Feature to format floating point numbers - #896

The length needs to be adjusted if trailing zeros are removed or else a 0-termination is copied into the output buffer.

    else if (item->type & cJSON_NumberIsFormatted)
    {
        precision = _get_precision_from_item(item, sizeof(number_buffer));
        if (item->type & cJSON_NumberFormatStyleFixedPoint)
        {
            sprintf((char*)number_buffer, "%.*f", (int)precision, d);
            remove_trailing_zeros((char*)number_buffer);
            length = strlen((char*)number_buffer); /*     <================    */
        }
        else
        {
            length = sprintf((char*)number_buffer, "%.*g", (int)precision, d);
        }
    }

Thank you for the review. I've updated it as suggested

Another project recently showed interest in this functionality, so I wanted to check whether this is something that could be considered for merge, or whether there are architectural concerns/preferences I should address first.

I've also been using this feature in my fork for quite some time now without issues, so I'm mainly trying to understand whether it makes sense to keep waiting for upstream inclusion, or whether the preference would be to close/reject the PR.