The code is certainly very short, but what about \u escape sequences in strings, parsing different representations of numbers, etc.? Since those things are part

The code is certainly very short, but what about \u escape sequences in strings, parsing different representations of numbers, etc.? Since those things are part of the JSON standard, you're not a JSON parser if you just leave them to the application to handle. Since this skimps out on half of the work, it won't even be able to tell you with certainty what's valid JSON and what isn't. (disclaimer: I also wrote a popular ANSI C JSON parser)