#internals-and-peps | Python | Page 57

pseudo cradle Jul 19, 2020, 3:46 AM

#

Oh, it automatically unpacks?

raven ridge Jul 19, 2020, 4:03 AM

#

When you return 5, 10 you're returning a tuple. When you a, b = foo() you're unpacking the returned object.

#

@pseudo cradle

pseudo cradle Jul 19, 2020, 4:08 AM

#

Mindboggling

#

def multi_arg_func(a,b,c):
    print( a + b +c )
    return a, b, c
myTup = (1,2,3)
d, e, f = multi_arg_func(*myTup)```

#

Like, that's so cool that that works

#

I don't know how I'm ever going to code anything again in C++

raven ridge Jul 19, 2020, 4:10 AM

#

!e ```python
first, *rest, last = 1, 2, 3, 4
print(first, rest, last)

fallen slateBOT Jul 19, 2020, 4:10 AM

#

@raven ridge :white_check_mark: Your eval job has completed with return code 0.

1 [2, 3] 4

pseudo cradle Jul 19, 2020, 4:11 AM

#

Python is making me so lazy it will be the death of me

#

That's also pretty cool

raven ridge Jul 19, 2020, 4:14 AM

#

!e ```python
def n_element_list(n):
return list(range(n))
a, b = n_element_list(2)
print(a, b)
a, *others, b = n_element_list(50)
print(a, b)

fallen slateBOT Jul 19, 2020, 4:14 AM

#

@raven ridge :white_check_mark: Your eval job has completed with return code 0.

001 | 0 1
002 | 0 49

pseudo cradle Jul 19, 2020, 4:14 AM

#

Very nice

raven ridge Jul 19, 2020, 4:14 AM

#

also kinda mind blowing that the tuple it returns isn't required to conform to one particular shape, and different types of unpackings might be valid on different calls.

#

a, b = n_element_list(N) will fail at runtime unless N is 2.

pseudo cradle Jul 19, 2020, 4:18 AM

#

Now, what about a, *others, b = n_element_list(2)?

raven ridge Jul 19, 2020, 4:18 AM

#

that's fine, others will be an empty list.

pseudo cradle Jul 19, 2020, 4:19 AM

#

...nice

raven ridge Jul 19, 2020, 4:19 AM

#

but that'd fail with N=1

#

!e ```python
a, *rest, b = range(2)
print(rest)
a, *rest, b = range(1)

fallen slateBOT Jul 19, 2020, 4:19 AM

#

@raven ridge :x: Your eval job has completed with return code 1.

001 | []
002 | Traceback (most recent call last):
003 |   File "<string>", line 3, in <module>
004 | ValueError: not enough values to unpack (expected at least 2, got 1)

void grail Jul 19, 2020, 5:03 AM

#

don’t quote me but i think *_ ignores the elements when declaring multiple variables

raven ridge Jul 19, 2020, 5:07 AM

#

it doesn't.

#

!e ```python
a, *, b = range(10)
print()

fallen slateBOT Jul 19, 2020, 5:08 AM

#

@raven ridge :white_check_mark: Your eval job has completed with return code 0.

[1, 2, 3, 4, 5, 6, 7, 8]

shy belfry Jul 19, 2020, 5:14 AM

#

What's the convention over here regarding decorators?

#

Say I have a decorator that can be used with or without arguments. Then is it normal to use if statements to allow the decorator to be used with or without braces?

raven ridge Jul 19, 2020, 5:15 AM

#

how else could you do it?

shy belfry Jul 19, 2020, 5:16 AM

#

Seeing Discord.py that isn't really the case. @commands.Cog.listener() doesn't expect arguments but @commands.Cog.listener doesn't work.

raven ridge Jul 19, 2020, 5:17 AM

#

right - you don't get that for free. it's easy to make a decorator that doesn't take any arguments, it's a bit harder to make a decorator that takes arguments, and it's harder still to make one that can be used either way.

shy belfry Jul 19, 2020, 5:17 AM

#

how else could you do it?
@raven ridge i.e. if args are present return wrapper else return decorated function/class.

charred wagon Jul 19, 2020, 5:17 AM

#

That's using an if statement

shy belfry Jul 19, 2020, 5:18 AM

#

right - you don't get that for free. it's easy to make a decorator that doesn't take any arguments, it's a bit harder to make a decorator that takes arguments, and it's harder still to make one that can be used either way.
@raven ridge at the moment all of mine can be used both ways. But is this conventional/accepted?

#

Or is it better for a decorator that could have arguments to symbolize that by using a wrapper?

raven ridge Jul 19, 2020, 5:19 AM

#

a lot of major libraries allow things taking only optional arguments to be called both ways. I'd probably do it if I were designing a class like that. But it's by no means required; it just may make things easier for your users.

#

the fact that @foo() and @foo work drastically differently is something that the author of the decorator knows, but the user might not, so users may be confused why the former works and not the latter if you don't make both work.

#

or you could sidestep it by requiring at least one arg 😄

shy belfry Jul 19, 2020, 5:22 AM

#

Makes sense! So there's nothing essentially against allowing @foo() and @foo is what I'm hearing? Of course it always ends up requiring me to use a factory function or class but that isn't too bad given the usability advantages. I just wasn't sure if maybe @foo was outdated syntax or something.

charred wagon Jul 19, 2020, 5:22 AM

#

No. It definitely isn't

#

It's used by every decorator that doesn't take args.

raven ridge Jul 19, 2020, 5:22 AM

#

no, there's no reason not to expose both if you don't mind doing the extra work - just make sure they actually do the same thing.

#

it'd be incredibly confusing for users if you decided to make @foo and @foo() both work but do something different 😄

shy belfry Jul 19, 2020, 5:27 AM

#

☝🏽 Definitely agreed on that one. Thank you for the feedback! I'll keep doing things this way then! 😄

open trout Jul 19, 2020, 7:28 AM

#

Is this weird?

from typing import NewType
FooType = NewType('FooType', object)

#

I want to make a type only for type checking later down the line

#

The example I saw was always using something like , int) or , str), never just inheriting from object

deep shard Jul 19, 2020, 7:41 AM

#

I just learned hot to text to speech its cool

#

*how

daring siren Jul 19, 2020, 8:07 AM

#

I remember some pep that defined how to mark some arguments as keyword only and some others as positional only. Could anyone link it? Google's failing me

languid dagger Jul 19, 2020, 8:09 AM

#

I believe these are the ones:
https://www.python.org/dev/peps/pep-0570/
https://www.python.org/dev/peps/pep-3102/

Python.org

PEP 570 -- Python Positional-Only Parameters

The official home of the Python Programming Language

Python.org

PEP 3102 -- Keyword-Only Arguments

The official home of the Python Programming Language

daring siren Jul 19, 2020, 8:25 AM

#

Thank you, @languid dagger !

unkempt rock Jul 19, 2020, 9:05 AM

#

hello, is there any active community for PyQt in discord?

manic river Jul 19, 2020, 9:11 AM

#

the list index is out of range , can someone help me?

📎 unknown.png

lost nexus Jul 19, 2020, 9:17 AM

#

this isn't a help channel (see the topic) check out #❓｜how-to-get-help @manic river

manic river Jul 19, 2020, 9:17 AM

#

this isn't a help channel (see the topic) check out #❓｜how-to-get-help @manic river
@lost nexus thanks

red solar Jul 19, 2020, 10:27 AM

#

is there any way to check if a memoryview has been released? (either in python or in C)

#

#define CHECK_MBUF_RELEASED(mbuf) \
    if (((_PyManagedBufferObject *)mbuf)->flags&_Py_MANAGED_BUFFER_RELEASED) { \
        PyErr_SetString(PyExc_ValueError,                                      \
            "operation forbidden on released memoryview object");              \
        return NULL;                                                           \
    }

#

hmm this is what C has

shy belfry Jul 19, 2020, 10:55 AM

#

So, uhh... I've been doing something with the descriptors and data models which seems to be going alright, but unfortunately my Sphinx documentation isn't loading any documentation for members that use that class I made. Does anyone know where to do the update_wrapper(...) stuff to get the documentation working?

cloud crypt Jul 19, 2020, 11:25 AM

#

>>> from enums import Flag, Order, StrFormat
>>> class Perm(StrFormat, Order, Flag, auto_on_missing=True, start=0):
...     Z, X, W, R
...
>>> Perm.Z
<Perm.Z: 0>
>>> Perm.X | Perm.W | Perm.R
<Perm.R|W|X: 7>``` working on my enum implementation, and keyword arguments in `meta_cls.__prepare__` and `meta_cls.__new__` are pretty useful! :D

cloud crypt Jul 19, 2020, 11:42 AM

#

@red solar flags don't seem to be exposed to python, so best thing would be to attempt some operation on it I guess

red solar Jul 19, 2020, 11:43 AM

#

yeah im writing a proposal for some modifications (additions) to memoryview

cloud crypt Jul 19, 2020, 11:44 AM

#

def is_released(memory: memoryview) -> bool:
    try:
        memory.readonly
        return False
    except ValueError:
        return True``` this can work, but yeah, better add some boolean like .released

red solar Jul 19, 2020, 11:46 AM

#

aww drat this means i'm gonna have to write unit tests :/

cloud crypt Jul 19, 2020, 11:48 AM

#

heheh

red solar Jul 19, 2020, 12:01 PM

#

if i'm adding a macro like #define PySomething _PySomething, i don't need to write tests right?

cloud crypt Jul 19, 2020, 12:10 PM

#

I guess

#

as long as _PySomething is tested, obviously

cloud crypt Jul 19, 2020, 12:31 PM

#

oh boy it's time to write tests oh FUUUU

last dune Jul 19, 2020, 12:33 PM

#

Is there any way to convert a list to an integer?

red solar Jul 19, 2020, 12:33 PM

#

lol

#

and yes

#

depends how you want to interpret it tho

#

you want the length of the list to be the value of the integer?

last dune Jul 19, 2020, 12:34 PM

#

No

#

I want the elements inside the list to be converted into an integer

red solar Jul 19, 2020, 12:35 PM

#

x = ["1", "2", "3"]
y = [int(z) for z in x]

#

(but also this isn't the channel to discuss this - #python-discussion or a help channel would be more appropriate)

last dune Jul 19, 2020, 12:37 PM

#

K thanks

red solar Jul 19, 2020, 12:59 PM

#

ugh i ran test on python and 1 failed :/

#

now i gotta go see what went wrong :/

paper echo Jul 19, 2020, 1:16 PM

#

@shy belfry can you give an example where sphinx fails

subtle pike Jul 19, 2020, 1:18 PM

#

i managed to crash sphinx over and over in my last project

#

but a less severe example of sphinx failing is with generic functions because it really isn't a fan of reaching into a function wrapper and using that doc string with typehints

#

you end up with loads and loads of identical docstrings which wastes space

#

unless you do ```py
my_docstring = """..."""
for f in [f, g, h]:
f.doc = my_docstring

spark forge Jul 19, 2020, 1:21 PM

#

Hello. I am new to Discord.. Do I ask a Python question here?

red solar Jul 19, 2020, 1:22 PM

#

#python-discussion or #❓｜how-to-get-help

magic python Jul 19, 2020, 1:25 PM

#

is there such a thing as a profiling context manager to see how long a subset of a script takes to run line by line?

red solar Jul 19, 2020, 1:28 PM

#

can a standard profiler not do that?

magic python Jul 19, 2020, 1:28 PM

#

@red solar i've no idea, i've only just used %lprun for the first time

#

%lprun -T <log file> -f <function name> <function with args>

red solar Jul 19, 2020, 1:32 PM

#

oh you're doing this from the command line :/

sacred tinsel Jul 19, 2020, 1:34 PM

#

The pycharm profiler is pretty good, it can draw a hierarchical tree graph showing how much time each function call took

magic python Jul 19, 2020, 1:34 PM

#

I don't have pycharm :/

sacred tinsel Jul 19, 2020, 1:34 PM

#

I think that's what you're looking for

#

mm, not sure then sorry

red solar Jul 19, 2020, 1:34 PM

#

uses cProfiler behind the scenes

#

what's lprun from?

magic python Jul 19, 2020, 1:35 PM

#

it's an ipython magic thing

red solar Jul 19, 2020, 1:35 PM

#

oh no

#

whats ipython?

#

oh jupyter notebook stuff

#

you should be able to use cProfiler in ipython?

#

oh wait it is cprofiler (sry this is the first time i'm looking at profiling)

#

@magic python you want more precise profiling than per function?

magic python Jul 19, 2020, 1:42 PM

#

@red solar just learnt about exec so am going to try that now, I need it line by line really, unless by "per function" you're referring to something else ( i want to see how long each call to pd.crosstab or whatever is taking )

red solar Jul 19, 2020, 1:42 PM

#

import cProfile

with cProfile.Profile() as pr:
    # ... do something ...

pr.print_stats()

this is an option (but it's just pure python)

#

if u wanna see how long each function call takes, the profiler should be showing you that already 🤔

magic python Jul 19, 2020, 1:43 PM

#

I'm just trying with quotes - profile.run( "f(x)") instead of .run( f(x) ), seems to have worked thanks

#

it's much more detailed than lprun actually

red solar Jul 19, 2020, 1:44 PM

#

i'm fairly sure lprun uses it in the background

#

however i still strongly recommend pycharm if u can get it

#

profiler is just a button and it shows u everything

#

no setup

magic python Jul 19, 2020, 1:45 PM

#

i feel as though I've done th setup now, what's the main thing pycharm offers over this line by line output of profile?

brazen jacinth Jul 19, 2020, 1:46 PM

#

Benefits of an IDE 😳

magic python Jul 19, 2020, 1:46 PM

#

although saying that, 99 0.000 0.000 0.001 0.000 base.py:729(__array__) isn't that clear I guess

#

neat

📎 unknown.png

red solar Jul 19, 2020, 1:48 PM

#

only issue is if you have stuff that takes less than 1ms, it doesn't really count it (gives it 0% of the time)

pearl river Jul 19, 2020, 1:49 PM

#

oh that's cool, I didn't find this button and instead just used cProfile like normal

magic python Jul 19, 2020, 1:50 PM

#

apparently profile.Profile is used for more granularity than profile.run, what if you want less 😅

red solar Jul 19, 2020, 1:50 PM

#

if u want less, pycharm gives u a %

#

wait wym by less?

magic python Jul 19, 2020, 1:51 PM

#

hrm, i wonder how lprun does it - the output from that just shows me code that I wrote, whereas profile.run shows me everything, base.py and core files and all sorts, I don't really need to see those, I just want to see my lines

#

https://stackoverflow.com/questions/3927628/how-can-i-profile-python-code-line-by-line

seems that lprun is what people suggest for that usecase

Stack Overflow

How can I profile Python code line-by-line?

I've been using cProfile to profile my code, and it's been working great. I also use gprof2dot.py to visualize the results (makes it a little clearer).

However, cProfile (and most other Python pro...

#

kernprof -l -v main.py from https://stackoverflow.com/a/62382967/3130747 was useful if it's of interest @red solar

red solar Jul 19, 2020, 2:02 PM

#

ok @profile does seem cool

#

but i'm happy with whole file profiling

magic python Jul 19, 2020, 2:02 PM

#

@red solar yeah - the default output of profile.run was (for me) a bit much

red solar Jul 19, 2020, 2:02 PM

#

also profiling is more a C++ thing for me than python - never profiled python before, i just try to make stuff readable

magic python Jul 19, 2020, 2:03 PM

#

well I'm estimating that this module will take 8 hours to run so I'd like to reduce it if possible lol

red solar Jul 19, 2020, 2:03 PM

#

wtf lol

#

what is it doing

magic python Jul 19, 2020, 2:03 PM

#

😄 running my code

visual shadow Jul 19, 2020, 2:03 PM

#

What kind of monster did you create 😅

red solar Jul 19, 2020, 2:03 PM

#

at this point shouldn't you be writing it as a C extension?

magic python Jul 19, 2020, 2:04 PM

#

Cython? I've considered it - or passing some stuff to Go / Rust or whatever (neither of which I know anything of)

red solar Jul 19, 2020, 2:04 PM

#

but?

magic python Jul 19, 2020, 2:04 PM

#

well but idk what i'm doing I guess

visual shadow Jul 19, 2020, 2:05 PM

#

Before changing languages, I'd try to address the code itself. Are there any bottlenecks or things to optimize

magic python Jul 19, 2020, 2:05 PM

#

that's why I was profiling, never really used it as a workflow before but thought I'd have a look

subtle pike Jul 19, 2020, 2:05 PM

#

scipy already makes extensive use of the CAPI for computational work so the benefit of writing it purely in C or another compiled language is only going to affect the pure python gluecode, which should already be minimised

visual shadow Jul 19, 2020, 2:05 PM

#

I see. Can you break the whole code down into a smaller chunks?

#

Profile a smaller chunk

magic python Jul 19, 2020, 2:06 PM

#

yeah - i can profile a representative sample of the task and try and improve that, all good

visual shadow Jul 19, 2020, 2:06 PM

#

Sounds good.

red solar Jul 19, 2020, 2:53 PM

#

Python build finished successfully!
The necessary bits to build these optional modules were not found:
_bz2 _dbm _gdbm
_hashlib _lzma _sqlite3
_ssl _tkinter _uuid
readline
To find the necessary bits, look in setup.py in detect_modules() for the module's name.
So this stuff is optional, and not having it should be ok

spark magnet Jul 19, 2020, 2:54 PM

#

ssl is often needed.

red solar Jul 19, 2020, 2:54 PM

#

shouldn't be for this tho

#

from _lzma import *
ModuleNotFoundError: No module named '_lzma'

== Tests result: FAILURE ==

1 test failed:
test_zoneinfo

#

how do i report this? like should lzma just not be part of the final product if _lzma doesn't exist?

#

and does this need to be fixed before i can submit a completely unrelated proposal? (wanna add 2 macros and corresponding documentation to the c-api)

#

(the test fails on origin/master before i make any changes)

fringe tusk Jul 19, 2020, 4:45 PM

#

Hello! :)

#

Is there someone very experienced in programming with Python who also invests in stocks?

#

Because I have a very odd concern that I'd love to talk about privately with him/her, it won't take much, not more than five minutes I swear!

grave jolt Jul 19, 2020, 5:20 PM

#

This is definitely not the right channel. I would suggest off-topic channels or #python-discussion.

modest hatch Jul 19, 2020, 5:21 PM

#

does dict.items() return a list or a tuple of items in a dict

grave jolt Jul 19, 2020, 5:22 PM

#

@modest hatch This is not the right channel. Please read the channel description and check out #❓｜how-to-get-help

#

dict.items() returns an iterable object containing tuples.

modest hatch Jul 19, 2020, 5:24 PM

#

okay thanks

red solar Jul 19, 2020, 5:54 PM

#

tfw you search for something, and you want the python result, but google only gives you c-api results 😢

#

https://docs.python.org/3/library/stdtypes.html#memoryview.release does this automatically call https://docs.python.org/3/c-api/buffer.html#c.PyBuffer_Release for you? (i think it does, but the wording isn't explicit enough for my liking)

#

sorry bad question - it does automatically call it

#

what i should have asked, does the documentation make it explicitly clear enough that it does?

light ember Jul 19, 2020, 6:28 PM

#

How my code worked even if it was wrong. (It's already fixed by others coders)

unkempt rock Jul 19, 2020, 6:28 PM

#

Luck

low lagoon Jul 19, 2020, 6:29 PM

#

@light ember This channel isn't for that type of discussion. Please read channel descriptions before proceeding.

light ember Jul 19, 2020, 6:51 PM

#

@low lagoon isn't my question related to design?

#

I mean I asked questions in #general but it's too fast for people to see

red solar Jul 19, 2020, 7:04 PM

#

this channel is more concerning the design of the language, or advanced language concepts, not just normal code

#

@grizzled vigil any chance you're free to walk me through contributing/making a PR to cpython? (like an overview, not in depth - just wanna see if i missed anything)

#

https://devguide.python.org/pullrequest/ been reading this

unkempt rock Jul 19, 2020, 7:17 PM

#

any PyQt community here?

red solar Jul 19, 2020, 7:18 PM

#

maybe in #user-interfaces or #game-development

#

(also a question for #community-meta)

naive hawk Jul 19, 2020, 8:52 PM

#

I need help with my python code

red solar Jul 19, 2020, 8:53 PM

#

#python-discussion or #❓｜how-to-get-help

marsh shoal Jul 19, 2020, 8:58 PM

#

any suggested place from where to learn python (with practice tasks) ?

pearl river Jul 19, 2020, 8:59 PM

#

People like https://www.w3schools.com/python/default.asp I think.

#

though not really the right channel, now that I notice. 🙂

marsh shoal Jul 19, 2020, 9:00 PM

#

ty

unkempt rock Jul 19, 2020, 10:46 PM

#

tho careful w3schools sometimes messes up on code but thats how i learned a bit c++

small siren Jul 20, 2020, 3:43 AM

#

I started as C++ programmer and when I saw most of the concepts like vector etc are used from python, I switched to python and never turned back since haha, I don't mind if its slower than C++ but I just enjoy building code in Python 😉 I wish 95% of the programmers on this planet just switch to Python and enhance it rather than spend time on building other languages etc.

#

@marsh shoal Amazing place to start is thenewboston python youtube videos; Bucky Roberts is a legend !

pseudo cradle Jul 20, 2020, 4:13 AM

#

This bothers me more than I should and I know I've already mentioned this, but I really miss ++ and -- in Python

raven ridge Jul 20, 2020, 4:20 AM

#

Start using +=1 in C++, too 😄

last pollen Jul 20, 2020, 4:28 AM

#

This bothers me more than I should and I know I've already mentioned this, but I really miss ++ and -- in Python
It's very easy to go wrong with those, whereas var += 1 and var -= 1 is explicit and easy to reason about

raven ridge Jul 20, 2020, 4:33 AM

#

When I'm using other languages I really miss if a < b < c: as a shorthand for if a < b and b < c: - that's an incredibly nice quirky Python feature

pseudo cradle Jul 20, 2020, 4:34 AM

#

Yeah, that's fair

#

@last pollen Why is it easy to go wrong with those?

#

Seems pretty intuitive and it's so much fewer keystrokes

crystal pebble Jul 20, 2020, 4:35 AM

#

That if a < b < c syntax needs to be in every language as of yesterday

raven ridge Jul 20, 2020, 4:40 AM

#

Postfix increment vs prefix increment is fairly confusing. And the syntax for overloading postfix increment in C++ is a glaring hack in the language specification

pseudo cradle Jul 20, 2020, 5:22 AM

#

Two of the same keypress is just so much sweeter than the mess of presses all over the keyboard that is the equivalent

last pollen Jul 20, 2020, 5:34 AM

#

how often do you do var += 1?

boreal umbra Jul 20, 2020, 5:35 AM

#

My Java instructor said to never use prefix ++, but it seems way less confusing conceptually than postfix

#

+= is also immediately obvious to a lot of people

gleaming rover Jul 20, 2020, 5:37 AM

#

I rarely even use ++ and --

crystal pebble Jul 20, 2020, 5:39 AM

#

I forget, what are some situations in which prefix and postfix increment / decrement would behave differently?

raven ridge Jul 20, 2020, 5:47 AM

#

*(++x) = 5; means x += 1; *x = 5;
*(x++) = 5; means *x = 5; x += 1;

unkempt rock Jul 20, 2020, 6:51 AM

#

I would prefer += as well since it's much more readable than prefix/postfix syntax

timid relic Jul 20, 2020, 7:47 AM

#

def printLines(line):
numTimes = int(input("Number of repetitions?"))

for i in range(numTimes):
    print(line)

return numTimes

could someone explain to me line 2 and why it is return numTimes

unkempt rock Jul 20, 2020, 8:09 AM

#

#python-discussion

#

I’m amazed python has a == b == c but not a < b < c

torpid bridge Jul 20, 2020, 8:12 AM

#

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

fallen slateBOT Jul 20, 2020, 8:12 AM

#

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

True

visual shadow Jul 20, 2020, 8:12 AM

#

Python has it.

unkempt rock Jul 20, 2020, 8:12 AM

#

Oh what @raven ridge you confused me :/

visual shadow Jul 20, 2020, 8:13 AM

#

He was saying he misses this in other languages

unkempt rock Jul 20, 2020, 8:13 AM

#

Ohhh I can’t read

#

Good to know 😂

#

Thought it was a little suspect python supported one but not the other

strange fog Jul 20, 2020, 8:17 AM

#

i think it's weird that in and not in are also considered comparison operators

#

and therefore chainable

#

cant really see any practical benefit to it either

#

like what's the chance you have to do this?

>>> 2 in [2, 3, 4] in [[2, 3, 4]]
True
>>>

visual shadow Jul 20, 2020, 8:20 AM

#

Very low chance

#

But special cases are not special enough to break consistency

strange fog Jul 20, 2020, 8:21 AM

#

it also opens up the door to weirder and potentially confusing stuff like this

>>> 2 in [2, 3, 4] < [4, 3, 2]
True
>>>

visual shadow Jul 20, 2020, 8:21 AM

#

If you want to write bad code, no one can stop you 😅

#

I'd be shocked to see anyone write stuff like that

deft pagoda Jul 20, 2020, 8:22 AM

#

given that all those can be defined by magic methods, there may exist some nice case where these expressions come up

strange fog Jul 20, 2020, 8:22 AM

#

>>> 2 in [2, 3, 4] == True
False

i could see a beginner making this mistake

visual shadow Jul 20, 2020, 8:23 AM

#

Oh beginners mess up chaining in general, that can always happen

strange fog Jul 20, 2020, 8:24 AM

#

@deft pagoda
true. still though, having membership checks in the same category as comparison just feels strange to me.

#

just from a language design point of view

undone hare Jul 20, 2020, 8:24 AM

#

That's because of CPython internals

strange fog Jul 20, 2020, 8:24 AM

#

this goes beyond cpython. it's part of the python language specs.

undone hare Jul 20, 2020, 8:25 AM

#

It was easier to make it a comp_op

visual shadow Jul 20, 2020, 8:25 AM

#

I think the impact of trying to stop someone from shooting their own foot on everyone else writing sensible code with unnecessary checks is significant enough to merit pythons stance of simplicity

#

Keep it simple. Assume people use things sensibly. Leave it at that

undone hare Jul 20, 2020, 8:26 AM

#

I mean, that is just good sense, nobody should try to chain those, or at least I don't see a reason to do so

unkempt rock Jul 20, 2020, 8:27 AM

#

It’s nice that it’s there if someone finds a good reason tho 🙂

undone hare Jul 20, 2020, 8:27 AM

#

It is okay to allow weird things to be done if nobody would do that

visual shadow Jul 20, 2020, 8:27 AM

#

Well the reason is simple. Chaining, in general, is awesome.

undone hare Jul 20, 2020, 8:28 AM

#

Yeah, that's a really cool functionality

visual shadow Jul 20, 2020, 8:29 AM

#

So, you write logic that implements this awesome feature. And then you notice that it can be used in not-so-awesome ways. This is the point where you decide : is this edge case significant enough to make an explicit check for it, taking a performance hit for every time chaining is used? Or not. That's it.

#

Python often leans towards "consenting adults" mentality in these cases.

#

In chaining in particular, I agree with the design choice.

deft pagoda Jul 20, 2020, 8:31 AM

#

hmm, it seems that ifyou return an arbitrary value from __contains__ it gets converted to a boolean

rigid holly Jul 20, 2020, 8:33 AM

#

does python not want to run multiple instances of the same script?

strange fog Jul 20, 2020, 8:34 AM

#

maybe if in/not in had different precedence to ==/>=/<=/and so on that'd be a better middle ground.

visual shadow Jul 20, 2020, 8:34 AM

#

I don't think python cares one way or another.

rigid holly Jul 20, 2020, 8:34 AM

#

I opened a second window of a working script and it did absolutely nothing

visual shadow Jul 20, 2020, 8:34 AM

#

Hmm

deft pagoda Jul 20, 2020, 8:36 AM

#

not associative:

In [38]: 1 in [1] in [[1]]                                                                                              
Out[38]: True

In [39]: (1 in [1]) in [[1]]                                                                                            
Out[39]: False

In [40]: 1 in ([1] in [[1]])                                                                                            
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-40-c04eb8a4d4b0> in <module>
----> 1 1 in ([1] in [[1]])

TypeError: argument of type 'bool' is not iterable

rigid holly Jul 20, 2020, 8:36 AM

#

it is supposed to open a pygame window but it does not

#

the first instance works

deft pagoda Jul 20, 2020, 8:38 AM

#

this is better than i thought

#

it is just like comp_ops

#

which is much more useful

strange fog Jul 20, 2020, 8:39 AM

#

yep they are actually considered comparison operators

#

https://docs.python.org/3/reference/expressions.html#comparisons

#

comp_operator ::=  "<" | ">" | "==" | ">=" | "<=" | "!=" | "is" ["not"] | ["not"] "in"

deft pagoda Jul 20, 2020, 8:40 AM

#

yeah, i would probably chain these in certain mathematical code

amber nexus Jul 20, 2020, 8:48 AM

#

@naive hawk this isn't a help channel

naive hawk Jul 20, 2020, 8:49 AM

#

oh sorry but where it's ?

amber nexus Jul 20, 2020, 8:49 AM

#

#❓｜how-to-get-help

naive hawk Jul 20, 2020, 8:49 AM

#

thx ^^

spice pecan Jul 20, 2020, 9:57 AM

#

I find it weird and honestly upsetting that in force-casts the output of __contains__ to a boolean, when == and other comparison ops do not

deft pagoda Jul 20, 2020, 9:59 AM

#

i do too, i think it should either raise an error or just allow anything --- like __len__

spice pecan Jul 20, 2020, 10:00 AM

#

I'd rather go with allowing anything to keep it in line with the other operators (and allow for some creative, albeit not exactly practically useful code)

red solar Jul 20, 2020, 11:48 AM

#

why is it upsetting that it force casts it to a boolean? __contains__ is supposed to return a boolean anyway

brazen jacinth Jul 20, 2020, 11:49 AM

#

^

spice pecan Jul 20, 2020, 12:17 PM

#

In the absolute majority of cases it doesn't matter, but it is technically limiting

#

You could say the same about any other comparison operator really, they're all supposed to return a bool

feral nest Jul 20, 2020, 12:42 PM

#

i am kinda new to python and i was just wondering how can you find the mode(most occuring number) and the median(mid value) from a set of inputs. thanks in advance

pearl river Jul 20, 2020, 12:43 PM

#

For the first, I'd use collections.Counter, then take the key with the highest value.

#

For the latter...not sure there are easier methods than sorting the array and taking the middle element.

#

there's bound to be a statistics library that can do all these, too

#

scipy probably can, maybe pandas too.

Also, just realised - this is #internals-and-peps, not #python-discussion or a help channel. It's rather specialized 😛 You might want to check #❓｜how-to-get-help instead.

peak spoke Jul 20, 2020, 12:46 PM

#

the statistics module provides functions for both

pearl river Jul 20, 2020, 12:46 PM

#

oh, of course stdlib has a statistics module 😩

obsidian bough Jul 20, 2020, 12:59 PM

#

Python syntax vs Kivy or Flask...
What is a examples of ONLY PYTHON CODE vs Kivy code or Flask? Things that IS PYTHON, vs a add on

radiant fulcrum Jul 20, 2020, 1:02 PM

#

what?

#

Flask is only python code

#

Kivy is a gui module
flask is a web framework

#

your line of question is a bit squiffed

grave jolt Jul 20, 2020, 1:25 PM

#

@strange fog I agree, the 1 in {0,1} in {0,1} being False made me heavily scratch my head, it's not obvious at all

minor sinew Jul 20, 2020, 1:25 PM

#

is that actually valid syntax? lol

spice pecan Jul 20, 2020, 1:26 PM

#

apparently yeah, expands into 1 in {0, 1} and {0, 1} in {0, 1}

grave jolt Jul 20, 2020, 1:26 PM

#

Yes, because in, just like < > <= >= == !=, is treated as a comparison operator

full jay Jul 20, 2020, 1:26 PM

#

Huh

spice pecan Jul 20, 2020, 1:26 PM

#

just like other chained comparison operators

full jay Jul 20, 2020, 1:26 PM

#

That's.... huh

spice pecan Jul 20, 2020, 1:26 PM

#

That's jank

full jay Jul 20, 2020, 1:26 PM

#

I didn't know you could do it with in twice

#

Thought it would freak out

minor sinew Jul 20, 2020, 1:26 PM

#

ya lmao it seems so boilerplate

spice pecan Jul 20, 2020, 1:27 PM

#

The output of __contains__ is also force-casted into a boolean if it's not invoked directly, unlike probably every other operator

grave jolt Jul 20, 2020, 1:28 PM

#

__eq__ as well, I believe

#

and other comparison operators

#

lol wtf

📎 unknown.png

#

excuse mee

📎 unknown.png

full jay Jul 20, 2020, 1:29 PM

#

Well yes

#

That makes sense

spice pecan Jul 20, 2020, 1:29 PM

#

eq, lt and the rest of comparison gang™ doesn't get forced into a boolean

#

but contains does

grave jolt Jul 20, 2020, 1:30 PM

#

📎 unknown.png

full jay Jul 20, 2020, 1:30 PM

#

That's so friggin weird

spice pecan Jul 20, 2020, 1:30 PM

#

But only if you do in, if you invoke A.__contains__ directly it returns 42 as expected

full jay Jul 20, 2020, 1:30 PM

#

I THINK I might know why?

spice pecan Jul 20, 2020, 1:30 PM

#

Why would that be?

strange fog Jul 20, 2020, 1:31 PM

#

does it matter if it's a standalone expression vs when used in an if?

grave jolt Jul 20, 2020, 1:31 PM

#

no

full jay Jul 20, 2020, 1:31 PM

#

Well, for the regular comparison ops, we can look at pathlib for why they might want you to customize what the symbols do

#

Example being they changed the division symbol from being, well, division to concatenating the paths together

grave jolt Jul 20, 2020, 1:31 PM

#

Well, division is a general-purpose operation

full jay Jul 20, 2020, 1:32 PM

#

But in is one of those situations where it's explicitly going to have to be a boolean. Much like how we typically want to do is_even() or something when writing functions

#

It's expected

#

But I'm also just pulling this out of thin air, sooooo

grave jolt Jul 20, 2020, 1:32 PM

#

And why wouldn't it be expected for == or < 🤔 ?

full jay Jul 20, 2020, 1:33 PM

#

I could see < being used for some weird piping

#

Not sure about equals

grave jolt Jul 20, 2020, 1:33 PM

#

same

#

also

#

A trick I read about on stackoverflow for C++:

strange fog Jul 20, 2020, 1:33 PM

#

i could imagine returning some weird 'maybe'like value for ==

#

to get approximate equality

#

if you want to get weird :p

full jay Jul 20, 2020, 1:33 PM

#

I'd love to see a ~= operator then

#

"I mean it's KIND of close"

spice pecan Jul 20, 2020, 1:34 PM

#

Isn't ~= technically already a thing

full jay Jul 20, 2020, 1:34 PM

#

Shhhh

grave jolt Jul 20, 2020, 1:34 PM

#

oh

full jay Jul 20, 2020, 1:35 PM

#

I forgot that it was

grave jolt Jul 20, 2020, 1:35 PM

#

yes

spice pecan Jul 20, 2020, 1:35 PM

#

wait no

modern night Jul 20, 2020, 1:35 PM

#

don't think it is, the other binary operators have assignment tho, it isn't in 3.7 at least

spice pecan Jul 20, 2020, 1:35 PM

#

~ is unary

#

it can't have a ~= form

strange fog Jul 20, 2020, 1:35 PM

#

~= is not a thing in python

#

^

full jay Jul 20, 2020, 1:35 PM

#

Was going to say, I didn't think I'd seen it before

strange fog Jul 20, 2020, 1:35 PM

#

i think there was a pep for it though

spice pecan Jul 20, 2020, 1:35 PM

#

My brain is lagging, mb

strange fog Jul 20, 2020, 1:35 PM

#

at some point

spice pecan Jul 20, 2020, 1:35 PM

#

There was a pep, yeah

#

using it as a shortcut to .is_close

grave jolt Jul 20, 2020, 1:36 PM

#

There's also an undocumented operator for extracting the value from a single-value container.

#

,=

#

!e

x = [42]
y ,= x
print(y)

fallen slateBOT Jul 20, 2020, 1:36 PM

#

@grave jolt :white_check_mark: Your eval job has completed with return code 0.

grave jolt Jul 20, 2020, 1:36 PM

#

:)

spice pecan Jul 20, 2020, 1:36 PM

#

oh wow

full jay Jul 20, 2020, 1:36 PM

#

Thanks, I hate it

unkempt rock Jul 20, 2020, 1:36 PM

#

I'm going to find some use for this

spice pecan Jul 20, 2020, 1:37 PM

#

trailing commas are confusing

full jay Jul 20, 2020, 1:37 PM

#

That feels really weird

#

Why not just y = *x?

#

That's at least a recognized syntax

grave jolt Jul 20, 2020, 1:37 PM

#

that doesn't work

full jay Jul 20, 2020, 1:37 PM

#

Wait, seriously?

#

I mean I don't doubt you

grave jolt Jul 20, 2020, 1:37 PM

#

nope

spice pecan Jul 20, 2020, 1:37 PM

#

Can't use starred expression here

full jay Jul 20, 2020, 1:37 PM

#

So duuuumb

spice pecan Jul 20, 2020, 1:37 PM

#

w o w

full jay Jul 20, 2020, 1:38 PM

#

It's literally just to save 2 keystrokes, then

#

To keep from having to do y = x[0]

spice pecan Jul 20, 2020, 1:38 PM

#

what if it doesn't support indexing

full jay Jul 20, 2020, 1:38 PM

#

Shit

#

Didn't think of that

#

Wait, actually

#

Does that work on dicts?

#

The ,= thing?

grave jolt Jul 20, 2020, 1:39 PM

#

of course

strange fog Jul 20, 2020, 1:39 PM

#

wasnt there a thing about not being able to unpack in return values either?

#

without an explicit tuple

spice pecan Jul 20, 2020, 1:39 PM

#

It just forces the left-hand side to be a tuple of one element, it's a trailing comma

strange fog Jul 20, 2020, 1:39 PM

#

thing that got addressed at some point

full jay Jul 20, 2020, 1:39 PM

#

My brain hurts

grave jolt Jul 20, 2020, 1:39 PM

#

no, it's an undocumented operator!!

#

😠

spice pecan Jul 20, 2020, 1:39 PM

#

~~,= is like x --> 0 in C, it's just abusing the spacing to make it look like something else~~ yes it's a secret operator

full jay Jul 20, 2020, 1:40 PM

#

You might call it a smooth operator

grave jolt Jul 20, 2020, 1:40 PM

#

It's a double-barrel sawn-off

unkempt rock Jul 20, 2020, 1:40 PM

#

print(“UwU”)

grave jolt Jul 20, 2020, 1:40 PM

#

,=

unkempt rock Jul 20, 2020, 1:40 PM

#

UwU

full jay Jul 20, 2020, 1:40 PM

#

@unkempt rock Is there a reason you keep spamming that here?

unkempt rock Jul 20, 2020, 1:40 PM

#

no

full jay Jul 20, 2020, 1:40 PM

#

Then I would ask that you don't

unkempt rock Jul 20, 2020, 1:40 PM

#

uwu?

full jay Jul 20, 2020, 1:40 PM

#

As in don't interrupt an active conversation if you have nothing to contribute to said conversation

strange fog Jul 20, 2020, 1:41 PM

#

def example_2():
    items = [1, 2, 3, 4]
    return 0, *items

yep looks like this got fixed in 3.8. 3.7 and below youd get a syntax error on the return line

full jay Jul 20, 2020, 1:42 PM

#

That's funky

spice pecan Jul 20, 2020, 1:42 PM

#

Ok, so what would you deem more readable, [y] = x, y, = x, (y,) = x or y ,= x?

full jay Jul 20, 2020, 1:42 PM

#

2

#

er, index 1

spice pecan Jul 20, 2020, 1:43 PM

#

[0] and [2] are the most explicit it seems, [1] looks less noisy but still readable, [3] looks like something you'd use to impress others

full jay Jul 20, 2020, 1:44 PM

#

It's weird to me that [0] still works on ones that aren't lists

#

I mean I know that's probably silly

spice pecan Jul 20, 2020, 1:44 PM

#

That's a fair point tbh

strange fog Jul 20, 2020, 1:44 PM

#

none of the above. y = x[0] all day :p

full jay Jul 20, 2020, 1:45 PM

#

Won't work on things that don't allow you to use indexes

#

Sets for example

strange fog Jul 20, 2020, 1:45 PM

#

oh true

#

hmm

spice pecan Jul 20, 2020, 1:45 PM

#

Type {InsertName} is not subscriptable

full jay Jul 20, 2020, 1:45 PM

#

(as I got pointed out to me earlier)

strange fog Jul 20, 2020, 1:45 PM

#

ah yes i was busy looking up the return unpacking error thing

full jay Jul 20, 2020, 1:45 PM

#

Interesting thing

strange fog Jul 20, 2020, 1:45 PM

#

didnt see that part of the convo

full jay Jul 20, 2020, 1:45 PM

#

No no, I was more just saying I was wrong, too

#

So if you do [y] = x where x is a dictionary, it'll give you the key

strange fog Jul 20, 2020, 1:46 PM

#

petition to make generator subscriptable? 😄

full jay Jul 20, 2020, 1:46 PM

#

And just the key

strange fog Jul 20, 2020, 1:46 PM

#

or all iterables rather

#

if they dont have indexing just iter() + next until you get the desired value

#

internally

spice pecan Jul 20, 2020, 1:46 PM

#

On a side note, the return unpacking is still funky in 3.8

strange fog Jul 20, 2020, 1:47 PM

#

oh is it

spice pecan Jul 20, 2020, 1:47 PM

#

>>> def func():
...     return *(i for i in range(5))
...
  File "<stdin>", line 2
SyntaxError: can't use starred expression here' # because discord markup doesn't like me
>>> def func():
...     return -1, *(i for i in range(5))
...
>>> func()
(-1, 0, 1, 2, 3, 4)```

strange fog Jul 20, 2020, 1:47 PM

#

oh dear

minor sinew Jul 20, 2020, 1:47 PM

#

at first glance, i could swear you're returning a pointer lol

strange fog Jul 20, 2020, 1:47 PM

#

yeah that doesnt work in 3.9 either

spice pecan Jul 20, 2020, 1:47 PM

#

idk why I went with i for i in range(5) instead of just *range(5)

#

small brain moment right here

crisp crystal Jul 20, 2020, 1:48 PM

#

!e

x = [42]
y ,= x
print(y)

@grave jolt How do you type the text with the rectangular outline and opaque background?

grave jolt Jul 20, 2020, 1:48 PM

#

!code

fallen slateBOT Jul 20, 2020, 1:48 PM

#

Discord has support for Markdown, which allows you to post code with full syntax highlighting. Please use these whenever you paste code, as this helps improve the legibility and makes it easier for us to help you.

To do this, use the following method:

```python
print('Hello world!')
```

Note:
• These are backticks, not quotes. Backticks can usually be found on the tilde key.
• You can also use py as the language instead of python
• The language must be on the first line next to the backticks with no space between them

This will result in the following:

print('Hello world!')

crisp crystal Jul 20, 2020, 1:48 PM

#

Thanks

grave jolt Jul 20, 2020, 1:48 PM

#

speaking of < >

#

!e

class _AppliedOperator:
    def __init__(self, value, fn):
        self.value = value
        self.fn = fn

    def __gt__(self, other):
        return self.fn(self.value, other)

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

    def __gt__(self, other):
        return _AppliedOperator(other, self.fn)

approx = CustomOperator(lambda a, b: abs(a - b) < 0.001)

print(0.123456 <approx> 0.123457)

fallen slateBOT Jul 20, 2020, 1:48 PM

#

@grave jolt :white_check_mark: Your eval job has completed with return code 0.

<__main__._AppliedOperator object at 0x7f172481d490>

grave jolt Jul 20, 2020, 1:48 PM

#

I don't understand the output

#

it should be True

#

And actually, (0.123456 <approx)> 0.123457 is True

spice pecan Jul 20, 2020, 1:49 PM

#

and and or return the operand that matches its truthiness by short-circuiting the logic

grave jolt Jul 20, 2020, 1:49 PM

#

oh, right

#

well, then some hacks will be needed

spice pecan Jul 20, 2020, 1:50 PM

#

Although I do have to compliment the <opname> idea, it's really creative

grave jolt Jul 20, 2020, 1:50 PM

#

I stole it from a stackoverflow answer on C++

spice pecan Jul 20, 2020, 1:52 PM

#

Ah, SO strikes again

grave jolt Jul 20, 2020, 1:54 PM

#

!e

_STACK = []

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

    def __gt__(self, other):
        if len(_STACK) % 2 == 0:
            _STACK.append((self.fn, other))
            return True
        else:
            fn, left = _STACK.pop()
            return fn(left, other)

approx = CustomOperator(lambda a, b: abs(a - b) < 0.001)
print(0.123456 <approx> 0.123457)

avg = CustomOperator(lambda a, b: (a + b)/2)
print(1 <avg> 4)

fallen slateBOT Jul 20, 2020, 1:54 PM

#

@grave jolt :white_check_mark: Your eval job has completed with return code 0.

001 | True
002 | 2.5

spice pecan Jul 20, 2020, 1:55 PM

#

I'm definitely ~~stealing~~ appropriating <opname> for my hobby projects, that's really neat

grave jolt Jul 20, 2020, 1:55 PM

#

well, it completely screws up when you don't put brackets

#

and even if you do...

flat gazelle Jul 20, 2020, 2:54 PM

#

what would you think about a total comparison operator like <=> which would return LT EQ GT, either as constants like True/False or -1 0 1?

peak spoke Jul 20, 2020, 2:55 PM

#

something like https://docs.python.org/2.7/reference/datamodel.html#object.__cmp__ ?

visual shadow Jul 20, 2020, 2:56 PM

#

I like the idea of a 3 way flag honestly

#

Though the symbol confuses me.

pearl river Jul 20, 2020, 2:57 PM

#

it would be kinda nice. Currently, it's mostly possible with something like sgn(a-b) if they are number-like, or some ifs if not.

flat gazelle Jul 20, 2020, 2:58 PM

#

that symbol I just took from C++

grave jolt Jul 20, 2020, 3:13 PM

#

I don't see any use for that, to be honest

red solar Jul 20, 2020, 3:25 PM

#

spaceship, not 3way flag

#

yay 🙂

📎 unknown.png

unkempt rock Jul 20, 2020, 3:32 PM

#

can i activate .py files from pressing a button in tkinter?

low lagoon Jul 20, 2020, 3:32 PM

#

@unkempt rock This isn't a help channel. Use #❓｜how-to-get-help to acquire help, and remember to read channel descriptions.

unkempt rock Jul 20, 2020, 3:32 PM

#

okay

brazen jacinth Jul 20, 2020, 4:20 PM

#

a = [1]
b ,= a

#

why does this exist

#

GWmythicalThonkCool

#

more importantly, how did you find out about it

light ember Jul 20, 2020, 4:41 PM

#

intresting ```>>> a = [1]
b ,= a

b
1
a
[1]

pseudo cradle Jul 20, 2020, 4:44 PM

#

huh, neat

grave jolt Jul 20, 2020, 4:52 PM

#

more importantly, how did you find out about it
our staff team has connections to the CPython core development team

brazen jacinth Jul 20, 2020, 4:53 PM

#

interesting little tidbit that i'll forget by tomorow but interesting nonetheless 😄

north root Jul 20, 2020, 4:55 PM

#

this might make it a bit more clear

(b,) = [1]

#

it's just unpacking the list

brazen jacinth Jul 20, 2020, 4:56 PM

#

thanks, makes more sense now 🤷‍♂️

#

the lack of paranthesis does things to your brain

#

the amount of times i accidentelly added a , at the end of a variable assignment 😄

north root Jul 20, 2020, 4:57 PM

#

it's also a bit odd to have the comma next to the equal sign

brazen jacinth Jul 20, 2020, 4:59 PM

#

a parsing oddity indeed

pearl river Jul 20, 2020, 4:59 PM

#

yeah, ,= sounds like it should do a ,=b == a = a,b

#

like, it's how the other <something>= operators work.

brazen jacinth Jul 20, 2020, 5:00 PM

#

, isn't an arithmetic operator though, but that would definitely be interesting

deep shard Jul 20, 2020, 5:02 PM

#

a = input('>>>')
b = {'.'}
if a == 1:
del(b[0])
b[0] = '1'
print(b)

#

why doesnt it update?

brazen jacinth Jul 20, 2020, 5:02 PM

#

this isn't a help channel, #❓｜how-to-get-help

#

i can tell you that you will need to do a=int(input()) though, as input() will always be a string

pseudo cradle Jul 20, 2020, 5:10 PM

#

Why does Pep8 hate camelCase?

#

I feel like it wouldn't be the worst thing to use camelCase either for function names or variables

#

and leave snake_case for the other

#

Fuck it. I'm using camelCase for my vars

radiant fulcrum Jul 20, 2020, 5:17 PM

#

I find camel case a bit odd tbh

#

like i dislike it because its not consistent with caps which annoys me having the lower starting letter and caps for each remaining

full jay Jul 20, 2020, 5:21 PM

#

I can kind of go either way

#

I think after doing Python for so long, my brain hesitates when I have to do camelCase and I end up taking longer to write it than pipe_case

#

Despite it being technically longer

pseudo cradle Jul 20, 2020, 5:26 PM

#

I also just like having vars/arguments be separated stylingwise from functions/methods but maybe that's just me

grave jolt Jul 20, 2020, 5:27 PM

#

Some people write functions/methods in PascalCase, and it drives me crazy a little bit

pseudo cradle Jul 20, 2020, 5:27 PM

#

and yeah maybe I'm coming from C++/matlab

#

and don't want to let go of that syntax

spice pecan Jul 20, 2020, 5:33 PM

#

camelCase looks kinda ugly to me, I prefer snake_case or PascalCase

grave jolt Jul 20, 2020, 5:33 PM

#

personally I prefer kebab-case or

`allowing spaces in identifiers come on its 2020 not 1950`

brazen jacinth Jul 20, 2020, 5:33 PM

#

eh, mostly you just kinda adapt in time with whatever code base you're working on

grave jolt Jul 20, 2020, 5:33 PM

#

true

full jay Jul 20, 2020, 5:34 PM

#

But mentally you're going "man, I wish it was this"

grave jolt Jul 20, 2020, 5:34 PM

#

And hopefully the codebase adheres to the convention of built-in names.

brazen jacinth Jul 20, 2020, 5:35 PM

#

if i would see a camel case codebase in python it would probably make me anxious 😄

#

but the less time you use to argue about personal styling conventions, the better

random path Jul 20, 2020, 5:37 PM

#

@grave jolt File "<ipython-input-5-ec7702362b26>", line 1
pygame (test)_STACK = []
^
SyntaxError: invalid syntax

brazen jacinth Jul 20, 2020, 5:37 PM

#

had a particular short internship, where alot of the code interviewing being done was based heavily on personal stylistic preference

grave jolt Jul 20, 2020, 5:37 PM

#

You copied it incorrectly @random path

brazen jacinth Jul 20, 2020, 5:37 PM

#

wasn't the best experience

#

i find pep8 a godsend, adhere to the language standard and it's readable, ciao

grave jolt Jul 20, 2020, 5:39 PM

#

Most cognitive dissonance I had was in C&&+Qt.
Legacy C names are written in withoutseparatorsandvwls-case like wcsxfrm
C++ names are in snake_case
and Qt names are in camelCase

glass quartz Jul 20, 2020, 6:36 PM

#

I have a question I REALLY need an answer to.

How different is Ada language from python?

red solar Jul 20, 2020, 6:36 PM

#

wow...

#

isn't ada a provable correct dependent typed language?

#

the only part of that sentence that applies to python is "language"

glass quartz Jul 20, 2020, 6:41 PM

#

In plain English that means....?

trail vessel Jul 20, 2020, 6:42 PM

#

Very

grave jolt Jul 20, 2020, 6:43 PM

#

@red solar no, that's Agda :)

#

https://en.wikipedia.org/wiki/Ada_(programming_language)
https://en.wikipedia.org/wiki/Agda_(programming_language)

red solar Jul 20, 2020, 6:43 PM

#

ohhhh

#

i got confused lol

#

i was close tho 🙂

tawny shoal Jul 20, 2020, 6:44 PM

#

hey what if you'd write identifiers like $my identifier$ (spaces are allowed) and it'd be a convention for every IDE to just remove the $ signs and turn the identifiers into bold text

red solar Jul 20, 2020, 6:45 PM

#

how would you unbold something?

#

also why?

tawny shoal Jul 20, 2020, 6:45 PM

#

probably would be annoying for the visually impaired

#

how would you unbold something?
@red solar e.g. placing the cursor in front of an identifier and pressing backspace

#

also why?
@red solar to cater to the demands of "special characters make code less readable" and still making identifiers easily recognizable

grave jolt Jul 20, 2020, 6:47 PM

#

Representing code as plain text might be a legacy idea we're not willing to rethink.
It requires us to resort to hacks like escaping with \.

red solar Jul 20, 2020, 6:47 PM

#

who's putting special characters in their code? wtf

#

if i see emoji characters in any code i read, i'm going to get angry

tawny shoal Jul 20, 2020, 6:48 PM

#

special characters are characters that are not alphabetical or numerical characters or punctuation

#

not just non-ascii characters

grave jolt Jul 20, 2020, 6:48 PM

#

what about ligatures?

red solar Jul 20, 2020, 6:49 PM

#

but why are these (non ascii chars) in the code in the first place?

pearl river Jul 20, 2020, 6:53 PM

#

They are allowed in variable names.

#

well, many of them. Not, say, emojis, but katakana is fair game 😛

red solar Jul 20, 2020, 6:54 PM

#

that doesn't mean people (should) use them

#

they were probably allowed to help with writing code in other languages

#

in which case it would be readable to anyone who knows the language

grave jolt Jul 20, 2020, 7:00 PM

#

I don't see a valid reason to allow special characters in variable names.

#

I once had a mysterious bug where I wrote

#

!charinfo с

fallen slateBOT Jul 20, 2020, 7:01 PM

#

Character Info

\u0441 : CYRILLIC SMALL LETTER ES - с

grave jolt Jul 20, 2020, 7:01 PM

#

instead of

#

!charinfo c

fallen slateBOT Jul 20, 2020, 7:01 PM

#

Character Info

\u0063 : LATIN SMALL LETTER C - c

grave jolt Jul 20, 2020, 7:01 PM

#

also

#

!e

Fаlse = True
Truе = False
Fаlsе = Fаlse
Fаlse = True
print(False, Fаlse, Fаlsе, Truе)

fallen slateBOT Jul 20, 2020, 7:03 PM

#

@grave jolt :white_check_mark: Your eval job has completed with return code 0.

False True True False

spice pecan Jul 20, 2020, 7:14 PM

#

what the fuck

north root Jul 20, 2020, 7:15 PM

#

uhhhh

#

magic ✨

spice pecan Jul 20, 2020, 7:16 PM

#

Why are True and False allowed to be var names

north root Jul 20, 2020, 7:17 PM

#

they aren't

flat gazelle Jul 20, 2020, 7:18 PM

#

!charinfo Fаlse

fallen slateBOT Jul 20, 2020, 7:18 PM

#

Character Info

\u0046 : LATIN CAPITAL LETTER F - F
\u0430 : CYRILLIC SMALL LETTER A - а
\u006c : LATIN SMALL LETTER L - l
\u0073 : LATIN SMALL LETTER S - s
\u0065 : LATIN SMALL LETTER E - e

Raw

\u0046\u0430\u006c\u0073\u0065

flat gazelle Jul 20, 2020, 7:18 PM

#

it is not False, but Fаlse

spice pecan Jul 20, 2020, 7:19 PM

#

oh

#

I should've guessed

#

I've been using that to bypass chatfilters for a loooong while bamboozle

red solar Jul 20, 2020, 7:37 PM

#

https://www.reddit.com/r/rust/comments/5penft/parallelizing_enjarify_in_go_and_rust/dcsgk7n/ sorry back to the special characters thing, i forgot to link this

paper echo Jul 20, 2020, 9:02 PM

#

!e

import unicodedata
print(unicodedata.normalize('NFKD', '\u0046\u0430\u006c\u0073\u0065'))

fallen slateBOT Jul 20, 2020, 9:02 PM

#

@paper echo :white_check_mark: Your eval job has completed with return code 0.

Fаlse

paper echo Jul 20, 2020, 9:02 PM

#

🤦 not sure what i expected

#

!e

import unicodedata
print(unicodedata.normalize('NFKD', '\u0046\u0430\u006c\u0073\u0065').encode('utf-8'))

fallen slateBOT Jul 20, 2020, 9:02 PM

#

@paper echo :white_check_mark: Your eval job has completed with return code 0.

b'F\xd0\xb0lse'

paper echo Jul 20, 2020, 9:03 PM

#

hm. interesting implications for language design

#

even if you force variable names into NFKD you don't solve the problem entirely

#

what if we had something like IDN for variable names

#

xn!flse_53d = 10
assert locals()['\u0046\u0430\u006c\u0073\u0065'] == 10

something like this, idk what the syntax would actually be

#

then again nooobody wants to write punycode variable names by hand

red solar Jul 20, 2020, 9:22 PM

#

nobody wants to read this horror either 😂

boreal umbra Jul 20, 2020, 9:43 PM

#

You know how you can get a function's signature with inspect? Is that an attribute of the function object itself?

#

I think I tried finding it once with dir and didn't see anything interesting.

flat gazelle Jul 20, 2020, 9:43 PM

#

it is in function.__code__ somewhere afaik

paper echo Jul 20, 2020, 9:44 PM

#

yep f.__code__.co_*

boreal umbra Jul 20, 2020, 9:44 PM

#

I wonder if it was something like function.__sig__, would that result in poorly conceived meta programming?

paper echo Jul 20, 2020, 9:44 PM

#

oh f.__annotations__

flat gazelle Jul 20, 2020, 9:44 PM

#

if type hints were used, you can also see __annotations__

#

In [68]: def fun(a, b, c): ...

In [69]: fun.__annotations__
Out[69]: {}

In [70]: fun.__code__.co_argcount
Out[70]: 3

boreal umbra Jul 20, 2020, 9:47 PM

#

that's not as fun as all the information you get from inspect

flat gazelle Jul 20, 2020, 9:49 PM

#

In [74]: def fun(a, b, c):
    ...:     d = a + b
    ...:     return e
In [75]: {k: getattr(fun.__code__, k) for k in dir(fun.__code__) if k[:2] == 'co'}
Out[75]:
{'co_argcount': 3,
 'co_cellvars': (),
 'co_code': b'|\x00|\x01\x17\x00}\x03t\x00S\x00',
 'co_consts': (None,),
 'co_filename': '<ipython-input-74-a5ec6893f41b>',
 'co_firstlineno': 1,
 'co_flags': 8259,
 'co_freevars': (),
 'co_kwonlyargcount': 0,
 'co_lnotab': b'\x00\x01\x08\x01',
 'co_name': 'fun',
 'co_names': ('e',),
 'co_nlocals': 4,
 'co_posonlyargcount': 0,
 'co_stacksize': 2,
 'co_varnames': ('a', 'b', 'c', 'd')}
```this is all you get

boreal umbra Jul 20, 2020, 9:51 PM

#

🤯

native lynx Jul 20, 2020, 10:28 PM

#

ok so after a lot of research i have come to conclusion i have no idea how bits work in python.

def LAST_BITS(k,n):
    return ((k) & ((1<<(n))-1))

def BitSlice(k,m,n):
    return LAST_BITS( ((k)>>(m)),((n)-(m)) )

SHA256="d632b3049081b413ecceed6fb80d56657b1c43a51ecad3e5ad02a6e720eced01"
BitSlice(SHA256,120,136)

this however returns various errors depending on if i pass it a "string". how do i convert this hex string to a malleable set of bits to i can return a subset of them?

pearl river Jul 20, 2020, 10:29 PM

#

def LAST_BITS(k,n):
    return ((k) & ((1<<(n))-1))

this takes and returns an int.

#

The string you can parse as a hexadecimal int if you really want.

In [90]: int("d632b3049081b413ecceed6fb80d56657b1c43a51ecad3e5ad02a6e720eced01",16)
Out[90]: 96884527485712115404056311403563845911353878330075242211070609857748849716481

native lynx Jul 20, 2020, 10:29 PM

#

ah ok i thoughts that where i went wrong.
ValueError: invalid literal for int() with base 10: b'\xd62\xb3\
casting the hex string litterally just gives me an error tho

pearl river Jul 20, 2020, 10:30 PM

#

not sure how to work with bytes like you want, though.

flat gazelle Jul 20, 2020, 10:30 PM

#

If you want bit sequences, your options are int or bytes or byarray

#

str is a sequence of Unicode characters, not of bits or bytes

native lynx Jul 20, 2020, 10:30 PM

#

O_e but at some level they are all just bits at heart

flat gazelle Jul 20, 2020, 10:31 PM

#

Yes, but not at a level you should be accessing.

native lynx Jul 20, 2020, 10:31 PM

#

lol indeed. trying to convert them over to a 16bit number for a color

flat gazelle Jul 20, 2020, 10:32 PM

#

Ye, use an int as per ConfusedReptile

native lynx Jul 20, 2020, 10:32 PM

#

thanks guys for the input 🙂 ill try that Reptile and see where i get. its been a frustrating night. but i think thats what i was missing

native lynx Jul 20, 2020, 10:51 PM

#

is there a way to print all characters (including hidden) its working as expected in a stand alone environment but the minute i pull the hashes from a sha256sum.txt file its giving me an invalid litterall. the only thing i can comeup with at this point is im passing unsensitized data thats not in the hex string format thats expected

pearl river Jul 20, 2020, 10:52 PM

#

you can just print it, or print(repr(text))

native lynx Jul 20, 2020, 10:57 PM

#

wow lol it may be time to take a break for a while. i just found hash = bytes.fromhex(hash_str) this just hanging out causing issues. its always the details the get you. >..<

#

BINGO 😛 thank you ConfusedReptile thats exactly what i did wrong. ^ i had that in there trying to work around the problem and forgot to take it out.

velvet gorge Jul 21, 2020, 12:46 AM

#

i cant type in #python-discussion

shy vine Jul 21, 2020, 12:46 AM

#

Yes you can

velvet gorge Jul 21, 2020, 12:46 AM

#

📎 unknown.png

#

stuck there forever

shy vine Jul 21, 2020, 12:47 AM

#

This is the wrong channel

vivid crow Jul 21, 2020, 1:27 AM

#

How do I run a function of a class, inside a different class?

#

ClassName.functionName doesn't seem to work

spark magnet Jul 21, 2020, 1:30 AM

#

do you have an object of the other class?

vivid crow Jul 21, 2020, 1:30 AM

#

you mean make it into an object but assigning it to a variable?

#

i'm new to python

near coral Jul 21, 2020, 1:31 AM

#

anyone using Python 3.9 Beta?

vivid crow Jul 21, 2020, 1:32 AM

#

no

#

3.9

spark magnet Jul 21, 2020, 1:36 AM

#

@near coral I have used it in tests

#

@vivid crow Can you show us the code you are worrking with? What is "the other class"?

near coral Jul 21, 2020, 1:38 AM

#

oh nice

#

I am interested in this new parser they will be releasing

#

although i don't fully understand the documentation for it

spark magnet Jul 21, 2020, 1:38 AM

#

i haven't looked into that at all

near coral Jul 21, 2020, 1:39 AM

#

oh ok

#

how risk is it to use the beta at this point?

spark magnet Jul 21, 2020, 1:40 AM

#

it's practically done. rc1 is slated for aug 10th or something

near coral Jul 21, 2020, 1:41 AM

#

oh wow

#

I might install the release candidate then

#

Hoping to test this out. I have my own function for this, but I trust python's more for removing prefixes and suffixes within a string

📎 unknown.png

boreal umbra Jul 21, 2020, 3:44 AM

#

So when you import a module, that whole module gets run, even if you're only importing certain objects, yes?

#

And subsequent imports don't cause the module to be re-executed? How is this managed as far as the reference count?

unkempt rock Jul 21, 2020, 3:49 AM

#

@worldly venture

worldly venture Jul 21, 2020, 3:49 AM

#

!ban 734970770333368406 Low effort troll

fallen slateBOT Jul 21, 2020, 3:49 AM

#

:incoming_envelope: :ok_hand: applied ban to @grizzled basin permanently.

#

:incoming_envelope: :ok_hand: applied mute to @gaunt light until 2020-07-21 04:01 (9 minutes and 59 seconds) (reason: mentions rule: sent 11 mentions in 10s).

velvet fossil Jul 21, 2020, 3:55 AM

#

Is anyone available tomorrow to help me on my collage python course?

#

Im falling behind cause I cant seem to add things together 🤦🏽‍♂️

unkempt rock Jul 21, 2020, 3:56 AM

#

im not sure bout homework help

#

dont think thats allowed

raven ridge Jul 21, 2020, 3:56 AM

#

@boreal umbra yes, that's correct. The first import adds the module to sys.modules, and the next time it's imported the existing module that's in sys.modules is returned instead of reexecuting the module's code

boreal umbra Jul 21, 2020, 3:57 AM

#

@velvet fossil you can get help with school-related work in the help channels as long as you're focusing on learning things conceptually; you can't get code written for you here.

velvet fossil Jul 21, 2020, 3:58 AM

#

@boreal umbra thanks for letting me know, yeah exactly what im in need of, as long as I understand the concept I'd be able to get the work done

raven ridge Jul 21, 2020, 3:59 AM

#

As far as reference count goes, the module itself is an object with a reference count. Import it once and it has a reference count of 2 (you're holding a reference to it, and so is the sys.modules dictionary. Import it again from another module and the reference count goes up to 3

unkempt rock Jul 21, 2020, 3:59 AM

#

helloo could i take a 'variable' from un txt file

#

?

boreal umbra Jul 21, 2020, 3:59 AM

#

@raven ridge is sys.modules in part for ensuring that each module has a non-zero reference count?

#

@unkempt rock take a look at #❓｜how-to-get-help

unkempt rock Jul 21, 2020, 4:00 AM

#

ok

boreal umbra Jul 21, 2020, 4:00 AM

#

sorry I was already typing that when you answered

#

yes, I see

raven ridge Jul 21, 2020, 4:00 AM

#

It's for caching already-loaded modules so that they can be reused...

#

So, yeah, I guess you could put it that way

boreal umbra Jul 21, 2020, 4:01 AM

#

I said "in part" since caching them would presumably be the primary goal.

raven ridge Jul 21, 2020, 4:01 AM

#

Yep.

#

But caching them necessitates keeping them alive, so there's a reference held to make that happen

boreal umbra Jul 21, 2020, 4:04 AM

#

I suppose that also means that manually deleting an element from sys.modules doesn't guarantee that it will be reloaded if you try to import it again, since it might still be alive via another reference.

#

(Not that you should write code that depends on modules being reloaded.)

raven ridge Jul 21, 2020, 4:06 AM

#

Au contraire - if you delete it fromsys.modules the next import will search for and load the module again.

#

And, if someone still had an existing reference, now you've got two different copies of that module used by different modules.

boreal umbra Jul 21, 2020, 4:07 AM

#

🤯

raven ridge Jul 21, 2020, 4:08 AM

#

When import is doing it's thing, it checks in sys.modules to see if it's previously been loaded, but if it's not found there, it doesn't have any way to know that it was already loaded

#

There's not some other cache it can check, nor can it reasonably inspect every object to see if it refers to that module.

boreal umbra Jul 21, 2020, 4:09 AM

#

Makes sense

raven ridge Jul 21, 2020, 4:11 AM

#

And if you do manage to get 2 copies of the same module loaded things can break in surprising ways, so, ya know, try to avoid it 😂

boreal umbra Jul 21, 2020, 4:11 AM

#

I was once working with spaCy, which is written in such a way that mutations to certain objects silently affect the whole module, so my unit tests inexplicably failed because they relied on things that happened in other tests not having happened.

#

Which is really annoying on their part.

raven ridge Jul 21, 2020, 4:13 AM

#

Indeed, that's far from ideal. There's a pytest plugin that runs each test in a separate processes that can help in situations like that, fwiw

boreal umbra Jul 21, 2020, 4:13 AM

#

good to know

#

I don't know if I'll be working on that project in the future

tawny shoal Jul 21, 2020, 4:53 AM

#

aw damn

lofty spear Jul 21, 2020, 5:01 AM

#

Guys, is there any website which offers some sort of decent python challenges or questions to practice and improve skills?

#

Excluding paid options

merry zodiac Jul 21, 2020, 5:24 AM

#

here is a simple little game i made using python ```python

word = input("Select the secret word!")
guess_count = 0
while True:
guess = input("Guess the secret word! ")
if guess.lower() == word.lower():
print("You win!")
break
elif guess.lower() != word.lower():
print("You have guessed incorrectly try again!")
guess_count += 1
print("Your amount of guesses is " + str(guess_count))
if guess_count > 3:
print("You lose!")``` its a guessing game

#

its not much but its honest work

unkempt rock Jul 21, 2020, 5:30 AM

#

print(word)

#

ez

merry zodiac Jul 21, 2020, 5:34 AM

#

oh go

#

d

#

oh fuck

#

you rascal

sullen widget Jul 21, 2020, 5:41 AM

#

Rotund, if you win and break the loop you will never reach the elif, so you don't need to check if guess.lower() != word.lower()

#

But good job

merry zodiac Jul 21, 2020, 5:49 AM

#

@sullen widget while its under the if that code is needed

#

its to tell you if you lose or not and the amount of guesses you have used

sullen widget Jul 21, 2020, 5:50 AM

#

You can use a for loop for that

unkempt rock Jul 21, 2020, 7:04 AM

#

Peoplew

#

people*

#

if i use the terminal version of python without any IDE

brazen jacinth Jul 21, 2020, 7:05 AM

#

Not the right channel, #❓｜how-to-get-help

unkempt rock Jul 21, 2020, 7:05 AM

#

How can i make "if then" stuff?

#

oh ok

stiff finch Jul 21, 2020, 7:11 AM

#

hey everyone i'm new here & also a little above than beginner. Is there any site or something where i can find a lot of realistic projects (not into web development) ? My main focus is on learning & gaining some knowledge please

sullen widget Jul 21, 2020, 7:13 AM

#

Well i would suggest just reading code from github since there is a lot of Python code there

#

Maybe even read python source code for the standard libraries lol

stiff finch Jul 21, 2020, 7:13 AM

#

okay

#

but only reading won't help right ?

sullen widget Jul 21, 2020, 7:14 AM

#

Yeah, try solving stuff in a different way than what they did

urban forum Jul 21, 2020, 7:14 AM

#

Hey, I wanna know can I automate data entry with python

stiff finch Jul 21, 2020, 7:15 AM

#

Hey, I wanna know can I automate data entry with python
@urban forum ofc you can

#

read selenium & requests module

#

Yeah, try solving stuff in a different way than what they did
@sullen widget ok gonna do that then

finite girder Jul 21, 2020, 8:17 AM

#

📎 unknown.png

#

hi, I am trying to find the definition of a function from a package I downloaded

#

thing is, I used n++ to search all files for the name of the function, It just returns a couple of lines calling it, but no decleration/definition

#

there is a .pyc binary which has a string referencing it, I am guessing that is the definition, but it's been compiled ?

languid dagger Jul 21, 2020, 8:30 AM

#

@finite girder This is not a help channel. It is intended for discussion of the Python language itself

finite girder Jul 21, 2020, 8:53 AM

#

ok

queen acorn Jul 21, 2020, 9:34 AM

#

re is a standard library module, right?

undone hare Jul 21, 2020, 9:34 AM

#

Yup

queen acorn Jul 21, 2020, 9:34 AM

#

Cheers :)

unkempt rock Jul 21, 2020, 11:10 AM

#

is there a way to hash a str without using a third party library? (Using Built in Library)

brazen jacinth Jul 21, 2020, 11:12 AM

#

there's a builtin func

#

hash()

unkempt rock Jul 21, 2020, 11:13 AM

#

noice

brazen jacinth Jul 21, 2020, 11:13 AM

#

i imagine it's not cryptographically secure, but you know, it'll do for your purposes probably?

torpid plinth Jul 21, 2020, 11:17 AM

#

hey guys, is this fizz buzz program correct?

#

number = 0

while number < 1000 :
    number += 1
    if number%2 != 0:
        print(f"fizz {number}")
    elif number % 2 == 0  :
        print(f"buzz {number}")

#

oh never mind that just shows which numbers are even or odd

brazen jacinth Jul 21, 2020, 11:19 AM

#

not the right channel, but fizzbuzz (is div by 3, is div by 5, and is div by 15) afaik?#❓｜how-to-get-help

torpid plinth Jul 21, 2020, 11:20 AM

#

ight

sullen widget Jul 21, 2020, 11:37 AM

#

Brunckek, there are collisions with the hash function but it is not that common, atleast not for strings

brazen jacinth Jul 21, 2020, 11:38 AM

#

like with most hash functions 🤷‍♂️

peak spoke Jul 21, 2020, 11:38 AM

#

Python provides hashlib if you need hashes that are unique, hashes of strings are random between processes and the hash func is more about hashing speed

unkempt rock Jul 21, 2020, 12:12 PM

#

I tried a = hash("A")output everytime I tried to run it was different

radiant fulcrum Jul 21, 2020, 12:13 PM

#

thats because that how it works

brazen jacinth Jul 21, 2020, 12:17 PM

#

i get the same one everytime

#

in the same interpreter loop anyway

#

gueesing that's the way the interpreter loop is setup

#

starts by initializing the hashing functions in some way

radiant fulcrum Jul 21, 2020, 12:20 PM

#

its different between processes (hash() that is)

brazen jacinth Jul 21, 2020, 12:20 PM

#

oh, he probably created a file and ran it everytime, heh

#

yea, in each process it will be different, that's how it's meant

radiant fulcrum Jul 21, 2020, 12:20 PM

#

which is why if you need to to be the same every time you want hashlib

dusk yacht Jul 21, 2020, 12:23 PM

#

Why is hash of two ints same?

modern night Jul 21, 2020, 12:24 PM

#

for small ints, their hash is their value

#

(with inbuilt hash)

#

except for -1, that I never understood, is their a reason why hash(-1) returns -2

dusk yacht Jul 21, 2020, 12:25 PM

#

Yes but for some reason either hash(0) and hash(-1) or hash(-1) and hash(-2) are same

#

That's exactly what i meant

brave badger Jul 21, 2020, 12:26 PM

#

iirc there's no -1 hash because of an implementation detail on CPython

#

Something about it being used to represent failure states

modern night Jul 21, 2020, 12:27 PM

#

ah i guess that makes sense, never thought of that just thought there was some weird bug

dusk yacht Jul 21, 2020, 12:28 PM

#

Same here. Doesn't it interfere with dict lookups?

modern night Jul 21, 2020, 12:30 PM

#

don't think so because for 2 hashes to over ride each other they need to compare to equal with == and their hashes (hash()) need to be equal so values like 1, 1.0 and True collide, thats why you get ```py

{1: 1, 1.0: 1.0, True: True}
{1: True}```

flat gazelle Jul 21, 2020, 12:34 PM

#

you can also add 1+0j to that

peak spoke Jul 21, 2020, 12:37 PM

#

Hashes for strings are randomised because of some vulnerabilities it exposed, otherwise it should always be the same (at least on that python version)

grave jolt Jul 21, 2020, 1:48 PM

#

Hashes for strings are coherent (a == b => hash(a) == hash(b)), but the algorithm is randomized on each run. If it was the same every time, it would be possible to find a long sequence of strings that causes a hash collision every time (two consecutive strings have the same hash mod something), and it would be possible to exploit this behaviour to make filling a dict with n values take O(n^2) time.

flat gazelle Jul 21, 2020, 2:06 PM

#

I wonder if other languages deal with this

red solar Jul 21, 2020, 2:07 PM

#

i don't think other languages randomize hash functions

#

especially compiled ones

full jay Jul 21, 2020, 2:16 PM

#

Wonder what JS does...

flat gazelle Jul 21, 2020, 2:18 PM

#

https://v8.dev/blog/hash-flooding found this link

red solar Jul 21, 2020, 2:22 PM

#

how interesting

narrow kettle Jul 21, 2020, 2:38 PM

#

C# apparently does both a ref compare and a value compare, i had to double check that in the c# server tho 😂

grave jolt Jul 21, 2020, 2:53 PM

#

In OCaml, you can choose whether to do that when you create a hashmap.

brazen jacinth Jul 21, 2020, 2:55 PM

#

;; though

grave jolt Jul 21, 2020, 3:13 PM

#

;; is only needed in the REPL

#

but yes, it's weird

brazen jacinth Jul 21, 2020, 3:21 PM

#

i mean i can understand it's a different mindset and all, also gotta get used to currying, but damn, functional languages are different

full jay Jul 21, 2020, 3:24 PM

#

I just have a hard time finding practical use cases for currying

#

Like I more or less get how they work but I can't seem to find when I should be using them

crystal pebble Jul 21, 2020, 3:26 PM

#

When you use a function with the same first argument a lot of the time

#

(Doesn't have to be the first argument, that just makes it easier)

full jay Jul 21, 2020, 3:27 PM

#

Sure sure, I just can't think of a situation where that's often the case. I'm sure I'll stumble across it

visual shadow Jul 21, 2020, 3:30 PM

#

I have no idea what currying is, but the conversation makes me think of things like bigrams or pair wise combinations or so on

#

Where we have a generic n gram function for example, but we are usually setting n to 2 when calling this function.

crystal pebble Jul 21, 2020, 3:31 PM

#

Let me get on my computer and I'll look for an example

flat gazelle Jul 21, 2020, 3:32 PM

#

for example, you can compose operations on lists like hs op = unique . map (+ 1) . filter even

crystal pebble Jul 21, 2020, 3:33 PM

#

curry (botUpdates newSeq) <$> takeTMVar origState``` I don't know if that helps

#

Probably not :/

brazen jacinth Jul 21, 2020, 3:33 PM

#

currying is basically from functools import partial right?

crystal pebble Jul 21, 2020, 3:34 PM

#

That's partial application. They're similar

flat gazelle Jul 21, 2020, 3:35 PM

#

currying essentially means that a functions takes a single argument and returns another function which takes the next argument, so

def add(a, b):
    return a + b
```becomes
```py
def add(a):
    def add(b):
        return a + b
    return add

brazen jacinth Jul 21, 2020, 3:37 PM

#

how often do you guys find yourself using something like this, or other similiar function-paradigm ideas in your non-functional code

#

i've been forcing myself to pick up a functional language for too long

#

but, tbh, haven't really found a decent reason why

#

just a different paradigm, or is there something more?

crystal pebble Jul 21, 2020, 3:38 PM

#

This? Not very often
Map, filter, reduce, etc a lot though

brazen jacinth Jul 21, 2020, 3:39 PM

#

can all be replaced via comprehensions though, even preffered if i recollect correctly

flat gazelle Jul 21, 2020, 3:39 PM

#

I do sometimes end up making essentially monadic error handlers

#

you cannot make reduce with a comprehension

crystal pebble Jul 21, 2020, 3:39 PM

#

List comprehensions are fine if you only have one thing to do, but when you start combining operations it becomes really messy

brazen jacinth Jul 21, 2020, 3:39 PM

#

^ fair

pearl river Jul 21, 2020, 3:40 PM

#

https://www.codewars.com/kata/53cf7e37e9876c35a60002c9

Codewars

Currying vs. Partial Application

Currying and partial application are two ways of transforming a function into another function with a generally smaller arity. While they are often confused with each other, they work differently. ...

visual shadow Jul 21, 2020, 3:40 PM

#

I googled currying and now I'm more confused than before.

crystal pebble Jul 21, 2020, 3:41 PM

#

getAttribute :: Monad m => m a -> (a -> m b) -> m b
getAttribute attr attrParser = uncurry (>>=) (attr, attrParser)```

This I would probably never do outside of FP though

brazen jacinth Jul 21, 2020, 3:41 PM

#

yea, the functional paradigm can be odd coming from oop or other conventional paradigms

#

at least the in depth stuff

flat gazelle Jul 21, 2020, 3:42 PM

#

uncurry (>>=) (attr, attrParser) why?. couldnt this just be getAttribute = (>>=)

crystal pebble Jul 21, 2020, 3:42 PM

#

I'm looking at my code right now trying to figure that out

brazen jacinth Jul 21, 2020, 3:43 PM

#

GWmemetownOMEGALUL , now i feel better being completely dumbfounded reading that

crystal pebble Jul 21, 2020, 3:43 PM

#

This was my first project ever. Man, it's embarassing looking at old code

brazen jacinth Jul 21, 2020, 3:44 PM

#

that just means you're growing :p

crystal pebble Jul 21, 2020, 3:47 PM

#

I wish I could go back in time and kill the old me. What an idiot that guy was

rigid snow Jul 21, 2020, 3:50 PM

#

Hey would people pay a webscraper money to scan for CISA alerts?

velvet fossil Jul 21, 2020, 4:28 PM

#

Is there anyone willing to help me through zoom please! Its basic python for most of you im sure

loud summit Jul 21, 2020, 4:30 PM

#

This is not a channel for asking for help with python, if you need help with python check #❓｜how-to-get-help @velvet fossil

velvet fossil Jul 21, 2020, 4:30 PM

#

@loud summit 👍

brazen jacinth Jul 21, 2020, 4:53 PM

#

this is related to the discussion at general, if you import a module inside a function, the module is tied to the function scope, yet sys.modules still includes it, but it's not available via reference in the global scope, does that mean it's imported just not available or something else?

spice pecan Jul 21, 2020, 4:53 PM

#

Imported, but there's no name referencing the module object outside of the function (and sys.modules obviously)

peak spoke Jul 21, 2020, 4:54 PM

#

sys.modules will hold all the imports you do since they're not shared across packages, if you don't import it whole that just means you don't have the name referencing it but the import mechanism is the same

brazen jacinth Jul 21, 2020, 4:55 PM

#

aha, got it

#

thanks

spice pecan Jul 21, 2020, 4:55 PM

#

import something
something = __import__('something')

these are essentially the same, not sure if that's how it works under the hood, but still

#

thus, importing inside a function does the exact same importing procedure, but assigns the module object to a local variable in that function/class

brazen jacinth Jul 21, 2020, 4:56 PM

#

python is a beginer friendly language, but damn does it have a lot of magic underneath

spice pecan Jul 21, 2020, 4:56 PM

#

It can be confusing, but its fairly consistent

brazen jacinth Jul 21, 2020, 4:57 PM

#

yea, don't mean it's not

#

but, damn, i thought i was fairly versed in the language a few years ago

#

im still not 😄

orchid sphinx Jul 21, 2020, 4:59 PM

#

Hi i have a class cup, and i want to make an inbuilt function that measures the volume of the cup, So i have a class Cup: And i want to find to get a value when i call Measure(Cup)

brazen jacinth Jul 21, 2020, 4:59 PM

#

#❓｜how-to-get-help

orchid sphinx Jul 21, 2020, 4:59 PM

#

So how do i manage to make a function that is inside the class?

#

Oh k thanks

spice pecan Jul 21, 2020, 5:00 PM

#

I'd say there isn't much magic, more hidden implementation details. Most things follow clearly established rules and don't work just because "uhh we needed that so we did some magic under the hood and there's no way for you to do the same"

brazen jacinth Jul 21, 2020, 5:00 PM

#

by magic, i meant hidden implementation details

#

by no means a small language

#

but yea i get what you mean

shrewd orbit Jul 21, 2020, 6:06 PM

#

is python IDLE bad?

unkempt rock Jul 21, 2020, 6:06 PM

#

its bad for big scale projects

#

but if u r testing sth out its fine

shrewd orbit Jul 21, 2020, 6:07 PM

#

well, i just began python coding in my school

#

they want me to take some good IDE

unkempt rock Jul 21, 2020, 6:10 PM

#

what i dont really like bout idle is that its line by line execution

full jay Jul 21, 2020, 6:10 PM

#

I mean... yes?

unkempt rock Jul 21, 2020, 6:11 PM

#

so if i mess up

full jay Jul 21, 2020, 6:11 PM

#

That's Python

brazen jacinth Jul 21, 2020, 6:11 PM

#

it's a REPL

#

that is it's purpose

#

Read Evalulate Print Loop

unkempt rock Jul 21, 2020, 6:11 PM

#

like with a text editor u can run all lines at once

#

when u tell it to run

#

like i dont like when ur running and editing at the same time

brazen jacinth Jul 21, 2020, 6:13 PM

#