Recognise a word inside a string

[Aramil of Elixia]

whats lambda i see that every where?

[chrisname]

Google search "lambda function C++" or something, there's plenty of information. It's not just a C++ thing, but since you're presumably more comfortable with C++ than other languages it's probably easiest for you to learn lambda functions that way.

[xantavis]

The same error messages appear.

I prototyped the bool in the 'Talk.h' under public

bool Search (std::string answer, std::string word);

and I have cut the bool out of the void.

[chrisname]

This code doesn't produce any errors.

#include <iostream>
#include <string>
#include <cctype>
#include <cstdlib>
#include <ctime>

using namespace std;

class Talk
{
    public:
        Talk();
        void Voidex ();
};

bool Search(string sentence, string word) {
    size_t pos = 0;
    while ((pos = sentence.substr(pos).find(word)) != string::npos)
    {
        if (!(isalpha(sentence[pos - 1])) || !(isalpha(sentence[pos + word.size() + 1])))
            return true;
    }
    return false;
}

Talk::Talk() {  }

void Talk::Voidex () {
    string x;
    getline (cin, x);
    int wert = 0;
    if (Search(x,"less"))
        wert = 1;
    else if (Search(x,"more"))
        wert = 3;
    cout << wert;
}

int main()
{
    // ...
    return 0;
}

[xantavis]

Yes, but I want to make it in a class and not in the main (alot of stuff will come in this class, so I dont want it to cramp my main.cpp

[chrisname]

Then split it into separate files.

[xantavis]

Oh, Im sorry!
I, accidentally, put the bool inside the constructor! So the code works now, but 1 last question: I thought that if you use the isalpha thing, only full words are being searched. But when I let the program look for the word "we", for example, the word well also gets listed. So why use isalpha then?

[chrisname]

Sorry, that was a mistake/oversight on my part. This code works:

bool contains_word(const std::string& sentence, const std::string& word)
{
	size_t pos = -1;
	// Find the start of 'word' in 'sentence'.
	while ((pos = sentence.find(word, pos + 1)) != std::string::npos) {
		// Make sure it's actually a word (e.g. "rain" could be part of
		// "drain"). isalpha checks if the given character is in the
		// alphabet. Here we check if the characters before and after
		// 'word' are alphabet characters. If neither of them is an
		// alphabet character then we've found a whole word.
		bool start = ((pos < 1) || !(isalpha(sentence[pos - 1])));
		bool end = ((pos + word.size() > sentence.size() || !(isalpha(sentence[pos + word.size()]))));
		if (start && end)
			return true;
	}
	// If we get to here then we didn't find any instance of 'word' that was
	// a whole word.
	return false;
}

isalpha() is a function which returns 1 (true) if the character you give it is in the alphabet. The variable
bool start = ((pos < 1) || !(isalpha(sentence[pos - 1])));
is false if there is an alphabet character before the word, while the variable
bool end = ((pos + word.size() > sentence.size() || !(isalpha(sentence[pos + word.size()]))));
is false if there is an alphabet character after the word. If both of them are true then it means that a whole word has been found, so the function returns true. If it can't find a match to "word" which is a whole word, it returns false.

[xantavis]

Actually, I just changed the code like this, so that it works:

    if ((pos = sentence.substr(pos).find(word)) != string::npos)
    {
        if (!(isalpha(sentence[pos - 1])) && !(isalpha(sentence[pos + word.size() + 1])))
            return true;
    }

So what I did was:
1. change the while to if
2. change the '||' to '&&'

And actually, I could merge the ifs to 1 if, if I wanted...

Still, Thanks for your Help!!

[chrisname]

Bare in mind that:
1. Since you're no longer doing it in a loop, it will only work if the first match is a whole word. What I mean is, without a loop, that function wouldn't find "rain" in the sentence "drain the rain" because the first match isn't a whole word, and it only looks at the first match. With a loop it looks at all matches.
2. You will get an error at sentence[pos - 1] if pos = 0 (because you'll be trying to access sentence[-1] which is not valid). This will be a problem any time the word is at the start of the sentence (like if you were searching for "hello" in "hello world").