It's a bit subtler than that.
When you write string literals, single-quoting them make them contain exactly the characters you typed, while double-quoting them activates the interpolation mechanism.
On the other hand, when you use a string as a regular expression, it gets interpreted by the regex engine.
For example:
$literal='abc@array$';
@array=(1,2,3);
$interpolated="abc@array\$";
print 'literal: ',$literal,"\n";
print 'interp.: ',$interpolated,"\n";
if ($literal =~ /$literal/) {
print "literal matches itself\n"
} else {
print "literal doesn't match itself\n"
}
$literal2='abc@array';
if ($literal2 =~ /$literal2/) {
print "literal2 matches itself\n"
} else {
print "literal2 doesn't match itself\n"
}
if ($literal2 =~ /$literal/) {
print "literal2 matches literal\n"
} else {
print "literal2 doesn't match literal\n"
}
This prints:
literal: abc@array$
interp.: abc1 2 3$
literal doesn't match itself
literal2 matches itself
literal2 matches literal
The first two lines are hopefully obvious. As for the others:
- $literal does not match itself because, seen as a regexp, it requires the string it matches to end with a y (the $ is interpreted by the regexp engine)
- $literal2 does match itself. This is to confirm that the regexp is not "double-quote-interpolated-expanded-whathaveyou": the @ is still a single character.
- $literal2 matches $literal because it does end in a y.
OK, for the pedants: the string between the // is interpolated, but its contents are not further expanded. I felt that this was more noise than signal, in this case.
|