i can not cin

[kryptkat]

i can not cin

ambiguous overloading

    cin >> "Text added to input stream here " >> variable >> endl ;

ok i get it you can only have

    cin >> variable ;

why does cin not allow you to add text <constant string> to the input stream like you can in other langs ?

second cin question

/* testcin.cpp*/

#include "testcin.h"
#include <iostream>

using namespace std;

// char made global so <in theory> it can be use with the for function inside the header.h 
// <sarcasmmeow> it only works in theory </sarcasmmeow>

    char a;
    char b;

int main() {
          
            cout <<  "input text a  " ;
            cin >>  a ;
            cout << endl ;
            cout <<  "input text b " ;
            cin >> b ;
            cout << endl ;

           dtext( &a , &b ) ;

// also if i use   cout << dtext (&a , &b) << endl ; i get errors
}  // end main

 
/* testcin.h */
#ifndef TESTCIN_H
#define TESTCIN_H

#include <iostream>
using namespace std;
void dtext ( char *pt , char *ytc ) {
 
    cout << "    *pt ==> " << pt << endl ;
    cout << "    *ytc ==> " << ytc << endl ;

    // also as soon as i add a for loop here i loose the vlaues of both variables inside the for loop

     for( int i = 0 ; i < 1000 ; i++ ) {
             cout << "    *pt ==> " << pt << endl ;
             cout << "    *ytc ==> " << ytc << endl ;
             cout << endl ;
             } // end for loop

// after for loop and after loss of value of the variables return answer is not correct
// see above cout note

} //end dtext

#endif

output

input text a g
input text b v

*pt ==> gv
*ytc ==> v

...

note the two letters. it was only supposed to print one. another error. why ? mehow ?

[brute]

ur compilor may hav data size probs...y dont u declare the variables as long......

[Daved]

>> why does cin not allow you to add text <constant string> to the input stream like you can in other langs ?
What would you expect it to do?

>> note the two letters. it was only supposed to print one. another error. why ?
You are outputting pt the pointer, not the character it points to. You just get (un)lucky that when you output a character pointer it is treated like a string and so all characters are shown one at a time from memory until it finds a null character (which in C is how the end of a string is indicated). You get even (un)luckier for ytc, because apparently the character after it in memory is a null.

The solution is to not output the pointer, but output the character it points to. Do you know how to dereference the pointer to get the character?

>> also if i use cout << dtext (&a , &b) << endl ; i get errors
dtext is a function that returns void. This code outputs the return value of the function, but since the function doesn't return anything it fails to compile. Either call dtext or cout a and b, don't mix the two. If a and b have been changed by the function, just outputting them will show it.

[laserlight]

why does cin not allow you to add text <constant string> to the input stream like you can in other langs ?

Because operator>> was not overloaded to perform such a thing, and for what you want to do (add text to the input stream?) it does not make sense anyway.

What would you expect it to do?

I would expect it to attempt to match the object

[kryptkat]

thank you daved that was interesting about the two char output. input formatting was what i was trying to do with the cin overloading.

Do you know how to dereference the pointer to get the character?

i am going to say no. but that is what i thought i did. cin << pt ; so i need something like a = pt ; cin << a ?
in thinking cpp the only thing i recall about dereferenceing is how to change the var ie

int a = 57
int  *ptra = &a;
deff
*ptra = 100 // change value dereff

dtext sends both char a and b to the function. does something with the chars. then returns the result. to do add result return. the main problem was the loss of the chars a and b when they were sent to the function.

thank you laserlight. i was reading about vector and think it can be done with vector. and other string functions. after the input is obtained.

i was shocked to see more than one char on the output. so add "\0" to string after cin ?
surprisingly it compiled.

[Elysia]

It is not surprising that you get this result. It is to be expected. The values are never lost.
The call is correct. But how you print the chars is incorrect.
Remember that the pointer only contains the address where the actual data is stored. Therefore, printing pt will not work as expected. In fact, it works in a rather strange way.
Usually, in C, strings are represented by a char*, where the pointer points to the first char in the string and the last char in the string is detonated by \0 to mark the end of the string.
The 'b' in your string just happens to be lying in memory after the 'a'. So cout prints the a, goes forward 1, prints the b and then just happens to find a \0.
You want to print the char it points to! So you will have to dereference it first! Like so: *pt.
Or you can use references

void dtext ( char& pt , char& ytc )

[Daved]

>> so add "\0" to string after cin ?
No, if you're only trying to do single characters, then don't bother with the '\0', it is just trying to fix faulty code with more faulty code.

Instead, just dereference pt and ytc when you output them. Hopefully by now you've looked up how to dereference a pointer, but to give you a hint you do it in this code:

*ptra = 100 // change value dereff

>> dtext sends both char a and b to the function. does something with the chars. then returns the result.
Is this result a single character, or a change to a and b? If it is a change to a and b, then you don't have to actually return anything, they will get changed automatically. Just call the dtext function in one line and on the next line call cout for a and b.

[kryptkat]

The call is correct. But how you print the chars is incorrect.

i am getting that.

Therefore, printing pt will not work as expected. In fact, it works in a rather strange way.

i am getting that.

You want to print the char it points to! So you will have to dereference it first! Like so: *pt.

like

 cout << "    *pt ==> " << *pt << endl ;

?
still working on how to deference with out changing value. opened up thinking in cpp ebook but it gave no examples of how. all it said was dreff same way as initialize pointer. a little confused.

void dtext ( char& pt , char& ytc )

i was going to try that next. opened up the thinking in cpp ebook saw that.
this is an example prog of the prob. the actual prog is a vigenere prog. char a and char b aka row and column return single char answer.

[Daved]

>> still working on how to deference with out changing value.

Dereferencing and changing the value are two separate things. You dereference a pointer when you want to access the value, but what you do with the value when you access it doesn't matter. So you could dereference to access a value so you can change it, or you can dereference a pointer to access a value just to output it. In this case you are just doing it for output, so it won't change.

[Elysia]

like

 cout << "    *pt ==> " << *pt << endl ;

?

That is correct.

[brewbuck]

I would expect it to attempt to match the object

class RequireMatch
{
public:
    RequireMatch( const std::string &pattern ) : mPattern( pattern ) {}

private:
    std::string mPattern;
    friend std::istream &operator>>( std::istream &, const RequireMatch & );
};

std::istream &operator>>( std::istream &stream, const RequireMatch &match )
{
    // cleverness goes here
}

...

std::cin >> RequireMatch( "This text must be matched" ) >> ...;

It comes in handy.

[laserlight]

It comes in handy.

Yes, in response to a comment you made I posted a similiar working example last year. When I submitted an interest check to the Boost mailing list, Scott McMurray pointed out that it would be simpler if this was actually integrated into appropriate overloads of operator>> instead of being a manipulator (but it would not be so good as overloads outside of the standard library itself due to the overloads possibly not being found by argument dependent lookup).

Going further off topic, would you like to work with me on a proposal to get this included into the C++ standard library?

[kryptkat]

ok got it working. thank you all.