Hope to keep the float type when the fractional part ends with '. 0' by using UnmarshalIntOrFloat

[Kish29]

when I use UnmarshalIntOrFloat to decode float number especially when the fractional part ends with '. 0', here is the test:

package main

import (
    "fmt"
    "strings"

    "gopkg.in/square/go-jose.v2/json"
)

func main() {
    dataStr := `
    {
        "float_zero": 0.0,
        "float_a": 1.0,
        "float_b": 0.123,
        "float_c": 21.00000,
        "integer_zero": 0
    }
    `
    dec :=json.NewDecoder(strings.NewReader(dataStr))
    dec.SetNumberType(json.UnmarshalIntOrFloat)
    var kv map[string]interface{}
    err := dec.Decode(&kv)
    if err != nil {
        panic(err)
    }
    for k, v := range kv {
        fmt.Printf("k=>%v(%T), v=>%v(%T)\n", k, k, v, v)
    }
}

and output is:

k=>integer_zero(string), v=>0(int64)
k=>float_zero(string), v=>0(int64)
k=>float_a(string), v=>1(int64)
k=>float_b(string), v=>0.123(float64)
k=>float_c(string), v=>21(int64)

So I hope that it could offer an option to decide whether keep these special float number still be float type like float_zero,float_a,float_c, thanks! :)

[csstaub]

This is a question for the Go standard library, the json package in go-jose is merely a copy of encoding/json from Go with updated semantics to make it more suitable for JOSE objects. It's not meant for use outside of this library.