rows of non key column are reversed after left join of two table

### Describe the bug, including details regarding any error messages, version, and platform.

Here table left and right joins by Idx column, "Series" column in right table has its original sequence **always reversed** in output table. How to preserve original ordering ? This happens in both c++ and python code. Example shown below

def test_hash_join_order():
# Test with different Series orders in right table
left = pa.table({'City': ["NEWYORK"], 'Number': [44], 'Idx':[1]})

# Test case 1: Original order [0,1,2,3]
right1 = pa.table({'Idx': [1, 1, 1, 1], 'Series': [0,1,2,3]})

# Test case 2: Reverse order [3,2,1,0]
right2 = pa.table({'Idx': [1, 1, 1, 1], 'Series': [3,2,1,0]})

# Test case 3: Random order [2,0,3,1]
right3 = pa.table({'Idx': [1, 1, 1, 1], 'Series': [2,0,3,1]})

def do_join(right):
left_source = Declaration("table_source", options=TableSourceNodeOptions(left))
right_source = Declaration("table_source", options=TableSourceNodeOptions(right))
join_opts = HashJoinNodeOptions("left outer", left_keys="Idx", right_keys="Idx")
joined = Declaration("hashjoin", options=join_opts, inputs=[left_source, right_source])
return joined.to_table()

print("Test 1 (Original order):", do_join(right1).column('Series').to_pylist())
print("Test 2 (Reverse order):", do_join(right2).column('Series').to_pylist())
print("Test 3 (Random order):", do_join(right3).column('Series').to_pylist())

test_hash_join_order()

Output:
Test 1 (Original order): [3, 2, 1, 0]
Test 2 (Reverse order): [0, 1, 2, 3]
Test 3 (Random order): [1, 3, 0, 2]
### Component(s)

C++