How to remove a specific keyword from a given user input? - c++

Question

I'm working on a program that takes user input two times. The user will input a sentence, and then enter a key phrase to remove each found match throughout the sentence, and return what's left. The code I have below for the 'logic' is...
char delim[AFFIX];
strcpy(delim, userInput); //storing the phrase user wants to remove
char *token = strtok(sentence, delim);
while (token)
{
cout << token << endl;
token = strtok(NULL, delim);
}
The issue I'm having is that strtok removes every single instance of a character found. For example, if the user wishes to remove all instances of 'pre' then the word "preformed" would turn into "fo m d" instead of formed. Is there a way to restrict strtok from removing EVERY instance of a character found, and only remove the series of characters?
I understand that working with strings or vectors would make my life much easier, but I want to work with arrays of chars. I'm really sorry if this isn't clear enough, I'm very new to c++. Any advice on how to approach this problem would be greatly appreciated. Thank you for your time.

Answer 1

A easy and more or less dirty solution would be to just add spaces before and after your search word. To ensure it's actually a full word and not part of a word.
Eg:
Input: "pre"
Check: " pre "
To ensure you also cut the first word you would need to check the first part of the string and add a space before it if the first word is also the inputted word.
Example:
#include <string>
#include <iostream>
void removeSubstrs(std::string& s, std::string& p) {
std::string::size_type n = p.length();
for (std::string::size_type i = s.find(p);
i != std::string::npos;
i = s.find(p))
s.erase(i, n);
}
void check(std::string& s, std::string& p) {
int slen = s.length();
int plen = p.length();
std::string firstpart = s.substr (0,plen);
std::string lastpart = s.substr (slen-plen,slen);
if (firstpart == p) {
s = " " + s;
}
if (lastpart == p) {
s += " ";
}
}
int main ()
{
std::string str = "pre test inppre test pre";
std::string search = "pre";
std::string pattern = " " + search + " ";
check(str, search);
removeSubstrs(str, pattern);
std::cout << str << std::endl;
return 0;
}

Answer 2

strtok isn't the right choice here, you can think of strtok's 2nd argument as a list of delimiters (each char in that list is a delimiter) and it breaks the string into tokens using those delimiters.
A better solution would be something along the lines of strstr and strcat.
maybe something like this (using MSVC++)
void removeSubstr(char *string, const char *sub, int n) {
char *match;
int len = strlen(sub);
while ((match = strstr(string, sub))) {
*match = '\0';
strcat_s(string, n, match + len);
}
}
int main() {
char test[] = "abcxyz123xyz";
removeSubstr(test, "xyz", sizeof(test) / sizeof(char));
cout << test << endl;
}
strcat_s is a more secure version of strcat (MSVC++ likes it way more than strcat :p)
There are better ways of doing this ofc but I'm restricting myself to char arrays here as you requested...