catching non-standard input & returning error

hi,

Wondering if it's possible to catch non-standard input & then use that for error checking/return ?

for example;

Code:

---

int input;
cout<<"Enter input: ";
cin>>input;

if (input!=int)
{
  cout<<"Not an int";
}

---

This is (a part of) something I've written, and while it does what I want, I was wondering if there's a more proper/elegant way of handling non-standard input.
Currently I've used an initialised int using -1 as a means of doing what I want (ie catching non ints).

Code:

---

//myflush1 & 2 added/used to handle non-int/standard input.
void myflush1 ( std::istream &in )
{
  in.clear();
}

void myflush2 ( std::istream &in )
{
  in.ignore ( numeric_limits<streamsize>::max(), '\n' );
  in.clear();
}
//other code



  int iNumoftimetype = 11; //number of time types in saTimetype[], counting from zero.
  int iInitialtimetype=-1, iConverttimetype=-1; //initialised to -1 in order for 0 > exit to work
  double dInitialtimevalue=-1, dConverttimevalue=-1; //used for display & conversion values
  string saTimetype[] = {"Exit Program","millisecond(s)","second(s)","minute(s)","hour(s)","day(s)","week(s)","month(s)"
                              ,"year(s)","decade(s)","century(ies)","millennium(s)"}; //used for text display
  string saFunctimetype[] = { "INVALID", "MILLISECONDS","SECONDS","MINUTES","HOURS","DAYS","WEEKS","MONTHS"
                              ,"YEARS","DECADES","CENTURYS","MILLENNIUMS" }; //used for function call
  string sInitialtimetype, sConverttimetype;
  string sFuncinitialtimetype, sFuncconverttimetype, sTime1_2_time2_func;

  while (1)
  {
    cout<<"Select base type & conversion type (enter value 1 or 2 or 3 etc)...\n";
        for (int n=0 ;n<=iNumoftimetype;n++ ) //numbers & displays the saTimetype array
        {
          cout<<n<<":"<<saTimetype[n]<<endl;
        }
    cout<<"Enter base type: ";
    myflush1 ( cin>>iInitialtimetype);
    myflush2 ( cin );

      if ((iInitialtimetype>=1)&&(iInitialtimetype<=iNumoftimetype)&&(iInitialtimetype!=0))
      {
        cout<<"Convert to: ";
        myflush1 ( cin>>iConverttimetype);
        myflush2 ( cin );
       
          if (iInitialtimetype == iConverttimetype)
          {
            cout<<"Conversion will be the same\n\n";
            iInitialtimetype=-1;
            iConverttimetype=-1;
            continue;
          }
        if ((iConverttimetype>=1)&&(iConverttimetype<=iNumoftimetype)&&(iConverttimetype!=0))
          //if reach here, exit the loop & moves onto switch.
        break;
       
          if ((iConverttimetype<-1)||(iConverttimetype>iNumoftimetype))
          {
            cout<<"Invalid selection\n\n";
            iConverttimetype=-1;
            continue;
          }

          //here, if input!=int return error & return control to loop
          if (((iConverttimetype<=-1)||(iConverttimetype>=iNumoftimetype))||(iConverttimetype!=0))
          {
            cout<<"Invalid selection type. Must be a (positive) integer.\n\n";
            continue;
          }

          if (iConverttimetype==0)
          {
            cout<<"Program exiting\n";
            cin.get();
            return 0;
          }
      break;
      }

      if ((iInitialtimetype<-1)||(iInitialtimetype>iNumoftimetype))
      {
        cout<<"Invalid selection\n\n";
        iInitialtimetype=-1;
        continue;
      }

      if (((iInitialtimetype<=-1)||(iInitialtimetype>=iNumoftimetype))||(iInitialtimetype!=0))
      {
        cout<<"Invalid selection type. Must be a (positive) integer.\n\n";
        continue;
      }

      if (iInitialtimetype==0)
      {
        cout<<"Program exiting\n";
        cin.get();
        return 0;
      }
  }

//other code

---

I thought I had read somewhere that there is a way to do this, but I can't seem to find where that was & I'm starting to think maybe I dreamt it up :rolleyes:

O_o

Well, I'm not going to bother with that code beyond a cursory glance.

You shouldn't arbitrarily clear the stream when an operation fails. (You are clearing the 'eofbit' flag.)

You should return a reference to the stream object. (You could then use the result of the function to control branching.)

You should store the result of stream operations--or otherwise use the result. If a stream operation leaves the stream in a good state, the relevant variable was modified--even if it has the same value. (With this implemented you don't need to employee a sentry value.)

I've updated my example. You should normally use the result of a stream operation to control branching, but here I've cleared everything but the 'eofbit' flag anyway so controlling termination with 'bool std::istream::eof()' is fine. (You can't use the implicit conversion to "boolean convertible type" because it will not loop if the first input is invalid.)

Soma

Code:

---

#include <iostream>
#include <limits>

std::istream & maybe_clear
(
  std::istream & in_f,
  bool & modified_f
)
{
  if(!(modified_f = !((in_f.fail()) && (!in_f.eof()))))
  {
      in_f.clear();
      in_f.ignore(std::numeric_limits<std::streamsize>::max(), '\n');
  }
  return(in_f);
}

int main()
{
  const int min(5);
  const int max(20);
  int i(0);
  for(bool updated(std::cin >> i); !std::cin.eof(); maybe_clear(std::cin >> i, updated))
  {
      if(updated && (min <= i) && (max >= i))
      {
        std::cout << i << '\n';
      }
  }
  return(0);
}

---

The thing you're missing is the checking of the return value of operator >> to see if the input succeeded.

Here's a simple method that only checks whether the input is an int or not:

Code:

---

int input;
cout<<"Enter input: ";

while (!(cin>>input))
{
  cin.clear();
  cin.ignore(numeric_limits<streamsize>::max(), '\n');
  cout<<"Not an int";
}

---

You can also use the functions in this post to handle almost all of your basic input needs in a more streamlined fashion: http://cboard.cprogramming.com/showp...4&postcount=24

Thank you both for the replies. This has greatly helped & is exactly what I was looking for. Still trying to understand many aspects, so thanks ;).