feat(python/sedonadb): Automatically give nameless dataframes an alias on create

Now that we have joins in Python, it's very easy to end up with an error about ambiguous references because almost everything has the default alias `?table?`. In most cases we can do a pretty good job guessing the alias (e.g., the filename when returning from a read function).

The issue names no files, tests, or entry points. Start by locating Python dataframe creation and existing alias handling, then clarify which nameless-dataframe cases should receive inferred aliases and add tests showing that joins no longer produce ambiguous-reference errors.

Written by the indexing model from the issue text.