Creating lists using a hash with the same key/values

Hagbone has asked for the wisdom of the Perl Monks concerning the following question:

Often, when coding scripts, I encounter situations where I need to test to see if a variable matches an item that's stored in a list of values.

One approach I've used is to maintain an array that is reviewed for the match using a "foreach" type iteration.

Another approach is to maintain a hash, and test for the presence of a hash value that matches.

It seems that using the second approach is more efficient from both an operating system standpoint, and a coding standpoint:

foreach (@whatever) {
    if (/^does this match$/) {
        do something;
    }
}
[download]

Vs.

if (/$hash{$whatever}/) {
    do something;
}
[download]

Assuming I'm thinking correctly so far, my next question has to do with creating the hash value. Is there anything "wrong" with populating a hash with keys and values that are the same? For example:

%SampleHash (
    'whatever' => 'whatever',
    'whoever' => 'whoever',
)
[download]

It feels like there's something wrong (or maybe more accurately, inefficient) using a hash this way, but I'm not smart enough to know if that is in fact true.

When I use hashes as described above, things seem to work OK ... I'm just wondering/hoping that some of the brainier participants here might be able to point out errors in my thinking, and/or more appropriate ways of how to handle this type of situation.

Comment on Creating lists using a hash with the same key/values Select or Download Code

Replies are listed 'Best First'.
Re: Creating lists using a hash with the same key/values by Trimbach (Curate) on Jan 01, 2003 at 23:53 UTC
The idiom of using hashes as a fast and easy lookup table is a common one in Perl. Hashes are quick, easy to setup and lend themselves well to a boolean-type comparisons like what you're talking about. The only thing is your example of `%SampleHash ( 'whatever' => 'whatever', 'whoever' => 'whoever', )` [download] is more commonly coded as `%SampleHash ( 'whatever' => 1, 'whoever' => 1, )` [download] ...simply because if you ain't gonna be using the values of the hash, there's no reason to make the value any longer than absolutely necessary. Making the value a "1" does the trick, and it doesn't waste memory. Of course, you're probably not likely to have a hash of values that's so big for the memory cost to matter much, but hey, waste not want not. :-) Gary Blackburn Trained Killer	[reply] [d/l] [select]
Re: Creating lists using a hash with the same key/values by tachyon (Chancellor) on Jan 01, 2003 at 23:55 UTC
Idiomatically you would do this: `my @list = qw( foo bar baz ); my %hash; $hash{$_}++ for @list; # now do a lookup if ( $hash{$look_this_up} ) { # blah } use Data::Dumper; print Dumper \%hash __DATA__ $VAR1 = { 'foo' => '1', 'baz' => '1', 'bar' => '1' };` [download] This populates the hash with identical values (ie 1). It means that $hash{$value} will be true if $value is in the hash or false if it is not. We are really only interested in the keys for the fast lookup but having a true value lets us avoid having to use exists cheers tachyon s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print	[reply] [d/l]
Re: Creating lists using a hash with the same key/values by atcroft (Abbot) on Jan 01, 2003 at 23:52 UTC
Why not do something like `$SampleHash{'whatever'}++;`, so all you would have to do is say something along the lines of `&whatever if ($SampleHash{'whatever'});` for instance...	[reply] [d/l] [select]
Re: Creating lists using a hash with the same key/values by djantzen (Priest) on Jan 02, 2003 at 00:01 UTC
I tend to agree with your hesitation Hagbone. I try to avoid using hashes merely for looking up keys and throwing away the values, but that's mostly a religious perspective on my part, and in many cases that's the cleanest way to code a lookup function. A couple of observations: first, in this situation it is common to initialize the hash using a hash slice where your values are '1'. So, `my %lookup_hash = (); %lookup_hash{@whatever} = (1) x @whatever;` [download] Secondly, a hash lookup is much faster than iterating through an array as the number of values increases. So, the "waste" of the values in a hash to gain lookup speed is sometimes a worthwhile tradeoff. Note that in your second example (`if (/$hash{$whatever}/)` ) you needn't do a match against the value returned from your hash; rather, merely check the key like so: `if ($hash{$whatever})`.	[reply] [d/l] [select]
Re: Creating lists using a hash with the same key/values by bart (Canon) on Jan 02, 2003 at 09:02 UTC
Often, when coding scripts, I encounter situations where I need to test to see if a variable matches an item that's stored in a list of values. `if (/$hash{$whatever}/) { do something; }` [download] Assuming I'm thinking correctly so far, my next question has to do with creating the hash value. Is there anything "wrong" with populating a hash with keys and values that are the same? For example: `%SampleHash ( 'whatever' => 'whatever', 'whoever' => 'whoever', )` [download] The idea to use a hash to check if an item can be found in a set, is a good one. What is wrong with your last approach, is that if the string evaluates to false, a simple `if($SampleHash{$string}){ ... }` [download] will fail to see it. The strings that do that, are "" and "0". I ordinarily use one of the following approaches: `foreach(LIST) { $checked{$_}++; } if($checked{$string}) { ... }` [download] or `@checked{LIST} = (); if(exists $check{$string}) { ... }` [download] but there are more. Use what seems most appropriate at the time.	[reply] [d/l] [select]
Re: Creating lists using a hash with the same key/values by gjb (Vicar) on Jan 02, 2003 at 00:22 UTC
You may want to have a look at the Set::* family of modules, especially Set::Scalar and Set::Object come to mind. Hope this helps, -gjb-	[reply]