py-tablib: updated to 3.6.0

3.6.0 (2024-03-23)

Improvements

- It's now possible to access a dataset row using its index without slicing
- The dataset `transpose()` method can be called on datasets without headers.
- The html format now supports importing from HTML content
- The ODS format now supports importing from .ods files
still a bit experimental.
- When adding rows to a dataset with dynamic columns, it's now possible to
provide only static values, and dynamic column values will be automatically
calculated and added to the row

Changes

- The html export format does not depend on MarkupPy any longer, therefore the
tablib[html] install target was removed also.

Bugfixes

- Fix crash when loading a databook from an XLS file
- `None` Python values are now converted to the empty string by the ODS formatter.
- When applying formatters, the internal data is no longer mutated
- Columns can be inserted even when a dataset has headers but no values

(adam)