Disadvantage of using 'eq' operator to compare numbers.

[paragkalra]

Hello All,

I have one basic question. Although it may sound silly.

What is the disadvantage of using 'eq' operator to compare two numbers instead of '=='

I mean what bugs are likely to creep in.

[marto]

See the discussion in eq vs. ==.

Martin

[biohisham]

Strings can look like 0 to the == operator and you can get into uncalled for hithces if you swap the string and numeric operators:

$apples = "apples";
$oranges = "oranges";
if($apples == $oranges){
           print "$apples and $oranges are the same";
           }else {
                 print "$apples and $oranges are not the same";
                 }

[JavaFan]

$_ = "Today is 01 Jan 2010";
my ($dayofmonth) = /(\d+)/;
print $dayofmonth eq 1 ? "ok 1\n" : "not ok 1\n";
print $dayofmonth == 1 ? "ok 2\n" : "not ok 2\n";
__END__
not ok 1
ok 2

[ikegami]

If you want one to be equal to one, you want the numerical operators:

>perl -le"print 1 == '1.0' ?1:0"
1

>perl -le"print 1 eq '1.0' ?1:0"
0

If you want two to be less than ten, you want the numerical operators:

>perl -le"print 2 < 10 ?1:0"
1

>perl -le"print 2 lt 10 ?1:0"
0

[dsheroh]

If you want two to be less than ten, you want the numerical operators:

You've got a typo there... s/numerical/string/ (The examples still demonstrate it correctly.)

The thinko was mine, not ikegami's. My comment was incorrect and adds nothing to the discussion; please reap.

[holli]

Not silly.

C:\>perl -e "print 12 == 0xC"
1
C:\>perl -e "print 12 eq 0xC"
1

What the f…?

[afoken]

Looking at print 12 == 0xC and print 12 eq 0xC with the eyes of a parser, I see the following:

• Compare two values using numeric compare (first case, ==) or string-compare (second case, eq)
• LHS value is not a string (no quotes, no bareword rules apply), but a decimal number literal (12), with a value of 12 (decimal)
• RHS value is not a string (no quotes, no bareword rules apply), but a hexadecimal (0x) number literal (0xC), also with a value of 12 (decimal).
• In the first case, both values are already numbers, no need to convert. Both values are constant and equal (12). In the second case, both values have to be converted to strings. After that, both values are constant and equal ("12").
• Optimizing leaves a single true value (1).
• Call the print function with that value.

This works exactly the same with octal constants (014), and makes the code look even more confusing because of the "innocent" zero: Both print 12 == 014 and print 12 eq 014 end printing a 1.

Alexander