Our current text iterator is not aware of multi-code point graphemes. Instead of simply incrementing an iterator one code point at a time, use our Unicode grapheme segmenter to break text into fragments.
