Element-wise multiplication & division of pl.Decimal silently fail and produce inconsistent results

[scur-iolus]

Checks

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of Polars (1.0.0-beta.1 to date).

Reproducible example

The following test passes with float but fails with Decimal as a datatype:

from decimal import Decimal as Dec

import polars as pl
import pytest


@pytest.mark.parametrize("datatype", [float, Dec])
def test_element_wise_multiplication_n_division(datatype) -> None:
    df = pl.DataFrame(
        [
            {
                "a": datatype(f"1.{'0' * 20}"),
                "b": datatype(f"1.{'0' * 20}"),
            }
        ]
    )
    df = df.with_columns(c=pl.col("a") * pl.col("b"))
    df = df.with_columns(d=pl.col("a") / pl.col("b"))
    # next line fails: I got a Decimal('0.0131811...') probably due to an overflow?
    assert df[0, "c"] == datatype("1")
    # next line fails but differently: the value is null
    assert df[0, "d"] == datatype("1")

Issue description

Some very basic element-wise operations between Decimal values fail, probably because of an overflow, without explicit warning nor error message.

This is a very nasty bug: in an element-wise multiplication, the output values may be plausible, but turn out to be completely wrong, making it really hard to debug. Here, I've tried to reduce the problem to its most transparent form (multiplying 1 by 1 must equal 1, not 0.013...), but I'm sure other people could tear their hair out over it.

I know polars only partially supports Decimal, I'm aware that it's a "work-in-progress feature", I tend to believe that this issue should be fixed before polars can claim to support this datatype. I marked this as a python bug since that is how I stumbled over it but I wouldn't be surprised if the same bug exists in Rust.

This could be related to #15037 in the sense that in both cases the root of the issue is likely a silent overflow.

Expected behavior

An explicit exception should be raised or a warning should be displayed.

Installed versions

--------Version info---------
Polars:               1.0.0-beta.1
Index type:           UInt32
Platform:             Linux-5.15.0-107-generic-x86_64-with-glibc2.31
Python:               3.11.9 (main, Apr  6 2024, 17:59:24) [GCC 9.4.0]

[ritchie46]

Set to p-medium as the decimal type is still unstable.

Note that if its is overflow, this is not something we check during execution. We follow numpy in that regard. Numeric arithmetic can overflow:

pl.Series([233], dtype=pl.UInt8) + 30
shape: (1,)
Series: '' [u8]
[
	7
]