BeautifulSoup.__init__() got multiple values for argument 'features' (when importing a custom BeautifulSoup wrapper function) - web-scraping

Question

Problem
I had been using this wrapper function with no issues.
def make_soup(response):
return BeautifulSoup(response.text, 'lxml')
response = requests.get('https://www.webscraper.io/test-sites/e-commerce/allinone')
soup = make_soup(response) # No Error
However, after upgrading to python 3.11 from 3.8, when importing the function and trying to use it, I get the following error:
from util import make_soup
response = requests.get('https://www.webscraper.io/test-sites/e-commerce/allinone')
make_soup(response) # Error
TypeError: BeautifulSoup.__init__() got multiple
values for argument 'features'
Attempted fix: Passing "features" as a keyword argument
I tried passing lxml as a keyword argument instead:
def make_soup(response):
return bs(response.text, features='lxml')
Initially, this appeared to fix the issue, as the function ran fine when imported.
However, I noticed that often the returned BeautifulSoup object was missing large parts of the HTML, that were not missing if I executed the same code without importing it.
Here's one more extreme example
from util import make_soup
session = requests.Session()
response = session.get('https://www.deviantart.com/') # One of the many sites I tested this on
soup = bs(response.text, 'lxml')
soup_from_imported_func = make_soup(response)
num_divs = lambda soup: len(soup.find_all('div'))
print(num_divs(soup))
print(num_divs(soup_from_imported_func))
These printed values should be equal, but they're not.
0
1120
If I print the broken soup this is all I get
<?xml version="1.0" encoding="utf-8"?>
I have tried using a different parser (html.parser and html5lib) but the issue persists.
Conclusion
As a temporary solution I can reluctantly repeat the contents of the make_soup function
However I would like to understand why this is happening, in case it generates more problems in the future.