In some cases it is useful to know what the original word form of a word or substring was after a string is normalized. This means that the changes that are made during normalization need to be tracked.
This PR does exactly that and allows users of the API to extract the original substring for indices in the normalized string.
In some cases it is useful to know what the original word form of a word or substring was after a string is normalized. This means that the changes that are made during normalization need to be tracked. This PR does exactly that and allows users of the API to extract the original substring for indices in the normalized string.