Describe the bug

itertools._tee_dataobject(it, None, None) creates an object on GraalPy, while CPython raises TypeError because values must be a list. Calls with fewer than three arguments are also accepted, where CPython requires exactly three.

Operating system

macOS

CPU architecture

ARM64

GraalPy version

master a3009f9 (Python 3.13.14); also reproduced on 25.1.3 (Python 3.12.8)

JDK version

No response

Context configuration

No response

Steps to reproduce

import itertools
print(type(itertools._tee_dataobject(iter([]), None, None)))
print(type(itertools._tee_dataobject(iter([]), [])))

GraalPy master a3009f9:

<class 'itertools._tee_dataobject'>
<class 'itertools._tee_dataobject'>

Expected behavior

CPython 3.13.14 (same on 3.12.13):

TypeError: teedataobject() argument 2 must be list, not None
TypeError: teedataobject expected 3 arguments, got 2

Stack trace

Additional context

_tee_dataobject(1) and _tee_dataobject() also create an object. _tee_dataobject(iter([]), None, 5) matches no specialization and fails with Unexpected values provided for TeeDataObjectBuiltinsFactory.InitNodeFactory.InitNodeGen instead of a TypeError.

Root Cause

TeeDataObjectBuiltins.InitNode requires only $self (minNumOfPositionalArgs = 1) and defaults values and nxt to None (TeeDataObjectBuiltins.java#L121-L124). Two isNone(values) specializations then treat None as an empty buffer (#L135-L153), and the TypeError specialization excludes None (#L184-L189). CPython checks for exactly three positional arguments, then rejects any non-list values (itertoolsmodule.c.h#L244-L251).

Fix Suggestion

Set minNumOfPositionalArgs = 4, remove the two isNone(values) specializations and their helper, and guard the TypeError specialization with !isList(values) alone.