Cannot properly eliminate string one from another character by character - c++

Question

Here the function (sub) takes two string as input, traversing two string I try to find out if there is any matches in string1 compared to string2. If any that character of string1 is replaced by NULL character. Now this works properly for non repeated character. But if string1 has more than one character that matches once it all replaced by NULL character where i needed only one replacement. For example if string1 and string2 are 122 and 2, after elimination i need 1 2 where i gets now a single 1.
void sub (string str1, string str2){
int i,j,k;
for(i=0; i<=str2.size() ; i++){
for(j=0; j<=str1.size() ; j++ ){
if( str2[i] == str1[j] )
str1[j] = NULL;
}
}
cout<<str1;
expected result is 1 2 instead of 1, if str1=122 and str2=2

Answer 1

You are making things more difficult on yourself than they need to be. The string library provides two functions that can do exactly what you need in a single call.
The member function std::basic_string::find_first_of will locate the first occurrence of a character from string2 in string1 returning the position where it is found.
The std::basic_string::erase function can remove all characters from string1 beginning at that position.
Your sub function will then reduce to:
void sub (std::string& s1, const std::string& s2)
{
s1.erase (s1.find_first_of (s2));
}
A short example using your given strings would be:
#include <iostream>
#include <string>
void sub (std::string& s1, const std::string& s2)
{
s1.erase (s1.find_first_of (s2));
}
int main (void) {
std::string s1 ("122"), s2 ("2");
sub (s1, s2);
std::cout << "s1: " << s1 << "\ns2: " << s2 << '\n';
}
Example Use/Output
$ ./bin/sub1at2
s1: 1
s2: 2
Look things over and let me know if you have further questions.

Answer 2

You can't remove a character from a string by setting it to NULL. The length of the string will remain the same. But one way to simulate the removal of the duplicates is to return a new string that matches the return conditions.
First iterate over the second string and use a hash table to map each character in s2 to true. Then iterate over s1 and add the current character to a new string only if the character in the hash table maps to false. Remapping the character to false after this condition ensures that all but one of the number of characters is written to the result string.
string remove_first_duplicates(string s1, string s2) {
unordered_map<char, bool> m;
string result;
for (char i : s2) m[i] = true;
for (char i : s1) {
if (!m[i]) result += i;
m[i] = false;
}
return result;
}

Answer 3

NULL is not a character-constant, even if \0 is the null character. It's a macro for a null pointer constant, and for historical reasons is often defined as 0, though it might be nullptr or any other null pointer constant.
Zeroing out characters does not stop them being part of the string. for that, you must move the remaining ones and adjust the length.
If you only want to do it once, on the first match, leve the function with return afterwards.
Consider separating it into two functions: One for finding a match, and one calling that and using the result for removing the first match.

Answer 4

As far as I understood your question, you want to remove one char from str1 corresponding to a match in str2.
void sub(string str1, string str2)
{
int i = 0, j = 0;
while (j < str2.size())
{
if (str1[i] == str2[j])
{
str1[i] = NULL; // could use str1.erase(i,1)
i = 0;
j += 1;
continue;
}
else
i += 1;
if (i == str1.size() - 1)
{
i = 0;
j += 1;
}
}
cout<<str1<<endl;
}
This will yield the output you desire. But this will produce NULL char in str1, better option is to use erase functionality from std::string.