Strings Starting with '0' and Decimal Numbers Incorrectly Converted During XML to JSON Transformation

[Justsvetoslavov]

This issue is similar to the previously resolved issue #826, which addressed the regression where strings starting with a '0' were incorrectly converted to integers. (fixed, as of 20240303, as noted in #868)

However, this new issue concerns decimal numbers that start with a '0', which are still being converted incorrectly in the latest version of the library.

XML

<datatypes>
    <data1>012345.6789</data1>
    <data2>0123456789</data2>
</datatypes>

we use:

JSONObject o = XML.toJSONObject(xml); // keep strings false

Output (on v20240303):

12345.6789 // leading zero omitted
0123456789

While issue #826 resolved the problem of handling integers starting with '0', this particular edge case with decimal numbers was not addressed. As a result, the current behavior strips the leading zero from the decimal number, converting 012345.6789 into 12345.6789, which is not the expected behavior in our case.

It seems there is a mismatch in behavior between how the library handles integer strings and decimal string numbers starting with '0'. Is it possible to handle decimal numbers in a way that preserves the leading zeros as part of the string, without requiring keepStrings=true, which would otherwise undesirably affect the conversion of other data types like booleans and integers?

Thank you for your attention to this issue.

[Justsvetoslavov]

As this has always been working like this and will be a breaking change, I believe the best option here is to handle it on our side.

[stleary]

@Justsvetoslavov This project tries to offer more flexibility for XML/JSON transformations. Feel free to reopen this issue if you would like this specific behavior addressed.