Open swartchris8 opened 3 years ago
Hmmm... yeah, that's a tough one API-wise. Any thoughts on how the API should look?
The way that would be most natural to me is if annotations could be nested
lowest_level_annotation_tuple = ("is", "verb", "#8ef")
overlapping_annotation = (["This", lowest_level_annotation_tuple, "annotated"], "example", "#afa")
So annotations could be constructed from a list of space separated tokens and annotations instead of just constructing them from strings.
I know it complicates things quite a bit but would be nice for my use case!
I would love it if overlapping annotations could be handled though this would require substantial changes to annotation