Show value in error of list.remove

[tdegeus]

Consider

foo = ["a", "b"]
for i in ["c"]:
    foo.remove(i)

This throws with a useless error message:

Traceback (most recent call last):
  File "/Users/tdegeus/Downloads/t.py", line 3, in <module>
    foo.remove(i)
ValueError: list.remove(x): x not in list

Instead it would be very helpful to show the value of x. Now one needs extra manual debugging, or extra lines of code.

Linked PRs

• gh-96844: Improve error message of list.remove #106455
• Revert "gh-96844: Improve error message of list.remove (gh-106455)" #116956

[pochmann]

What if x is large, so large that it pushes the previous error-information off-screen or even out of the screen buffer?

What if the conversion of x to string takes a long time or causes another error? (Think of the current large int issue).

Maybe it could be done for known small good types and values.

Then again, for example set.remove already does this:

---

x = 1; {0}.remove(x) shows:

KeyError: 1

---

x = 10**1000000; {0}.remove(x) takes a long time and then prints a lot (testing with Python version without the new string conversion limit).

---

class C:
    def __repr__(self):
        raise Exception('foobar')
{0}.remove(C())

shows:

KeyError: <exception str() failed>

[ericvsmith]

Wouldn't we already have this problem with KeyError on dict lookups:

>>> {}[1234]
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
KeyError: 1234

?
I guess you could argue that the value should be removed from KeyError.

[corona10]

Python 3.13.0a0 (heads/main:217f47d6e5, Jul  6 2023, 07:20:36) [Clang 14.0.0 (clang-1400.0.29.202)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> foo = ["a", "b"]
>>> for i in ["c"]:
...     foo.remove(i)
...
Traceback (most recent call last):
  File "<stdin>", line 2, in <module>
ValueError: 'c' is not in list

Now the error message is updated, here is the rationale why I agree with the updating error message even if for the possibility of large input.

#106455 (comment)

cc @tdegeus @pochmann @ericvsmith