I don't think you can construct a reference to a builtin. References are to things with memory addresses and builtins don't have any (if i am wrong, someone will yell at me).
This will more or less accomplish what you want to do, though:
my %hash = (
foo => sub {uc shift},
);
print $hash{foo}->("hello\n");
HTH, HAND. | [reply]
[d/l] |
This seems like an odd thing to do. Just out of curiosity, why do you want to do it?
When's the last time you used duct tape on a duct? --Larry Wall | [reply] |
I'm validating input to a function, via named parameters,
that inserts stuff into a database using DBI. For each
parameter passed (say %param), I run it through a bunch of validation
routines, massage it and spit out what, if all tests pass,
should be inserted into the database...my validation routines
for each parameter might be different and are stored as
references in a hash. ie.
%validationHash = (
"name" => {
"regexp" => "string",
"status" => "required",
"outputsub" => \&massage;
},
"phone" => {
"regexp" => "phone",
"status" => "optional",
"outputsub" => \&mungephonenum,
},
);
then I iterate over the keys in the hash (ie.$param{'name'})
(thereby ignoring any parameters passed that I don't care
about), make sure it is defined if
$validationHash{"name"}{"status"} eq "required" otherwise
return undef etc. then make sure that it looks like a
"string" with a routine that runs a regexp match and finally,
return a value $validationHash{'name'}{'outputsub'}->($param{'name'})
as what should be inserted into the database.
Needless to say, I would love to hear about nice ways of
validating input and even better, ways of constructing SQL
statements given such input (I'm considering using
DBIx::Recordset for the latter, but I've done enough work
already that I might have to come back to it at another time).
BTW, the first solution given worked very well and I am
using it successfully, thanks.
-Adi | [reply]
[d/l] |