in reply to getting the name of a variable to use as string

You can't.

The internal reason is that the exact same scalar may have multiple names by which it can be accessed at the same time. For instance in the following code the exact same variable is $foo and $bar: 

my $foo = "example";
foreach my $bar ($foo) {
  # Right now $foo and $bar are the same
}

[download]
There is no general way to know which one you'd want back, so Perl doesn't even think of trying to answer that question.

Instead you'll need to step back and think about the problem that you're trying to solve and then solve it in a different way. My feeling is that you're working too hard to avoid a simple line of code. But if you really feel strongly about it, you can achieve the effect with something like this: 

foreach my $var_name (qw(foo bar baz)) {
  $hash{$var_name} = eval "\$$var_name";
}

[download]
but that would probably be a bad idea.

UPDATE: Thinking about it I had the awful idea that someone actually might be able to write a function that stares at the optree for the caller and figures out what the passed in variable was called (in simple cases). This would be an even worse idea than the eval in my books.

Replies are listed 'Best First'.
Re^2: getting the name of a variable to use as string
by pmneve (Novice) on Sep 17, 2004 at 02:26 UTC
    I'm really trying to avoid heavily repetitious code of the form 
    $hash{'foo1'} = $foo1;
$hash{'foo2'} = $foo2;
...
$hash{'foon'} = $foon;

    [download]
    where foo1..foon are column names of a thoroughly denormalized relational table each row of which I want to put into a hash of hashes. Looks like a job for a code generator ;-) Thanks all for your help!
[reply]
[d/l]

      Your thinking appears totally left field. First with DBI you can retrireve rows as hashrefs where the key is already the column name. Next how does the data get into $foo1 in the first place. Why not just put it into a hash in the first instance and then use the hash?

      cheers

      tachyon

[reply]
        I work in left field ;-). I don't always have access to the data directly from a DBI call. I have to deal with unloads that don't carry column names. Being of the mind that the fewer keystrokes I have to type, the fewer dumb errors I'll have to correct I'm looking for a way to take the results of a split 
        my ($a, $b, $c, $d) = split(/\|/,$unloadrow);

        [download]
        and build a hash keyed by the column name (a, b, etc) without having to type each name twice more as in 
        $hash{'a'} = $a;
$hash{'b'} = $b;
...

        [download]
        This may seem lazy but I often have a dozen tables with twenty or more columns to deal with and it would be nice to avoid a lot of error prone and repetitious typing or even cobble up a subroutine to handle building the hash.

        The goal is to build a hash of those hashes so I can look up individual field values to do analysis of the data.

        Thanks!

[reply]
[d/l]
[select]
      First, as tachyon says, I'm quite curious how you've gotten this information into all these variables, and why you didn't just put them into a hash to begin with. But assuming that's done by somebody else in code you don't control...

      You're approaching the problem backwards. Instead of taking a variable and getting the name out, take the name and turn it into a variable, with a symbolic reference. Something like this should work, though I haven't tested it: 

      foreach my $i (qw(foo1 foo2 foo3 ... foon))
{
  $hash{$i} = $$i;
}

      [download]
[reply]
[d/l]
        Not only does it not work, but I already explained why in this thread.

        Furthermore if recommending symbolic references I'd strongly suggest mentioning the problems with using them, and also mentioning how to turn off strict.pm for that block (which they should be using).

[reply]
        I'm trying to put data in the hash as noted in my reply to tachyon. Can't always use DBI (which solves the issue when I can).

        Thanks!

[reply]

In Section Seekers of Perl Wisdom