# previously, I've already stripped anything following the first @, converted anything
# not \w to a -, removed repeating -, and stripped opening and trailing -
$_ = 'csdude';
# split letters and dashes from any trailing numbers
($actualName, $append) = /([a-zA-Z-]+)([0-9]*)$/;
# create hash from MySQL results
$hash = $dbh->selectall_hashref("SELECT col FROM table WHERE col REGEXP ?", 'col', undef,
# regex looks for anything beginning with $actualName, then optionally
# followed by numbers to the end
'^' . $actualName . '[0-9]*$'
);
# if $_ isn't in the hash, no need to append anything
if (exists($hash->{$_})) {
my %numList;
$append //= 1;
for $key (keys %{$hash}) {
$numList{$key =~ s#$actualName##r} = 1;
}
# $safety[0] will be the highest number in the list
@safety = sort { $b <=> $a } keys %numList;
while ($append <= $safety[0] && exists($numList{$append})) {
$append++;
}
}
$finalResult = $actualName . $append;
Is there something elsewhere in the code that sets a minimum length?
Does this mean you can't have overlapping names, like lucy24 and lucy2424 (form that actually occurs in one of my email addresses because reasons)?
perl (of which I don't speak a solitary word).
example.com/article/type-what-you-want/1234
In Apache I created a rewrite for it to /article/view.php?id=1234, and more or less just ignore the type-what-you-want text. It's pretty and search engines like it (or did, I don't know if they still do), though.
