thinking about it would probably m...
!e ```py
print("\n".join("".join(map(chr,range(97,98+i)))for i in range(26)))
@versed eagle :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | a
002 | ab
003 | abc
004 | abcd
005 | abcde
006 | abcdef
007 | abcdefg
008 | abcdefgh
009 | abcdefghi
010 | abcdefghij
011 | abcdefghijk
... (truncated - too many lines)
Full output: https://paste.pythondiscord.com/sicanetazi.txt?noredirect
versed eagle
print("\n".join("".join(map(chr,range(97,98+i)))for i in range(26)))
print('\n'.join('abcdefghijklmnopqrstuvwxyz'[:_]for _ in range(1,26)))
it's shorter
because you don't need the start argument for the range call
last locust
```
print("\n".join("".join(map(chr,range(97,98+i)))for i in range(26)))
print('...
Ah yeah, just
slow marsh
s
!e ```py
import matplotlib.pyplot as plt
x = [1,3,5,6]
y = [1,6,7,8]
plt.plot(x,y)
plt.show()
@slow marsh You've already got a job running - please wait for it to finish!
Sorry, an unexpected error occurred. Please let us know!
TimeoutError:
unique heath
@slow marsh
unique heath
yeah ok
unique heath
eager ore
https://github.com/mTvare6/hello-world.rs
there needs to be more rocket emoji
wait until you see enterprise fizzbuzz though
versed eagle
wait until you see enterprise fizzbuzz though
it's unmaintained though :(
versed eagle
https://github.com/mTvare6/hello-world.rs
huh so apparently linus torvalds has contributed to that one
low lynx
encode_rail_fence_cipher,decode_rail_fence_cipher=lambda s,n:''.join([''.join(map(''.join,zip(s[i::(k:=2*~-n)],s[k-i::k]+'็ง'))),s[::k],s[n-1::k]][(i%~-n<1)+(n-i<2)].rstrip('็ง')for i in range(n)),lambda s,n,c=0:s and''.join(map(''.join,zip(*[s[:(c:=(g:=1+(l:=len(s)-1)//(k:=2*~-n)))],*(m:=[[[],[]],[*zip(*[zip(*zip(*[iter(f'{s[c:(c:=c+l//k*2+(l%k>=i)+(l%k>=k-i))]:็ง<{g*2}}')]*2))for i in range(1,n-1)])]][n>2])[0],f'{s[c:]:็ง<{g}}',*m[1][::-1]]))).rstrip('็ง')
one liner for this https://www.codewars.com/kata/58c5577d61aefcf3ff000081
cannot be bothered to golf further
eager ore
huh so apparently linus torvalds has contributed to that one
i love git-blame-someone-else
versed eagle
i love git-blame-someone-else
oh interesting
keen thicket
hey everyone! I am running an Obfuscated Python Code Competition. Should be lots of fun! Please enter or dm me if you want to be a judge: http://pyobfusc.com
dusty zodiac
Can anyone help me shorten this code I wrote.
def square_digits(num):
return int("".join([str(i) for i in [int(i)**2 for i in list(str(num))]]))
It looks like a monstrosity IK hahah
versed eagle
square_digits=lambda n:int("".join(map(str,[int(i)**2for i in str(n)])))
@versed eagle ```py
square_digits=lambda n:int(''.join(str(int(i)**2)for i in'%i'%n))
That's much shorter DANG ty
I see '%i'
unique heath
hi
tough willow
hi
last locust
!e ```py
import matplotlib.pyplot as plt
x = [1,3,5,6]
y = [1,6,7,8]
plt.plot(x,...
-
plt.show() is blocking which is why it times out
-
even if you pass
block=False, it opens a new window and doesn't save a file so you wouldn't be able to see on snekbox anyway. -
The correct way to do this is to use
plt.savefig, as follows:
!e ```py
import matplotlib.pyplot as plt
x = [1,3,5,6]
y = [1,6,7,8]
plt.plot(x,y)
plt.savefig("plot.png")
@last locust :white_check_mark: Your 3.11 eval job has completed with return code 0.
pure dew
oh nice
vast wave
<@696824456697479240> :white_check_mark: Your 3.11 eval job has completed with r...
!e print("\n".join("#"*n for n in range(10,0,-1)))
night quarryBOT
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | ##########
002 | #########
003 | ########
004 | #######
005 | ######
006 | #####
007 | ####
008 | ###
009 | ##
010 | #
discord's search fucking sucks
rugged owl
discord's search fucking sucks
It gives up
worn sage
```
print("\n".join("".join(map(chr,range(97,98+i)))for i in range(26)))
print('...
What is the use of [:_] in the context you used?
rugged owl
What is the use of `[:_]` in the context you used?
slicing at _ through the following loop list comp, [start:stop:step] iirc
tough willow
slicing at _ through the following loop ~~list comp~~, [start:stop:step] iirc
Start Stop step
sick hound
!e
# This comment is necessary for the code to run.
# This comment is also necessary.
# So is this one.
q = [36, 29, 0, -36, -69, -184, -165, -177, -210, -252, -296, -399]
n = 0
def __a():
pass
def __b():
global n
for i in __a.__code__.co_lines():
n += i[0] + (i[1] * i[2])
for i in range(len(q)):
print(str(__b())[10:] + chr(q[i] + n), end="")```@sick hound :white_check_mark: Your 3.11 eval job has completed with return code 0.
ยรฝลฦวซวจษซหฬอคฮจฮฑ
Haha it doesn't even run properly in here
vague cairn
Was b supposed to return something?
zenith geode
!e
s = (x for x in range(10))
print(list(s).__sizeof__())
print(list(s).__sizeof__())
print(list(s).__sizeof__())
@zenith geode :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | 168
002 | 40
003 | 40
restive void
The generator is just exhausted the second and third time?
!e empty list takes less space than non-empty list
s = (x for x in range(10))
print(list(s))
print(list(s))
print(list(s))
@restive void :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
002 | []
003 | []
oh lmfao thanks
sick hound
Was b supposed to return something?
Nope, the code runs fine in a .py file. And removing the line the comments are on makes it throw an exception :)
!e @commands.group( help="Shows boosts count", usage="boosts", aliases=["bc"]) @blacklist_check() async def boosts(self, ctx): await ctx.send( embed=discord.Embed(title=f"Boosts Count Of {ctx.guild.name}", description="**`%s`**" % (ctx.guild.premium_subscription_count), color=0x977FD7))
night quarryBOT
@sick hound :x: Your 3.11 eval job has completed with return code 1.
001 | File "/home/main.py", line 5
002 | @blacklist_check()
003 | IndentationError: unexpected indent
what
terse mortar
!e ```@commands.group(
help="Shows boosts count",
...
That wonโt work.
The bot uses a sandbox environment so itโs not actually connected to the bot
rugged owl
Nope, the code runs fine in a .py file. And removing the line the comments are o...
!e
q = [36, 29, 0, -36, -69, -184, -165, -177, -210, -252, -296, -399]
n = 0
def __a():
pass
def __b():
global n
for i in __a.__code__.co_lines():
n += i[0] + (i[1] * i[2])
for i in range(len(q)):
print(str(__b())[10:] + chr(q[i] + n), end="")```@rugged owl :white_check_mark: Your 3.11 eval job has completed with return code 0.
dยรรรปรฤลลฎฦฦลฑ
Doesnt seem to throw an exception
quartz wave
Nope, the code runs fine in a .py file. And removing the line the comments are o...
what python version?
quartz wave
!e
```py
# This comment is necessary for the code to run.
# This comment is also...
oh it's 3.10
!e ```py
'some' r'eally' 'useless' + str('ings') \
- int('4nd integers'[:3 is 3]) + f'or' r'eal' +
str('this' is not 'a' 'joke')
q = [36, 29, 0, -36, -69, -184, -165, -177, -210, -252, -296, -399]
n = 0
def __a():
pass
def __b():
global n
i = next(it := __a.__code__.co_lines()) and next(it)
n += i[0]-2 + (i[1] * i[2])
for i in range(len(q)):
print(str(__b())[10:] + chr(q[i] + n), end="")
@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
001 | /home/main.py:2: SyntaxWarning: "is" with a literal. Did you mean "=="?
002 | * int('4nd integers'[:3 is 3]) + f'or' r'eal' + \
003 | /home/main.py:3: SyntaxWarning: "is not" with a literal. Did you mean "!="?
004 | str('this' is not 'a' 'joke')
005 | Hello World!
cipher = lambda t, s: ''.join(chr((ord(c) - 65 + s) % 26 + 65) if c.isupper() else chr((ord(c) - 97 + s) % 26 + 97) if c.isalpha() else c for c in t)
Wondering if this can be shortend further??
tough willow
cipher=lambda t,s:''.join(chr((ord(c)-65+s)%26+65)if c.isupper()else chr((ord(c)-97+s)%26+97)if c.isalpha()else c for c in t)```I'm assuming there won't be any non-alphabetic characters?
oh nvm
gleaming linden
cipher=lambda t,s:''.join([c,chr((ord(c)-65+s-c.islower()*32)%26+97-c.isupper()*32)][c.isalpha()]for c in t)
``` does this work?
it didn't work, it should be fixed nowcipher=lambda t,s:''.join([c,(d:=65+c.islower()*32),chr((ord(c)-d+s)%26+d)][-c.isalpha()]for c in t)
``` 100This one works
dusty zodiac
```py
cipher=lambda t,s:''.join([c,(d:=65+c.islower()*32),chr((ord(c)-d+s)%26+d)...
I dont think this one works
gleaming linden
oh wait I'm stupid
cipher=lambda t,s:''.join([c,chr((ord(c)-(d:=65+c.islower()*32)+s)%26+d)][c.isalpha()]for c in t)
97
dusty zodiac
dang 97 length
sudden osprey
c.islower()*32 -> ord(c)&32 I think should work
gleaming linden
cipher=lambda t,s:''.join([c,chr((ord(c)-(d:=65+(ord(c)&32))+s)%26+d)][c.isalpha()]for c in t)
``` works but needs brackets, 94cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s-(d:=(n&96)+1))%26+d)][64<n&95<91]for c in t)
92 if it works
cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s-(d:=-~(n&96)))%26+d)][64<n&95<91]for c in t)
Same length but looks cooler
oh wait
cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s+(d:=~(n&96)))%26-d)][64<n&95<91]for c in t)
91
cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s+~(d:=n&96))%26-~d)][64<n&95<91]for c in t)
90
wheat river
cipher=lambda t,s:''.join([chr(((n:=ord(c))-s-~(d:=n&96))%26-~d),c][not n^32]for c in t)
86^
sick hound
What are we trying to golf now
fleet bridge
>>> cipher('abcdefgh',1)
'bcdefghi'
``` it is a caesar cipher i guessthis doesn't work nvm
sudden osprey
I'd ignored non-ASCII btw, given the original code only works properly for ascii. (could use [96<n|32<123] or [64<n&~32<91] to fix that, although it's +1 character)
wheat river
cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s+~(d:=n&96))%26-~d)][n>>5>1]for c in t)
86
gleaming linden
```py
cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s+~(d:=n&96))%26-~d)][n>>5>1...
doesn't work for characters after z (cipher("abc{", 1) -> "bcdb")
sick hound
@wheat river what are we trying to golf
gleaming linden
caesar cipher
sick hound
```py
cipher = lambda t, s: ''.join(chr((ord(c) - 65 + s) % 26 + 65) if c.isuppe...
Shorten this?
gleaming linden
pretty much
gleaming linden
What
when the key is 1, all the letters get shifted by 1 (so a -> b, b -> c, etc. and z -> a)
when the key is 2, all letters get shifted by 2 (a -> c, b -> d, ..., y -> a, z -> b)
etc.
sick hound
!e
exec(bytes('ๆฅฃๆกฐ็ฅใด ๆฐ ๆตกๆขโกโฑด็ โบโงๆจฎๆฅฏโกฎๆกฃโกฒๆผจๆฒๆจโฉโญใถโ ็ซโฉโฅใฒโฌ ใ โคตๆค โฆโนฃ็ฉ็ตๆฐโกฒโฉๆฑฅๆณๆ ็จโ จ็ฏโกคโฅฃโด ใค โทโซโฅณโ ใ โถโซในโฉๆฉๆ ๆคฎๆ
ณ็ฌๆ
จโคจๆ ็ฌโฅโฃๆฝฆโฒโฃๆนฉ็ โฉ','u16')[2:])
@sick hound :warning: Your 3.11 eval job has completed with return code 0.
[No output]
hehe bytes
gleaming linden
!e
```py
exec(bytes('ๆฅฃๆกฐ็ฅใด ๆฐ ๆตกๆขโกโฑด็ โบโงๆจฎๆฅฏโกฎๆกฃโกฒๆผจๆฒๆจโฉโญใถโ ็ซโฉโฅใฒโฌ ใ โคตๆค โฆโนฃ็ฉ็ตๆฐโกฒโฉๆฑฅๆณๆ ็จโ จ็ฏโกคโฅฃโด ใค โทโซโฅณโ ใ โถโซในโฉๆฉ...
that's longer though
sick hound
oh
gleaming linden
also, we usually count in bytes and not characters
sick hound
not sure what that means
gleaming linden
ASCII characters take up only 1 byte, but other characters (usually in degen solutions) take up more than 1 byte
eg ๆฅฃ is 1 character, but it might take up 2 or 3 bytes instead
gleaming linden
- the function is called
cipher, you can't change that - it has to preserve case (uppercase should remain uppercase and lowercase should remain lowercase)
- it should leave non-alphabetic characters alone
gleaming linden
>>> cipher("abcABCuvw@{", 5)
'fghFGHzab@{'
``` here's a testcase if you need oneyea, i cant get the range right with the bit shifts
eager ore
can you help me with steal sticker command (actually i want it)
wrong channel
sick hound
wrong channel
wrong world ๐ฆ
languid night
? ๐ญ
fleet bridge
!e
```py
exec(bytes('ๆฅฃๆกฐ็ฅใด ๆฐ ๆตกๆขโกโฑด็ โบโงๆจฎๆฅฏโกฎๆกฃโกฒๆผจๆฒๆจโฉโญใถโ ็ซโฉโฅใฒโฌ ใ โคตๆค โฆโนฃ็ฉ็ตๆฐโกฒโฉๆฑฅๆณๆ ็จโ จ็ฏโกคโฅฃโด ใค โทโซโฅณโ ใ โถโซในโฉๆฉ...
what is a u16? utf-16?
flint hollow
low-level type. guessing unsigned 16-bit int
quartz wave
```python
cipher=lambda t,s:''.join([c,chr(((n:=ord(c))+s+~(d:=n&96))%26-~d)][64...
88 ```py
cipher=lambda t,s:''.join(chr([n:=ord(c),(n+s+~(d:=n&96))%26-~d][64<n&95<91])for c in t)
no it is actually an alias for utf16
wow, docs are now in dark theme! awesome!
since when?
last locust
since when?
python/python-docs-theme#44
glass drumBOT
last locust
It's been in progress for... 3 years apparently. Was finally merged last week
fleet bridge
i love it
last locust
Looks like it only applies to the dev docs (i.e. py3.12) atm though
sick hound
I kind of like how the dark theme uses the Monokai theme for syntax highlighting despite the fact I think it's kind of overrated
sudden osprey
88 ```py
cipher=lambda t,s:''.join(chr([n:=ord(c),(n+s+~(d:=n&96))%26-~d][64<n&9...
Nice, good find
radiant anchor
!e
unused=b'\x50K\3\4'+b'\0'*26+b'+(\xca\xcc+\xd1P\xcfHL\xceNMQ\xc8\xc9\xcfQ\xd7\4\0PK\1\2'+b'\0'*6+b'\1'+b'\0'*9+b'\x15'+b'\0'*7+b'\13'+b'\0'*17+b'__\x6da\x69n__.\x70y\x50K\5\6'+b'\0'*8+b'9\0\0\0003\0\0\0'
i=__import__
i("runpy").run_path(i("py_compile").compile(__file__))
@radiant anchor :white_check_mark: Your 3.11 eval job has completed with return code 0.
hacked lol
Files with disallowed extensions can't be uploaded: .pyc
radiant anchor
golden finch
!e ```py
'some' r'eally' 'useless' + str('ings') \
* int('4nd integers'[:3 is 3]...
Is this somehow encoding text in the operators?
lunar marsh
!e
```py
unused=b'\x50K\3\4'+b'\0'*26+b'+(\xca\xcc+\xd1P\xcfHL\xceNMQ\xc8\xc9\xc...
thatโs so stupid lol. why would it prioritise a zip when it has a pyc headerโฆ
spring rose
I have sever, root and intermediate certs. Where should these be uploaded so that i don't get SSL error when I establish a connection to the url endpoint using python get and post methods?
spring rose
Basically i have a python script to access a url endpoint
That fails as we have a self signed certificate
But i have all the self signed certificate with me
The root intermediate server and chain
Now which among these should I add to make the secure post request to the url?
And where should these be added?
restive void
Basically i have a python script to access a url endpoint
That fails as we have ...
This is the wrong channel for this sort of question; #โ๏ฝhow-to-get-help
quartz wave
Is this somehow encoding text in the operators?
no
finite blaze
Hey
I'm trying to golf something
Basically i need to find all numbers in a file that are a power of 3
What I'm doing rn is
print(len(*"".join(["","1"][math.log(int(i.strip()),3)] for i in open("file.txt")))
And i get some weird results from it
finite blaze
I forgot that i do .is_integet() after that log()
It gives me a generator object genexpr
spring rose
What is this queue parameter in airflow??
golden finch
this isn't a help channel
sly ibex
Depends
vast wave
!e ```py
class Whatever:
def init(self, msg, amount):
self.msg = msg
self.amount = amount
def print(self):
print(f"#{self.amount}: {self.msg*self.amount}")
fake_inst = type("", (), {})()
Whatever.init(fake_inst, "message ", 10)
Whatever.print(fake_inst)
@vast wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
#10: message message message message message message message message message message
instances are for losers
quartz wave
instances are for losers
this is a โ instance
fake_inst = type("", (), {})()
shhh
no one needs to know
restive void
no one needs to know
You can just do Whatever.__init__(Whatever, "message ", 10), no need for the fake instance :P
quartz wave
How can we append a item to a list inside a list?
โซ,l[i]+=x
rain sluice
?
quartz wave
!e ```py
l=l[0]=[0]
l[0]+=l,
print(l)
@quartz wave :white_check_mark: Your 3.11 eval job has completed with return code 0.
[[...], [...]]
@rain sluice
rain sluice
thx lemme try it
is l supposed to be the value
or the list
beginner python programmer so pretty new
quartz wave
beginner python programmer so pretty new
you shouldn't be in this channel then
rain sluice
then where
versed eagle
rain sluice
that'll take ages
quartz wave
then where
maybe try and ask in #python-discussion or open a thread in #1035199133436354600 because this channel will only confuse you the further you try to investigate here
vague cairn
How can we append a item to a list inside a list?
X=[1,2,3,4,5]
y=[(X.insert(xi*2, 'inserted'), 5-xi)[-1]
for xi in tuple(X)]
print(X, y, sep='\n')```
?Is it possible to shorten this into one line?
class Solution(object):
def searchInsert(self, nums, target):
x = nums
if target in nums:
return x.index(target)
else:
x.append(target)
return x.index(target)
I tried doing this but it turns out x.appen(target) returns a None.
return nums.index(target) if target in nums else nums.index(nums.append(target))
This should work
return nums.index(target) if target in nums else (len(nums), nums.append(target))[0]
it works thank you
fleet bridge
return (nums.append(target), nums.index(target))[1]
The thing is, it shouldn't append if it's there
dusty zodiac
I just realized I forgot to add .sort() to the thing rip
dreamy pier
This should work
```py
return nums.index(target) if target in nums else (len(num...
(+class)
Solution = type("Solution", (object,), {"searchInsert": lambda self, nums, target: nums.index(target) if target in nums else (len(nums), nums.append(target))[0]})
Solution=type("Solution",(),{"searchInsert":lambda s,n,t:(t in n and(n.index(t),)or(len(n),n.append(t)))[0]})
Nice
gleaming linden
Solution=type("Solution",(),{"searchInsert":lambda s,n,t:(t in n or n.append(t),n.index(t))[1]})
``` does this work?Solution=type("Solution",(),{"searchInsert":lambda s,n,t:[n.append(t)]*(t in n)and n.index(t)})
gleaming linden
~~Solution=type("Solution",(),{"searchInsert":lambda s,n,t:[n.append(t)]*(t in n...
I'm pretty sure that appends t no matter what
ivory lichen
def perfect(n):
divs = [1]
i = 2
while i != n and n != 1:
if n % i == 0:
divs.append(i)
i +=1
else:
i +=1
dsum = sum(divs)
if dsum == n and n != 1:
return True
else:
return False
is this a "good" answer to the perfect number question?
finite blaze
```python
def perfect(n):
divs = [1]
i = 2
while i != n and n != 1:
if...
That's not the right channel for such a question
but yes, this is a good answer
!e
def perfect(n):
divs = [1]
i = 2
while i != n and n != 1:
if n % i == 0:
divs.append(i)
i +=1
else:
i +=1
dsum = sum(divs)
if dsum == n and n != 1:
return True
else:
return False
print(perfect(6))
@finite blaze :white_check_mark: Your 3.11 eval job has completed with return code 0.
True
i think the class name isn't needed
also short circuiting optimizations (87) ```py
Solution=type("",(),{"searchInsert":lambda s,n,t:1>(t in n)==n.append(t)or n.index(t)})
just using the class is shorter (82) ```py
class Solution:
def searchInsert(s,n,t):
if(t in n)<1:n+=t,
return n.index(t)
I'd say a "good" answer would be not using brute force and using euclid-euler and lucas-lehmer
and for your current solution first loop can be replaced by a for loop
and
if cond:
return True
else:
return False
is just return cond
dreamy pier
just using the class is shorter (82) ```py
class Solution:
def searchInsert(s,n...
class Solution:searchInsert=lambda s,n,t:1>(t in n)==n.append(t)or n.index(t)
what if list.appendindex existed
versed eagle
what would its behaviour be?
quartz wave
actually what about list.indexdefault
versed eagle
what would its behaviour be?
quartz wave
what would its behaviour be?
same thing as n.indexdefault(t) in what we're trying to do above
versed eagle
yeah ik that. im asking which part of that you're trying to do
quartz wave
yeah ik that. im asking which part of that you're trying to do
what
versed eagle
i already knew that it would be for what you were doing above
what im asking is which part of that it would be used for
i.e., what would it do (hence the question, "what would its behaviour be")
quartz wave
like append if the value isn't in the list and return its index?
versed eagle
ah
quartz wave
it would be so simple
we could just do staticmethod(list.indexdefault)
finite blaze
hmm, is there a way to add other functions to lists etc?
versed eagle
it would be so simple
we could just do `staticmethod(list.indexdefault)`
staticmethod(list.indexdefault)
lambda s,n,t:n.indexdefault(t)
quartz wave
ok
versed eagle
that'd be interesting, but not very practical i think
quartz wave
yay ```pycon
a = []
a.indexdefault(2)
0
a
[2]
!e fishhook ```py
from fishhook import hook_cls
@hook_cls(list)
class _:
def indexdefault(self, value):
for i, obj in enumerate(self):
if obj == value:
return i
i = len(self)
self.append(value)
return i
a = []
print(a.indexdefault(2))
print(a)