How do I count keys in a BerkeleyDB *quickly* (using DB_File)?

[Damon Hastings]

I have a berkeley db which I'm accessing via the DB_File interface (a
btree). I would like a quick way to count how many keys are in the
file. If I just use "scalar keys %hash", then it takes forever! (The
file has millions of keys.) Can anyone help me?

 

[Sam Holden]

Damon Hastings ([email protected]) wrote on MMMDCLVIII September
MCMXCIII in <URL: I have a berkeley db which I'm accessing via the DB_File interface (a
btree). I would like a quick way to count how many keys are in the
file. If I just use "scalar keys %hash", then it takes forever! (The
file has millions of keys.) Can anyone help me?


'scalar keys %hash' is the way to go. The only way to get the number
of keys from a db hash is to count them.

Of course if the hash is only used by you script (or scripts) then
you could add a 'count' key to the db hash, and keep it up to date
when you modify the hash.

See sig below for the big drawback of that approach...

 

[Lack Mr G M]

|>
|> I have a berkeley db which I'm accessing via the DB_File interface (a
|> btree). I would like a quick way to count how many keys are in the
|> file. If I just use "scalar keys %hash", then it takes forever! (The
|> file has millions of keys.) Can anyone help me?

my $n = 0;
while (my ($dnc1, dnc2) = each %DB_hash) {$n++};

might be quicker - but you'll have to test it...