BUG: Fix to_dict(orient='index') to apply 'into' to nested mappings

[C1-BA-B1-F3]

What does this PR do?

Fixes #65778

This PR fixes a bug where DataFrame.to_dict(orient='index', into=MyDict) only applied the into parameter to the outer mapping, leaving nested row mappings as plain dict.

Example

from collections import OrderedDict
import pandas as pd

df = pd.DataFrame({"A": [1, 2], "B": ["x", "y"]})
result = df.to_dict(orient="index", into=OrderedDict)

# Before fix: result[0] is dict
# After fix: result[0] is OrderedDict
assert isinstance(result[0], OrderedDict)  # Now passes

Changes

• Modified all three return paths in the orient='index' branch to use into_c for nested mappings instead of dict()
• Added tests for:
  • OrderedDict with basic DataFrame
  • defaultdict with basic DataFrame
  • Mixed dtypes (object and non-object columns)
  • All object dtype columns

Testing

Added comprehensive tests covering:

1. test_to_dict_index_into_nested - Tests OrderedDict and defaultdict
2. test_to_dict_index_into_nested_mixed_dtypes - Tests with mixed column types
3. test_to_dict_index_into_nested_all_object - Tests with all object columns

All new tests pass.