string::size_type definition

Can and will someone please give me an explanation and some simple examples on the use of the

Code:

---

string::size_type

---

definition or function?
By the way, which is it a definition or a funcition?

A function has parameters(or void) and a return type.

Btw, your subject doesn't match your message. This is what MSDN says about size_type:

basic_string::size_type
typedef A::size_type size_type;
The unsigned integer type describes an object that can represent the length of any controlled sequence.

There doesn't appear to be anything called string::string_type.

I meant to type "size_type". Thanks, I've corrected it now.
So, where are the examples?

Code:

---

string::size_type index = str.find("XXX");
if (index < 0)
  cerr << "No porn located!\n";

---

if(index < string::npos) cerr<<"No Vin Desil either" << endl;

int i = str.find("ugly hack");
if(i<0) cerr << "No hacks but this one" << endl;

This is still a little fuzzy to me. The :: (scope resolution operator) tells me (and compiler) that size_type is within the scope of the string class. Considering:

Code:

---

string::size_type index = str.find("XXX");

---

index is a string of size_type, which is big or long enogh to "represent the length of any controlled sequence". Is this the correct interpretation?

and,
"string::npos" in

Code:

---

if(index < string::npos) cerr<<"No Vin Desil either" << endl;
int i = str.find("ugly hack");
if(i<0) cerr << "No hacks but this one" << endl;

---

does what?

I would appreciate more explanation and examples.

Quote:

---

index is a string of size_type

---

string::size_type does not mean that size_type is a string.

class string is defined this way:

Code:

---

typedef basic_string<char> string;

---

template class basic_string is declared this way:

Code:

---

template<class E,
    class T = char_traits<E>,
    class A = allocator<T> >
    class basic_string {
    ...
    };

---

this is the declaration of type size_type:

Code:

---

typedef A::size_type size_type;

---

A is the instantiated type of allocator. template class allocator is used to manage storage allocation and freeing for arrays of type T. allocator<T>::size_type is simply an unsigned int describing the length of the array that the allocator is managing.

this is the definition of npos:

Code:

---

static const size_type npos = -1;

---

as you can see for yourself, it is simply the value -1.

Quote:

---

Originally posted by LuckY
allocator<T>::size_type is simply an unsigned int describing the length of the array that the allocator is managing.

this is the definition of npos:

Code:

---

static const size_type npos = -1;

---

as you can see for yourself, it is simply the value -1.

---

This is almost correct. The implemnetors are taking advantage of the fact that they know size_type is unsigned, and that they are using a 2's compliment system. Thus -1 = 0xFFFFFFFF or the largest possible 32 bit unsigned value. However it's best to forget that you knew this, consider.
unsigned i = -1; // ok(though should warn)
if(i == -1) // true
if(-1 < 0) // true
if(i < 0) // FALSE
A big part of the reason for providing size_type's and npos is so that you don't accedentally introduce unexcpected comparisons between signed and unsigned types.

Code:

---

#include <iostream>
using namespace std;

using string::size_type;
int main()
{
        size_type number = 10;
        cout<<number<<endl;
       
        return 0;
}

---

Code:

---

#include <iostream>
using namespace std;

int main()
{
        string::size_type number = 10;
        cout<<number<<endl;

        string sentence = "Matrix Reloaded";
        if(sentence.find('y') == string::npos)
                cout<<"Character not found"<<endl;
        //string::npos is a large number that's returned
        //to indicate not found.
       
        return 0;
}

---