pub fn split_sections_by_boundary_owned( sections: Vec<Section>, ) -> Vec<Vec<Section>>
Split owned sections into documents based on ENDPOINT boundaries.
Each ENDPOINT with meaningful content before it starts a new document.