Fix fragile doctest sorting in multi_polynomial_ideal

[Hapsa21]

Description

This PR fixes a fragile doctest failure reported in #41193.

The Issue:
The failing test in src/sage/rings/polynomial/multi_polynomial_ideal.py used sorted(..., key=str) on a list of dictionaries. In Python, the string representation of a dictionary depends on the insertion order of its keys. Since the upstream library msolve likely changed its internal insertion order in newer versions, key=str produced a different sort order than expected, causing the test to fail despite the mathematical results being correct.

The Fix:
I updated the test command to replace the unstable sort key with a robust one. This sorts the dictionary items before stringifying, ensuring the sorting relies on the dictionary contents rather than their internal memory layout.

# Old (fragile)
key=str

# New (robust)
key=lambda d: str(sorted(d.items()))

Verification
I verified that this change makes the sorting immune to dictionary insertion order.

Proof of Fix: Running the following script demonstrates why key=str failed and why the new key succeeds:

dict_A = {'x': 100, 'y': 500}
dict_B = {'y': 500, 'x': 100}

#  ---Old Behavior (key=str)--- 
# str(dict_A) is "{'x': 100, 'y': 500}"
# str(dict_B) is "{'y': 500, 'x': 100}"
print(f"Old method equal? {str(dict_A) == str(dict_B)}")
# Result: False (This caused the test failure)

# --- New Behavior (The Fix) ---
my_key = lambda d: str(sorted(d.items()))
print(f"New method equal? {my_key(dict_A) == my_key(dict_B)}")
# Result: True (The sort is now stable)

This ensures the doctest passes reliably regardless of the upstream library's behavior.

Issue

Fixes #41193 Relates to #41192

📝 Checklist

• The title is concise and informative.
• The description explains in detail what this PR is about.
• I have linked a relevant issue or discussion.
• I have created tests covering the changes.
• I have updated the documentation and checked the documentation preview.

⌛ Dependencies

None.

[cxzhong]

Thank you. LGTM