I second this. My biggest problem with it is that I have no idea which operators accept byte slices, which accept strings, which accept both and which accept both, but treat both as byte streams (which might or might not matter). For many operators the documentation simply does not say and it's hard to say from the source as the types are often inferred and so not mentioned in code either.
It would be particularly nice to modify the
examples to contain non-ascii characters in all places that work on strings and chars to demonstrate they are handled properly (from brief look at the code I suspect it would find some bugs, too)—or clear information that it works on bytes, not utf-8 sequences.