How to get certain values from user input and ignore others - c++

Question

Is there any way of getting only the numbers from this string input:
h8 f7
If tried fscanf. But, probably i'm using it wrong.
int positionOfTheKnight = 0, finalDestination = 0;
fscanf("%*s%d %*s%d", &positionOfTheKnight, &finalDestination);
cout << positionOfTheKnight << " " << finalDestination;
It shows this following error:
cpp|11|error: cannot convert 'const char*' to 'FILE* {aka _iobuf*}' for argument '1' to 'int fscanf(FILE*, const char*, ...)'|

Answer 1

I made a small sample to demonstrate multiple ways to read the example of OP:
#include <cstdio>
#include <iostream>
#include <string>
int rcToInt(char c, char r)
{
if (c < 'a' || c > 'h' || r < '1' || r > '8') {
std::cerr << "Wrong input!\n";
return -1;
}
return (c - 'a') * 8 + (r - '1'); // convert characters to index
}
int main()
{
{ // using scanf
std::cout << "Using scanf():\n";
char r0, c0, r1, c1;
if (scanf(" %c%c %c%c ", &c0, &r0, &c1, &r1) < 4) {
std::cerr << "Reading move with scanf() failed!\n";
} else {
const int pos0 = rcToInt(c0, r0), pos1 = rcToInt(c1, r1);
if (pos0 >= 0 && pos1 >= 0) {
std::cout << "Got " << c0 << r0 << " -> " << c1 << r1 << '\n';
}
}
}
{ // Yet another way to use scanf():
std::cout << "Using scanf():\n";
char c0[2], r0[2], c1[2], r1[2];
if (scanf("%1[a-h]%1[1-8] %1[a-h]%1[1-8] ", c0, r0, c1, r1) < 4) {
std::cerr << "Reading move with scanf() failed!\n";
} else {
const int pos0 = rcToInt(*c0, *r0), pos1 = rcToInt(*c1, *r1);
if (pos0 >= 0 && pos1 >= 0) {
std::cout << "Got " << c0 << r0 << " -> " << c1 << r1 << '\n';
}
}
}
{ // using std::cin
std::cout << "Using operator>>():\n";
char r0, c0, r1, c1;
if (!(std::cin >> c0 >> r0 >> c1 >> r1)) {
std::cerr << "Reading move with operator>>() failed!\n";
} else {
const int pos0 = rcToInt(c0, r0), pos1 = rcToInt(c1, r1);
if (pos0 >= 0 && pos1 >= 0) {
std::cout << "Got " << c0 << r0 << " -> " << c1 << r1 << '\n';
}
}
}
// There is still the [ENTER] in input queue which must be consumed:
std::cin.ignore(std::numeric_limits<std::streamsize>::max(), '\n');
// ...before proceeding
{ // using std::cin
std::cout << "Using std::getline():\n";
std::string line;
if (!std::getline(std::cin, line) || line.size() < 5) {
std::cerr << "Reading move with std::getline() failed!\n";
} else {
const char c0 = line[0], r0 = line[1], c1 = line[3], r1 = line[4];
const int pos0 = rcToInt(c0, r0), pos1 = rcToInt(c1, r1);
if (pos0 >= 0 && pos1 >= 0) {
std::cout << "Got " << c0 << r0 << " -> " << c1 << r1 << '\n';
}
}
}
// done
return 0;
}
Output:
Using scanf():
h8 f7↵
Got h8 -> f7
Using scanf():
f7 h8↵
Got f7 -> h8
Using operator>>():
h8 f7↵
Got h8 -> f7
Using std::getline():
f7 h8↵
Got f7 -> h8
Live Demo on coliru
All four alternatives follow the basic idea to read the input as characters and convert them to a corresponding index in the range of [0, 63] afterwards.
While char might represent characters, it's as well an integral value – specifically the smallest available integral type. In (C and) C++, this is no difference. A character '1' is stored as integer value 33 (assuming ASCII code)). '1' as well as 33 represent the same value. Hence, it is fine (and even recommended) to use character constants for arithmetic computations if appropriate (like in rcToInt() in the above example).
if (scanf(" %c%c %c%c ", &c0, &r0, &c1, &r1) < 4) {
Formatter starts with a space () to consume optionally pending white space.
%c stores one character. Hence, a char* argument (pointing to sufficient storage) has to be provided.
Formatter ends with a space () to consume the terminating ENTER as well.
The return value of scanf() is checked to grant that all 4 arguments will be assigned.
if (scanf(" %1[a-h]%1[1-8] %1[a-h]%1[1-8] ", c0, r0, c1, r1) < 4) {
Formatter starts and ends with a space () to consume surrounding white space (like in the previous example).
%1[a-h] reads a string of length 1 which may consist of characters a, b, ..., h only. There has to be provided storage for an extra byte as that formatter will always store an additional 0-terminator (\0). Hence, the declations char c0[2], r0[2], c1[2], r1[2]; in this case.
%1[1-8] similar like above for the characters 1, 2, ..., 8. Please note, that digits are read as characters.
if (!(std::cin >> c0 >> r0 >> c1 >> r1)) {
Read four char variables using stream input operators.
The iostream header provides a variety of overloaded operator>>s (e.g. std::istream& operator>>(std::istream&, char&)) for this.
The whitespace (occurring before reading c1) is skipped by default setting of std::cin.
if (!std::getline(std::cin, line) || line.size() < 5) {
Read a whole line into a std::string (which is grown appropriately).
Analyze characters in that std::string.
A combination of 3. and 4. would be to read a line with std::getline(), wrap the read line in a std::istringstream to parse it with input stream operators.
Sorry, if I swapped row and column in the above code. I'm no chess expert and not aware of the usual convertions.