[penguindev]

"So for example, a string containing only the G clef character (U+1D11E) may be represented as "\uD834\uDD1E""

eww, UTF-16 surrogate pairs. Die, die, die, you scourge. (Yes, I've been working with the win32 api lately, which was designed for UCS2/UTF-16).

when has 16 bit anything been a good idea, really. TCP port ranges, looking at you too.

[quink]

Char can be anything not " or \.

You don't have to escape unicode. If you want to just use Unicode and encode as UTF-8 then go nuts.

This is valid JSON:

{ "äöü": "大家好", "clef": "𝄞" }

[penguindev]

I realize that. I'm just sad to see UTF-16 vestiges living on, in current specs. Who can fathom how much misery and billions of dollars in wasted productivity (byte order marks - fun!) that wrought upon the IT landscape....

Python has an upper case \UXXXXXXXX literal that seems a lot cleaner.