Closed lizadaly closed 1 year ago
Specifically, there are lots of business rules embedded in how elisions (or any annotation) is treated when intermingled with case text—the computed offsets are fine with user-generated content, but not case text, which is more complex and has transformations in a few places.
I'll have to reverse-engineer how the offsets were computed when it comes to how case elements are presented so the character-based offsets match.
Tanaka, 3.2, in the platform versus the reader view Zittrain, 3.1.1 in the platform versus the reader view