Perl's a great language for special variables - variables that are set up without the programmer having to intervene and providing information ranging from the number of lines read from the
current input file ($.) through the
current process ID ($$) and the
operating system ($^O). Other special variables effect how certain operations are performed (
$| controlling output buffering / flushing, for example), or are fundamental in the operation of certain facilities - no more so than $_ and @_.
Lets clear a misconception.
$_ and
@_ are different variables . In Perl, you can have a list and a scalar of the same name, and they refer to unrelated pieces of memory.
$_ is known as the "default input and pattern matching space". In other words, if you read in from a file handle at the top of a
while loop, or run a
foreach loop and don't name a loop variable, $_ is set up for you. Then any regular expression matches,
chop s (and
lc s and many more) without a parameter, and even
print s assume you want to work on $_. Thus:
while ($line = <FH>) {
if ($line =~ /Perl/) {
print FHO $line;
}
print uc $line;
}
Shortens to:
while (<FH>) {
/Perl/ and
print FHO ;
print uc;
}
@_ is the list of incoming parameters to a sub. So if you write a sub, you refer to the first parameter in it as
$_[0] , the second parameter as
$_[1] and so on. And you can refer to
$_# as the index number of the last parameter:
sub demo {
print "Called with ",$#_+1," params/n";
print "First param was $_[0]/n";
Note that the
English module adds in the ability to refer to the special variables by other longer, but easier to remember, names such as @ARG for @_ and $PID for $$. But use English; can have a detrimental performance effect if you're matching regular expressions against long incoming strings.
