Question

I run this piece of code on Visual C++ 2010

char c[10];
cin.get(&c[0],5);
cin.get(&c[2],4);
cout << c << endl;

and if I feed “123456789” to cin, the cout clause will print “12567”, which is the result I expected.

But if I write:

char c[10];
cin.getline(&c[0],5);
cin.getline(&c[2],4);
cout<< c <<endl;

and feed the same string, it will only show me “12”, where c=={‘1′,’2′,’\0′,’4′,’\0’}

According to the documentation, the difference between cin.get and cin.getline is that cin.get does not discard the delim character as cin.getline does, so I don’t know why this happens. Can anyone give me hints?

Answer 1

What is happening is that if basic_iostream::getline() reaches the limit of characters to be read (the streamsize argument minus 1), it stops reading then places a null character after the data it has read so far. It also sets the failbit on the stream.

So assuming that the stream has "123456789" ready to read, when you call cin.get(&c[0],5) the array will get {'1','2','3','4','\0'} placed into elements 0 through 4. And the failbit is set on the stream.

Now when you call cin.get(&c[2],4), the failbit is set on the stream, so nothing is read. The getline() call does nothing but place the terminating null into the array at index 2 (even if nothing is read from the stream, getline() will place the null character – even if the non–read is because of the failbit). So the array now looks like:

{'1','2','\0','4','\0'}

The documentation you link to mentions this:

If the function stops reading because this size is reached, the failbit internal flag is set.

But getline() does a lot, so it’s easy to miss that detail.