1

Running the following, I would expect to receive N, Y, Y.

I understand why I'm not, because '0.00' != '0' for the second example, but is there a consistent way of testing for 0 without casting back to a float/double, and without dropping the === to a ==.

echo bcmul( '5.1', '2.234', 2 );
echo bcmul( '5.1', '2.234', 2 ) === '0' ? '  Y  ' : '  N  ';
echo "<br/>";

echo bcmul( '0.00', '000.00', 2 );
echo bcmul( '0.00', '000.00', 2 ) === '0' ? '  Y  ' : '  N  ';
echo "<br/>";

echo bcmul( '0', '0', 2 );
echo bcmul( '0', '0', 2 ) === '0' ? '  Y  ' : '  N  ';
echo "<br/>";

Notes

Why don't I want to drop the ===?

If I'm providing out functionality as part of a grander project, via a method such as get_total_cost(), I don't feel that it's intuitive to other developers to have to drop the strict comparison, when they are expecting the function to return a numeric value as a string.

Charles
  • 50,943
  • 13
  • 104
  • 142
Chris
  • 54,599
  • 30
  • 149
  • 186

2 Answers2

2

A. Yes 0.00 !== 0 is valid because they are not the same type 

var_dump(0.00,0);

Output 

float 0
int 0

B. 0 !== "0" is valid because they are not the same type

var_dump(0,"0");

Output 

int 0
string '0' (length=1)

C. Why don't I want to drop the ===

var_dump("hello" == 0 );  true
var_dump("hello" === 0 );  false

Conclusion

$a === $b TRUE if $a is equal to $b, and they are of the same type.

I guess this is what you want 

echo (int) bcmul('0.00', '000.00', 2) === (int) '0' ? '  Y  ' : '  N  '; 
       ^                                    ^
Baba
  • 94,024
  • 28
  • 166
  • 217
0

You could try using ctype_digit() to determine if the returned string contains a "clean" int, or if it contains a floating point dot somewhere, and then just have two sets of Y/N functions; One for string-ints and one for string-floats.

Desty Nova
  • 34
  • 1