[Python][Parquet] Allow to select columns of a list field without requiring the list component names

Subtask for ARROW-14196.

Currently, if you have a list column, where the list elements itself are nested items (eg a list of structs), selecting a subset of that list column requires something like `columns=["columnA.list.item.subfield"]`. While this "list.item" is superfluous, since a list always contains a single child. So ideally we allow to specify this as `columns=["columnA.subfield"]`.

This also avoids relying on the exact name of the list item (item vs element), for which the default differs between Parquet and Arrow.

**Reporter**: [Joris Van den Bossche](https://issues.apache.org/jira/browse/ARROW-14286) / @jorisvandenbossche
#### Related issues:
- [[C++][Parquet] Default to compliant nested types in Parquet writer](https://github.com/apache/arrow/issues/29781) (is related to)
#### PRs and other links:
- [GitHub Pull Request #11385](https://github.com/apache/arrow/pull/11385)

_{**Note**: *This issue was originally created as [ARROW-14286](https://issues.apache.org/jira/browse/ARROW-14286). Please see the [migration documentation](https://github.com/apache/arrow/issues/14542) for further details.*}

Start by reviewing pull request #11385 and the Python Parquet column-selection handling. Verify the current path form for nested list-of-struct fields, then confirm that selecting `columnA.subfield` works without the list component name and that the existing form remains supported.

Written by the indexing model from the issue text.