Re: SQL Results

[Speedy]

If you use a hash technique to remove duplicates, be careful of the case. Otherwise [e-mail] will be different from [e-mail], and John john will get two copies of the email.

When saving email addresses from Web forms I generally make the stored results lower case.

[theorbtwo]

Careful here. Though in most cases, the username part of email addresses is case insensitive, I'm fairly certian that this isn't gaurnteed.

(The domainname part /is/ gaurnteed to be case insensitive, however.)

TACCTGTTTGAGTGTAACAATCATTCGCTCGGTGTATCCATCTTTG ACACAATGAATCTTTGACTCGAACAATCGTTCGGTCGCTCCGACGC

[jeffa]

Most correct about case!

Here is one way to elimate 'would-be' duplicates who simply are not because of case sensitivity:

use strict;

my %unique;
my @email = qw(
   [e-mail]
   [e-mail]
   [e-mail]
   [e-mail]
   [e-mail]
   [e-mail]
);

for (@email) {
   my $key = lc $_;
   $unique{$key} = $_ unless ($unique{$key} and $unique{$key} =~ /[A-Z
+]/);
}

print join("\n", values %unique), "\n";

Yields the list:

[e-mail]
[e-mail]
[e-mail]
[e-mail]

Of course, this is NOT perfect. Just some random thoughts if you will. ;) If you are not concerned with the extra overhead of space, i recommend storing the email address twice: once as it is, and once all lower-cased. Then you could issue a SQL statement like:

SELECT distinct lc_email, email, ...

and only use the 'email' column. Actually, even better might be performing a comparison before you insert the email address - make sure that a lower case version of the candidate email address does not already exist in the database first.

jeffa

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)