cin not allowing input after first use of function

Hi,

I've been searching around the internet for an answer to my problem for a little bit now but haven't been able to find anything. This being my first big C++ project I've been running in to lots of trouble, most notably with the inability to take input from a cin statement the second time I call one of my functions. The first time through the function everything works as it should but the second time through the cout text is output put cin acts as if the user has already typed in something (although when using my VS 2003 .net debugger I can see that the values remain at 0). Any help would be greatly appreciated.

The file in question contains 3 functions so I'll truncated it down to the one I'm having problems with. The problems are occurring directly at the begining of the file with the cin statements but I included the whole thing just to make sure although I'm doubting there's anything wrong (at least with regards to the cin statements) at the end of the file.

Thanks for any help anyone can bestow upon me.

PHP Code:

---

#include <iostream>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h>
#include "Structures.h"

using namespace std;

int CreateUnit()
{
    // String input code referenced to test code created by Justin Biller

    // ********* So many errors occur with the error checking, I've been messing with it for a while but I don't
    // understand what's going on half of the time.

    char input[5], delimiters[4] = { '-', ' ', ',', 0 };
    char *tokenOne, *tokenTwo, *point ;
    char CUnit = 0;
    int CreateX = 0, CreateY = 0, Team = 0;
    int pause = 0;

    cout << "Enter unit team (1 or 2): ";
    cin  >> Team;

    Team = Team << 7;

    cout << "Enter unit to be created: ";
    cin  >> CUnit;

    if(CUnit > 96)            // if character selection is lower case converts to uppercase so it only checks
        CUnit -= 32;        // one value during the switch statement, needs error checking for invalid input

    if((CUnit == 'I' || CUnit == 'F' || CUnit == 'R' || CUnit == 'B' || CUnit == 'T' || CUnit == 'A' || 
        CUnit == 'H' || CUnit == 'J' || CUnit == 'N' || CUnit == 'S' || CUnit == 'V') != 1)
    {
        cout << "Ivalid input" << endl;
        CreateUnit();
    }

    /* Will take in a string, seperate it into two variables, then set those
    variables equal to numbers that can be used for settings in an array.*/

    cout << "Enter position of created unit (Y,X)(Y-X)(Y X): ";
    cin  >> pause;            // *************** I don't know why this works, more testing to be done tomorrow
    gets(input);            // without it "gets" will not work ********************************************* 

    if(pause != 0)
    {
        cout << "Ivalid input" << endl;
        CreateUnit();
    }
    
    tokenOne = strtok(input, delimiters);
    tokenTwo = strtok(NULL, delimiters);

    if(*tokenOne <= 85)        // converts to array equivilent of the ASCII letter (IE A converted to 0 for position 1)
        CreateY = *tokenOne - 65;
    else
        CreateY = *tokenOne - 97;

    /* I use strtod here to take out the int from tokentwo.  Using strtod on the
    pointer allows us to remove the actual number and place it into an int.  Other-
    wise you would be stuck just trying to use the first digit of the number since
    strtok sets each char in the string to one digit, so if tokentwo took in '10'
    you would only have access to the number '1' since the '1' and the '0' are in
    their own section of memory of the array.  These can be printed with %s, but
    you can't actually use the value itself. */

    CreateX = strtod(tokenTwo, &point);
    CreateX--;

    if (CreateY < 0 || CreateY > 19 || CreateX < 0 || CreateX > 19)    // ****************************
    {                                                                // THESE VALUES WILL NEED TO BE 
        cout << "Ivalid input" << endl;                                // REPLACED BY THE GRID VALUES
        CreateUnit();                                                // ****************************
    }

    //*********************************************************************************************
        
    Infrant.HP=5;            // Health values for every unit
    Flame.HP=5; 
    Rocket.HP=6; 
    ABuggy.HP=10; 
    Tank.HP=20; 
    Art.HP=10; 
    AHeli.HP=8; 
    Jet.HP=12; 
    Bomber.HP=15; 
    SInfran.HP=5; 
    SBuggy.HP=8; 

    switch ( CUnit )
    {
    case 'I' :            // Infantry
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Infrant.HP | Team;
            break;
        }
    case 'F' :            // Flame Infantry
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Flame.HP | Team;
            break;
        }
    case 'R' :            // Rocket Infantry
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Rocket.HP | Team;
            break;
        }
    case 'B' :            // Buggy
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = ABuggy.HP | Team;
            break;
        }
    case 'T' :            // Tank
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Tank.HP | Team;
            break;
        }
    case 'A' :            // Artillery
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Art.HP | Team;
            break;
        }
    case 'H' :            // Attack Helicopter
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = AHeli.HP | Team;
            break;
        }
    case 'J' :            // Jet
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Jet.HP | Team;
            break;
        }
    case 'N' :            // Bomber
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = Bomber.HP | Team;
            break;
        }
    case 'S' :            // Spy Infantry
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = SInfran.HP | Team;
            break;
        }
    case 'V' :            // Spy Buggy
        {
            array[CreateX][CreateY] = CUnit;
            HealthArray[CreateX][CreateY] = SBuggy.HP | Team;
            break;
        }
    }
    
    return 0;
} 


---

Remember I'm able to compile the code and run it once just fine, just those first cins asking for team, unit, and position are giving me problems. Mucho frustrating.

-Peter

***edit***

I forgot to mention... all I recieve as output the second time through the function is a constant barrage of error messages. The "error testing" picks up a bad input somewhere and because cin isn't working the cycle repeats ad infinatum until I close my program. Oh... and I'm compiling in Visual Studio 2003 .net

Note: gets() is a C function. getline() is the C++ way.

A line like this:

cin >>num

waits for the user to enter some input. Then the user can type something in and hit Return. However, when the user hits Return, an invisible \n or "newline" character is entered after the input ('\n' is considered a single character). So, if the user types in 10 and hits Return, the input actually looks like this:

10\n

Now, let's examine how that input is handled. The >> operator is programmed to skip all leading whitespace(e.g. spaces, tabs, and newlines), then read in data, and finally stop reading in data when the first white space character is encountered. And, very importantly, the terminating whitespace character is not read in. So, with this input:

10\n

1) cin>> skips any leading whitespace(in this case there's no whitespace in front of 10)
2) cin>> reads in 10, and
3) cin>> stops reading when it encounters the whitespace character \n
4) cin>> does not read in the \n, and leaves the \n in the input stream.

Sometime thereafter, if these lines are executed:

cout << "\nEnter your name: ";
getline(cin, name);

then that is an instruction to read in more input. Specifically, getline() grabs all the input it can including whitespace until it hits the end of the line, which is indicated by a newline character '\n'. Very importantly, unlike the >>operator, getline() reads in the \n character at the end of the line--thereby removing it from the stream--and discards it. In your case, there is still a \n left in the input stream, and as far as getline() is concerned that is good input, so it doesn't need to wait for any user input, and therefore getline() does its thing and reads in the \n.

The lesson is: if you are going to be switching between the >>operator and getline(), you need to remove the '\n' character that is inserted at the end of the input when the user hits Return. You can remove a trailing whitespace character like this:

cin>>data;
cin.ignore(1);

cin.ignore() will remove the designated number of characters from the stream.

Note that cin.ignore() is not necessary when you do something like this:

cin>>data;

and then the user enters:

10\n

and then you do something like this:

cin>>data2;

Since the >> operator skips leading whitespace, it skips over the \n remaining in the stream and searches for some input, and when it can't find any, it stops and waits for the user to enter something.
-----

Quote:

---

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h>

---

In C++, no legal header file names end in ".h". Those should be cstdio, cstdlib, cstring, and cmath respectively.
---

Also, in C++ we use the <string> type for strings instead of char arrays. string types can be added together with a + sign, and are generally much easier to use than char arrays.

Thank you very much 7stud, worked like a charm. I was also able to get rid of my little "cin >> pause" that I was using as a means to get meaningful input into my "gets" statement.

Another question though... do those headers you mentioned contain the same information as the ones I'm using already? This is my first full year programing and I've only "learned" C thusfur and I'm taking C++ concurrently. Do I lose anything if I go with cmath as opposed to math.h? I'll try throwing them in there but I just want to make sure that it's not going to create problems down the line if I'm trying to write some of my code in C.

Thanks again!

-Peter

***Edit****

Never mind about the header question... once I started typing them in I figured it out after typing in the "c" in front of all my previous #includes

Quote:

---

In C++, no legal header file names end in ".h".

---

->

Code:

---

In C++, no [standard] header file names end in ".h".

---

You can still call your header files header.h, and many people do.

FAQ on flushing the input buffer: http://faq.cprogramming.com/cgi-bin/...&id=1043284392

You might want to use whatever the C++ equivalent of clearerr(stdin) is, too.

>> In C++, no [standard] header file names end in ".h".

This is also wrong. The headers used in the original program were all standard C++ header files (except of course Structures.h). The ones with the .h in the above example are from the C library, and they are allowed to have the .h. However, the use of the C version of the headers is deprecated, which means that it is standard but it might be removed from the standard in the future.

Basically, you should use the versions without the .h (<cstdio>, <cmath>, etc.), but for the C library headers it will still work on standards compliant compilers if you use the C way. The only difference between the versions is the use of the std namespace in the C++ version.