Filequery

- escape data in Rich tables so it treats the data as literal instead of trying to render it if it contains square brackets
- if a file name does not conform to DuckDB's [rules on unquoted identifiers](https://duckdb.org/docs/sql/keywords_and_identifiers.html), wrap it in double quotes

- Since DuckDB 0.10.0, `read_csv` is preferred to `read_csv_auto`, so filequery uses that for reading in CSVs now
- Improved CSV export from TUI. Filequery now uses a DuckDB `copy` statement behind the scenes to create the CSV instead of building the CSV from the contents of the result table in the UI

- menu option to load a SQL file
- brings up a file browser modal where a user can select a file to load into the current editor tab
- `FileDb` support for exporting query result to a JSON file as a list of JSON documents
- added `dict_records` property to QueryResult to allow users to retrieve the records as a list of dictionaries

- Table list on the left pane is now a interactive tree that can be expanded to see columns and data types
- Added menu screen dialog
- The "save SQL" and "save result" options are now buttons in the menu instead of key bindings, this helps reduce clutter in the UI
- Menu has a button for exiting the program

What's Changed
- added tabs in the TUI to allow for multiple SQL editors
- added keyboard controls in the TUI for switching between panes
- improved styles to make it easier to tell the currently active pane in the TUI

What's Changed
- can now run filequery TUI without specifying files or a file directory, enabling users to to write their own SQL scripts for loading files with any options they want to specify
- use sample_size=-1 when reading csv, json and ndjson files so all records are sampled when DuckDB is determining the table structure
- in the TUI, the table list is now reactive, so creating/dropping tables causes the list to update
- dropping support for Python 3.7