JSON does not specify a mechanism to determine the scope of numbers, so these have to be derived by looking at the actual values when importing the JSON. The importer will use either the Decimal or Double data type when encountering a number, depending on the presence or absence of the e (exponent) value. Decimal is assumed rather than integer because the size of the number could be larger than an Integer (or even 64-bit Long). In addition, the number could actually have a fractional component in the actual data. Decimal will thus always work in these situations. Double is assumed for floating point numbers (having the exponent value).