How can I use a "goto" statement to break out of a loop
for(i = 0; (i < 9); i++)
{
for(j = 0; j < 9; j++)
{
//cout << " " << Matrix[i][j];
//cout << "i: " << i << endl;
if(Matrix[i][j] == 0)
{
//temp = 10;
[goto] ;
//break;
}
}
}
I wanted to keep the values at which i and j were when I left the nested for loop. How can I use a goto statement for that?
Like this:
int i,j;
for(i = 0; i < 9; i++)
{
for(j = 0; j < 9; j++)
{
//cout << " " << Matrix[i][j];
//cout << "i: " << i << endl;
if(Matrix[i][j] == 0)
{
//temp = 10;
goto end;
//break;
}
}
}
end:
cout << i << " " << j << endl;
goto to out of the current scope (specially a loop's scope) finishes it properly?
goto that prevent programmers from making improper jumps, e.g. jumping over initialization.Just as you use goto in any other situation. As long as you don't cross scopes with local variables in them, you can actually think of it as "goto this and that line":
for (/* ... */) {
/* ... */
if (/* ... */)
goto finalise;
}
finalise:
foo = bar; //...
However, there are many situations when goto is an indicator for not well designed code. By no means always, but often.
I suggest you use gotos big brother return and factor out your code into a function:
inline std::pair<int,int> findZeroEntry(std::vector matrix) {
for (int i = 0; i < 9; i++)
for (int j = 0; j < 9; j++)
if (Matrix[i][j] == 0)
return std::make_pair(i,j);
return std::make_pair(9,9); // error
}
Well, @bitmask's answer has already said most of what I thought of saying when I read the question.
But for completeness, here's another technique:
Matrix m;
Index2D< 9, 9 > pos;
for( ; pos < pos.end(); ++pos )
{
if( m( pos.x(), pos.y() ) == 0 )
{
break;
}
}
cout << pos.x() << " " << pos.y() << endl;
IMHO this is far more clear code.
Also, the matrix can then be made to support indexing via Index2D values, thus reducing the above to just …
Matrix m;
Index2D< 9, 9 > pos;
for( ; pos < pos.end(); ++pos )
{
if( m[pos] == 0 )
{
break;
}
}
cout << pos.x() << " " << pos.y() << endl;
Since there’s nothing like Index2D in the standard library, it needs to be defined somewhere, e.g. like
template< int width, int height >
struct Index2D
{
int i_;
int x() const { return i_ % width; }
int y() const { return i_ / width; }
void operator++() { ++i_; }
bool operator<( Index2D const& other ) const
{
return (i_ < other.i_);
}
Index2D(): i_( 0 ) {}
Index2D( int const x, int const y )
: i_( width*y + x )
{}
static const Index2D endValue;
static Index2D end() { return endValue; }
};
template< int width, int height >
Index2D< width, height > const Index2D< width, height >::endValue( 0, height );
But then it can be reused everywhere that you need this functionality.
endValue local to end().You don't need goto to leave the nested for loop and save off those variables. Merely, you want to break out of each loop successively. Just have a boolean at the appropriate scope level that you check to know if need to break out of the loops.
Corrected example
IE:
bool HasFoundZero = false;
for(i = 0; i < 9; i++)
{
for(j = 0; j < 9; j++)
{
//cout << " " << Matrix[i][j];
//cout << "i: " << i << endl;
if(Matrix[i][j] == 0)
{
//temp = 10;
HasFoundZero = true;
}
if(HasFoundZero)
{
break;
}
}
if(HasFoundZero)
{
break;
}
}
goto. It's still a goto, just with a boolean to obfuscate and slow things down. Why obfuscate? Just use goto. I'd prefer to see that verboten goto used to break out of an inner loop over a boolean that hides the fact that a goto is being used. Other languages provide a multi-level break. C/C++ does not. That C/C++ does not is a language defect.
i and j where the object was found (i in this case will be the location plus one)
gotos. In fact, why ever use break, by that logic? break is safer in usage. There's a reason why it exists. Your logic could even extend to switch or while being unnecessary... @DavidRodríguez-dribeas: I corrected the example. Any good compiler will optimize further. More importantly, this is easier to maintain than gotos and labels. As well, in this simple example, is there anything to suggest that one condition test is a performance concern?
