Hi,
First I want to appreciate your effort of making this library. It fits our use case perfectly and we're considering using this library on our UI.
When trying it, we found something that is not quite straightforward in the real world paragraph comparison.
For example, in following cases, we are expecting word-by-word diffing instead of character-by-character:
I changed "some" to "same"
image

I changed from "I love accessibility" to "I have accountability"
image