Right now, the output of SDU is aggregated into a JSON array.
It would be great if SDU could return the parsed structure of the document. This way we can know for sure if this word in the document is a 'Table of Content' type, or 'Subtitle' type.
Why is it useful?
|Who would benefit from this IDEA?||If I had this, I would be able to identify sections of a document and treat it accordingly. For example, if I combined this with TTS service, I would read different sections of the document in different styles.|
How should it work?