Fixes:esp8266#8938Also increase accuracy of mapping and fixing map issues with large values.Still only using integer operations.Test function to see absolute error compared to floating point map function using `double` precision floating point values.```c++# include <iostream># include <string>long test_map(long x, long in_min, long in_max, long out_min, long out_max) {    const double out_length = out_max - out_min;    const double in_length  = in_max - in_min;    if (in_length == 0) { return in_min; }    if (out_length == 0) { return out_min; }    const double delta = x - in_min;    const double value = (delta * out_length + (in_length / 2)) / in_length + out_min;//    return std::round(value);    return value;}long check(long x, long in_min, long in_max, long out_min, long out_max) {    const long floatvalue = test_map(x, in_min, in_max, out_min, out_max);    const long map_value = map(x, in_min, in_max, out_min, out_max);    return floatvalue - map_value;}int main(){    long input = 0;    for (size_t i = 0; i < 15; ++i) {        input = (i / 3) * 2500 - 1 + i % 3;        const long values[] = {            check(input, 0,        1000,   255,      0),            check(input, 0,        1000,   0,        255),            check(input, 10000,    0,      100,      10100),            check(input, 10000,    0,      10100,    100),            check(input, 10000,    0,      0,        1024),            check(input, 10000,    0,      1024,     0),            check(input, 1000,     0,      0,        10240),            check(input, 1000,     0,      10240,    0),            check(input, 0,        1000,   0,        10240),            check(input, 0,        1000,   10240,    0),            check(input, 0,        10000,  10240,    0),            check(input, 10234567, -12345, 10234567, 100),            check(input, 10000,    0,      10234567, 0),            check(input, 50,       1,      10234567, 0)        };        std::cout << input << ":";        constexpr size_t nrvalues = sizeof(values) / sizeof(values[0]);        for (size_t i = 0; i < nrvalues; ++i) {            std::cout << "\t" << values[i];        }        std::cout << "\n";    }}```Output:```-1:0-1000-102200-1220:00000000000-1011:-1000-1010011-1002499:1-10000111100002500:110000100100102501:000000200210014999:1-10000111100005000:100000100100115001:000000200210107499:1-10000111100117500:110000100100007501:000000200210109999:1-1000-11111001010000:1000001001000010001:0000-1020022001```

Not really a need to make a recursive call for simply swapping values.