C++ Primer 5th Edition Exercise 1.23 关于我找到的解决方案的问题

Question

The exercise: Write a program that reads several transactions and counts how many transactions occur for each ISBN.

The answer I found on Github:

#include <iostream>
#include "include/Sales_item.h"

int main()
{
    Sales_item currItem, valItem;
    if (std::cin >> currItem)
    {
        int cnt = 1;
        while (std::cin >> valItem)
        {
            if (valItem.isbn() == currItem.isbn())
            {
                ++cnt;
            }
            else
            {
                std::cout << currItem << " occurs " << cnt << " times " << std::endl;
                currItem = valItem;
                cnt = 1;
            }
        }
        std::cout << currItem << " occurs "<< cnt << " times " << std::endl;
    }
    return 0;
}

For example if the input ISBN is like this:

A A B A B C

The output would be:

A occurs 2 times
B occurs 1 times
A occurs 1 times
B occurs 1 times
C occurs 1 times

Which isn't quiet right because they're not properly grouped. The excercise didn't specify that records for each ISBN should be grouped together. (Though it's specified in the next exercise.)

I haven't found a right solution online yet. I've been thinking about this excercise for a while and still don't have much clue on how to solve it.

I've only learned about while, for, if so far. This seemingly easy exercise is kind of a challenge for me. Hope you guys can give me, a beginner some clue about it. Thanks in advance!

Answer 1

use a std::map<char, int> chcounts; where the key is the character.

You can then do

  chcounts[ch]++;

this takes advantage of the fact that if you reference a map element using [] and the entry does not exist it will get created and default initialized

or if you have strings do

 std::map<std::string, int> counts;

 ....
 counts[isbn]++;