C getchar vs scanf

Question

I am confused by a piece of code found in a function I am studying:

char GetCommand( void )
{
    char command;

    do {
        printf( "Enter command (q=quit, n=new, l=list):  " );
        scanf( "%c", &command );
        Flush();
    }
    while ( (command != 'q') && (command != 'n')
           && (command != 'l') );

    printf( "\n----------\n" );
    return( command );
}

void Flush( void ) {
    while ( getchar() != '\n' )
        ;
}

What I don't quite understand here is the usage of the Flush() function. I mean, the book I am reading explains it by saying that it prevents the user from inputting more than a single character and then having that character read when they are prompted for input the 2nd time.

What I don't understand is how Flush() is preventing this from happening. It doesn't DO anything. All it is is a while command. (While this is true......what?????) Doesn't make sense.

Answer 1

getchar() has the side effect of removing the next character from the input buffer. The loop in Flush reads and discards characters until - and including - the newline \n ending the line.

Since the scanf is told to read one and only one character (%c) this has the effect of ignoring everything else on that input line.

It would probably be more clear if the scanf was replace with

command = getchar();

but it's actually a generally bad example as it does not handle End Of File well.

In general scanf is best forgotten; fgets and sscanf work much better as one is responsible for getting the input and the other for parsing it. scanf (and fscanf) try to do too many jobs at once.