problem understanding a panagram code

[nor]

#include <iostream>
#include <vector>
#include <cctype>
using namespace std;

const int ALPHA_SIZE = 26;

int main()
{
string line;

while (getline(cin, line))
{
vector<int> letter_count(ALPHA_SIZE);
// int letter_count[ALPHA_SIZE] = { 0 }; // alternativ

// Räkna förekomsten av varje bokstav på raden.
for (unsigned int i = 0; i < line.length(); ++i)
{
char c = line[i];

if (isalpha(c))
{
++letter_count[toupper(c) - 'A']; <------- dont understand this this function and why -A?
}
}

// Kontrollera om raden uppfyller kraven för panagram.
bool is_anagram = true;
bool is_perfect = true;
int n_letters = 0;

for (int i = 0; i < ALPHA_SIZE; ++i)
{
n_letters += letter_count[i];

if (letter_count[i] > 1)
{
is_perfect = false;
}

if (letter_count[i] == 0)
{
is_anagram = false;
break; // out of for loop
}
}

if (is_anagram)
{
cout << n_letters << ": " << line;
if (is_perfect) cout << " [PERFEKT]";
cout << '\n';
}

}

return 0;
}

[Athar]

The letters A-Z are next to each other in any character set. See the ASCII table, for instance:
http://www.asciitable.com/

So by subtracting 'A', you get the alphabet index for an uppercase character, i.e. 0 for A, 1 for B and 25 for Z.
This is then used as an array index to increase the count for that particular letter.

[nor]

Thank you very much for the information!