Well, of UTF-16 itself, not specific to Javascript. What they're exposing are the 16-bit Code Units of UTF-16, which happen to be (reserved) Unicode Code Points

Well, of UTF-16 itself, not specific to Javascript. What they're exposing are the 16-bit Code Units of UTF-16, which happen to be (reserved) Unicode Code Points, but are not Unicode Scalar Values. I guess I don't know what this software is for? But the references to the normalisation forms makes me assume it's about Unicode and so it ought to work in the actual Code Points, not these surrogates.