On May 21, Paul Cotter said:
>> The problem is: PROTOTYPES MUST BE SEEN BEFORE THE FUNCTION IS CALLED.
>
>I've seen this statement before and do not really understand it, having come
>from a 'true-compiler' background. It is the 'seen' that puzzles me.
>
>If I have a 'require' then I believe the prototype is inspected.
>If the subroutine occurs at the front of the 'main' code (ie before being
>used) then I believe the prototype is inspected
>
>However if I have code that goes:
>
>&dummysub( )
>&realsub('paulus')
>.
>.
>sub realsub($)
>{....}
>sub dummysub( )
>{...}
What prototypes do is modify the call to your function:
sub foobar ($$) { ... }
foobar(@this, @that);
# becomes
foobar(scalar(@this), scalar(@that));
foobar(1,2,3,4);
# becomes
# a compile-time error -- 4 args sent, 2 expected
foobar(%this, @that);
# becomes
foobar(scalar(%this), scalar(@that));
sub arrays (\@\@) { ... }
arrays(@a, @b);
# becomes
arrays(\@a, \@b);
arrays(%c, $d);
# becomes
# a compile-time error -- expected '@' leading sigil
That's what they do. Really. And Perl can't modify your source code if
it hasn't seen the instruction to modify it. These modifications happen
at compile time, when Perl is reading in your code for the first time.
Here, compare the opcodes for these two lines:
japhy% perl -MO=Terse
sub foo (\@); foo(@a);
bar(@a); sub bar (\@b);
__END__
UNOP (0x142d00) pp_entersub [2]
UNOP (0x13bbb0) pp_null [141]
OP (0x1287d0) pp_pushmark
>>> UNOP (0x12e280) pp_refgen
UNOP (0x13bb80) pp_null [141]
OP (0x1287e8) pp_pushmark
UNOP (0x12e240) pp_rv2av [1]
GVOP (0x12e220) pp_gv GV (0xce8ac) *a
UNOP (0xbede0) pp_null [17]
GVOP (0xbed80) pp_gv GV (0x12b7d4) *foo
UNOP (0x142dc0) pp_entersub [4]
UNOP (0x13bac0) pp_null [141]
OP (0x128788) pp_pushmark
UNOP (0x142de0) pp_rv2av [3]
GVOP (0x142ca0) pp_gv GV (0xce8ac) *a
UNOP (0x142cc0) pp_null [17]
GVOP (0x142ce0) pp_gv GV (0x14c3d0) *bar
--
Jeff "japhy" Pinyan [EMAIL PROTECTED] http://www.pobox.com/~japhy/
Are you a Monk? http://www.perlmonks.com/ http://forums.perlguru.com/
Perl Programmer at RiskMetrics Group, Inc. http://www.riskmetrics.com/
Acacia Fraternity, Rensselaer Chapter. Brother #734
** I need a publisher for my book "Learning Perl's Regular Expressions" **
