Printable View
Hi,
Please take a look at this
When i input a non integer value such as a char 'a' , this segment of the loop executes infinitely. I believe the cin stream have set failbit because of the incorrect format of input. However, how can i set it back to goodbit? I tried cin.clear() but it didnt work.Code:
int choice;
do{
cin >>choice;
switch{
case 1: cout << "Valid choice" <<endl;
case 2: cout<< "leaving loop" << endl;
default: cout << "Invalid choice" << endl;
}
}while(choice !=2)
Thanks
sorry some error there. here is the amended code
Code:
int choice;
do{
cin >>choice;
switch(choice){
case 1: cout << "Valid choice" <<endl;
case 2: cout<< "leaving loop" << endl;
default: cout << "Invalid choice" << endl;
}
}while(choice !=2)
urgh!!! I forget the breaks too
Code:int choice;
do{
cin >>choice;
switch(choice){
case 1: cout << "Valid choice" <<endl;
break;
case 2: cout<< "leaving loop" << endl;
break;
default: cout << "Invalid choice" << endl;
}
}while(choice !=2)
When you enter a character instead of an integer cin goes into a failed state. To recover from this you need to clear the stream flags and remove any characters left in the stream, then try again. Here's one simple (but slightly imperfect) way to do this, you get the idea:
Code:#include <iostream>
#include <limits>
using namespace std;
int main()
{
int number;
cout << "Enter an integer value: " << flush;
while(!(cin >> number)){
cin.clear();
cin.ignore(numeric_limits<int>::max(), '\n');
cout << "\nInput Invalid. Please Try Again: " << flush;
}
cin.ignore(numeric_limits<int>::max(), '\n');
cout << "You entered: " << number << endl;
}
Sorry, i am new to C++ and just have some knowledge of C.So what this mean:
To my knowledge, ignore() is used to ignore the next character, unless u specify the number of characters to ignoreCode:cin.ignore(numeric_limits<int>::max(), '\n');
ok i think i made some understanding.
When cin takes in an inproper format, it sets failbit and the input still remains in the buffer stream. After we set it back to good state, we have to discard the inproper input. But how do we know how many characters to discard?Code:cin.ignore(numeric_limits<int>::max(), '\n');
This line:
cin.ignore(numeric_limits<int>::max(), '\n');
will remove everything from the stream up to the next newline. It look more complicated than it is; that's C++ syntax for you. It's effectively:
cin.ignore(some_large_number, '\n');
You could instead use:
cin.ignore(1000, '\n');
which doesn't look as frightening, but then you still run the (slight) risk that there will be more than 1000 characters of garbage left in the stream.
As you rightly suggest, used on its own:
cin.ignore();
will remove the next (single) character from the stream.
Allow me to clarify my understanding:Code:cin.ignore(numeric_limits<int>::max(), '\n');
So the code above will ignore a max number of characters determined by numeric_limits<int> or unless it encounters the delimiter '\n' first. I tried reading up, but i cant find what does the max() mean. Could you explain? Is it necessary there?
And also is there any other C++ functions that can flush the stream like fflush() in C,so that i dont have to specify the number of characters to ignore? It seems more efficient this way and i suppose <iomanip> has such function for flushing stream?
Thanks
:)
Actually you should not be flushing input streams with fflush() in C. fflush() is not defined for input streams according to the standard. Some compilers have allowed it to work but it's one of those bugs waiting to happen if you change compilers.Quote:
Originally posted by Raison
And also is there any other C++ functions that can flush the stream like fflush() in C,so that i dont have to specify the number of characters to ignore?
oh my!!! i been flushing since i learned C.
Shame on you. Go to your corner and think about what you've done! ;)
I am just a newbie but wouldn't it be easier just to change the while condition to <=2 so that any other value greater then 2 would exit the loop?
I guess the other option would be to set choice=2 in the default section of the switch so it would then exit the loop on any invalid output;
This would work, and is a better option than above. What if you added 4 as a good option?Quote:
Originally posted by manofsteel972
I guess the other option would be to set choice=2 in the default section of the switch so it would then exit the loop on any invalid output;
i was taught to just do:
to flush the buffer...Code:while ( cin.get() != '\n' ){}
this sounds like a logical method too. Thanks for the information.
I also have a question regarding this isssue.
I recently had problem that I wasn't able to solve:
consider this:
And this works fine if you expect user to enter integer and user enters "fsdfsdfs"Code:
#include <limits>
#include <iostream>
using namespace std;
int main()
{
int x;
cin>>x;
while(!cin.good())
{
cout<<"something is bad";
cin.clear();
cin.ignore(numeric_limits<int>::max(), '\n');
cout << "\nInput Invalid. Please Try Again: " << flush;
cin>>x;
}
return 0;
}
but what if user enters 4.56 as value, then everything goes silently and rest of number stays in
stream so next calling ( for example cin>>x) whil pick up the rest from stream. How can I test
to see it it was integer that userentered or not (then I need to take appropriate action like
ignore() etc)?
I can't believe that nobody tried to answer this!
OK, here's a shot without a compiler to assist.
expected input: int like 3
actual input: decimal value like 3.14
input type: extraction operator, >>
outcome: the >> operator will extract the 3 from the input buffer and place it in some temporary storage to see if there is another digit next in line, to make 33 for example, as opposed to just 3, or whether there is something else there. It will then extract then extract the decimal point. Since the decimal point isn't valid in an int the fail bit for cin will be flipped and the decimal point will probably remain in the input buffer together with the 14. The success/failure of the action can be assessed, the failbit can be reset if needed, and the input buffer could then be cleared of any remaining material the same way as before.
Another alternative would be to accept user input only as a string, not as a numerical value. Then you can parse the string to see if erroneous material is present. If input is appropriate for the type you eventually want to use, then you can convert the string input into a numerical value using atoi() or atol() as appropriate. In fact, this may be how cin evaluates input and decides that inappropriate material is present and therefore resets the fail bit of the state variable in the first place, but I can't say that for sure.
That's ok, but if you can give me some code example
//version trying to use built in methods of cin
version using input as string and manually parsing string for validationCode:int num;
bool acceptInput = true;
while(acceptInput)
{
cout << "enter an integer" << endl;
cin >> num;
if(cin.fail())
{
//not sure if why failed so don't accept input
cin.clear();
cin.ignore(50000, '/n');
cout << "inappropriate input, try again" << endl;
}
else
acceptInput = false;
}
NOTE: none of the above code has been compiled or tested.Code:char input[100];
int i;
int num;
bool acceptInput = true;
bool accpeptable = true;
while(acceptInput)
{
//reset flag each time through loop
acceptable = true;
//request input
cout << "enter an integer" << endl;
//store input as string
cin >> input;
//look at every char of string
for(i = 0; i < strlen(input); ++i)
{
if(isdigit(input[i]))
{
//do nothing
}
else
{
//there can be nothing but digits in ints
acceptable = false;
//stop for loop
i = strlen(input);
//notify user
cout << "inappropriate input. try again." << endl;
}
}
//if all char in input were digits
if(acceptable)
{
//stop while loop
acceptInput = false;
//convert string to int
num = atoi(input);
}
}
Possible bug with second version:
what if user input 123456789. This input would pass the manual validation process but be too big for a valid int. Suspect that output would be same as if user input same value and code accepted input directly into int variable, but I won't swear to it. The second version might be more robust if input were converted to type double first, the value of the double is evaluated, if the value of the double is above the upper limit of an int, then the user is notified of inappropriate input and the process continues; but I'll leave that up to you to do.
I don't think the fail bit will be set in this case. I believe the stream will still be in a good state; it will just stop extracting from cin when it encounters the first non-digit character, i.e. the decimal point.Quote:
Originally posted by elad
expected input: int like 3
actual input: decimal value like 3.14
input type: extraction operator, >>
outcome: the >> operator will extract the 3 from the input buffer and place it in some temporary storage to see if there is another digit next in line, to make 33 for example, as opposed to just 3, or whether there is something else there. It will then extract then extract the decimal point. Since the decimal point isn't valid in an int the fail bit for cin will be flipped and the decimal point will probably remain in the input buffer together with the 14.