Not to mention that stuff like Excel loves to export CSV files in "locale specific ways".

Sometimes commas to delimiter, sometimes semicolons, floating point values might have dots or commas to separate fraction digits.

Not to mention text encodings, Ascii, western european character sets, or maybe utf-8 or whatever...

It's a bloody mess.

That’s why we just email the sheets around like it’s 1999 :)

You need more than a standard; that standard has to be complete and unambiguous. What you're looking for is https://github.com/nst/JSONTestSuite

EDIT: The readme's results are from 2016, but there's more recent results (last updated 5 years ago). Of the 54 parsers /versions tested, 7 gave always the expected result per the spec (disregarding cases where the spec does not define a result).

It falls down under very large integers -- think large valid uint64_t values.

JSON doesn't fail for very large values because they are sent over the wire as strings. Only parsers may fail if they or their backing language doesn't account for BigInts or floats larger than f64, but these problems exist when parsing any string to a number.

And indeed applies to CSV as well: it's just strings at the end of the day, its up to the parser to make sense of it into the data types one wants. There is nothing inherently stopping you from parsing a JSON string into a uint64: I've done so plenty!