Hello
Some time ago I read Effective C++ and it advises programmers to use fgets() instead of get because you can specify how many characters to read, and thus avoid memory corruption if the input is too long.
Several new books (such as Prata or Professional C++) don't speak about fgets.
Is it because we don't need it anymore?
cin.get() and cin.getline() can also specify the size to read, and I think other functions too.
What's the difference between cin.get and fgets?
In C++, C functions aren't that important because you can use C++ libraries.
The idea behind the typical fgets() loop in C can be expressed in C++ as
That's pretty safe.Code:#include <iostream> #include <string> using namespace std; string buff; while (getline(cin, buff)) ...
OK, thanks. And what happens if the input is bigger than the size that can be fit in memory?
If a string needs more memory it will automatically reallocate itself. If that allocation should fail, the string is unaltered, and a type of runtime exception called bad_alloc is thrown.
There are hard limits on the size of a string, which come from the range of numeric values it uses: The size_type used to report the length of the string and access characters does have a maximum value. This number is quite large. If you are curious about the limits of your string implementation, you can call max_size().
If you use other methods, like cin.getline() or cin.get(), they can be called to act much like fgets() does. The major difference between the two is that the newline, if it can be stored, is actually discarded by cin.getline() before returning, and fgets() does not do this. If you use cin.get() to read a line, then the newline will be retained.
Last edited by whiteflags; 08-13-2013 at 09:45 PM.
On the C side, still always avoid gets() like the plague. It's actually been removed from C11, as it's just such a terrible function. fgets() is fine.
But yes, in C++ there's no need to worry about gets/fgets if you're happy using get() and getline().
