getting specific filename from bash - mysql

Question

So I have a perl module that uses a bash command to obtain the file(s) with certain "table" names. In my specific case, it is looking for tables with the name "event", but I need this to work with all names too.
Currently, I have the following code in my perl script to obtain MYI files with the name table, and I am receiving not only event_* but also event_extra_data_* as well. For my example, I only need the 2nd table that exists in my database for event_. As my test info, I have, currently,
event_1459161160_0
event_1459182760_0
event_extra_data_1459182745_0
event_extra_data_1459182760_0
which are partitioned tables from tables "event" and event_extra_data which is the value that the $table variable sees below.
Anyways, my question is, how do i limit this to only receiving event_1459182760_0.MYI and not event_extra_data_1459182760_0.MYI which it is currently getting?
elsif ($sql =~ /\{LAST\}/i )
{
$cmd = 'ls -1 /var/lib/mysql/sfsnort/'.$table.'_*MYI | grep -v template | tail -n1 | cut -d"/" -f6 | cut -d"." -f1';
$value = `$cmd`;
print "Search Value: $value\n";
if ($value eq "")
{
$sql = ""; # same as with FIRST
}
else
{
$sql =~ s/\{LAST\}/$value/g;
}
}

Answer 1

Don't parse ls - there's no point, and it's prone to causing problems.
I would point out this - the glob function within perl allows you to do to a limited number of "regex-like" patterns. (Note - they aren't regex, so don't get them mixed up).
foreach my $filename ( glob "event_[0-9]*" ) {
#do something with $filename
}
If you're just after the last - when sorted numerically:
my ( $last ) = reverse sort glob "event_[0-9]*";
Given you have a single path, then you should be able to:
my ( $last ) = reverse sort glob "/var/lib/mysql/sfsnort/event_[0-9]*.MYI";
Note - that this works, assuming you're working with time() numeric values - it's doing an alphanumeric sort (and on directory names too).
If that isn't a valid assumption, you'll need a custom sort - which is quite easy, you can feed sort a subroutine to sort by.
Either:
sort { my ($a1) = $a =~ /(\d+)/; my ($b1) = $b =~ /(\d+)/; $b1 <=> $a1 }
To extract the first 'string of digits' from the path. (note - also includes directories).
Or use the -M file test:
sort { -M $a <=> -M $b }
Which will read modification time from the file (technically -M is age in days).
You can remove the reverse if you custom sort, just by swapping $a and $b.

Answer 2

Though I think this would be better done all in perl, to answer your specific question about how to get event_* but not event_extra*, you could of course add that to your grep to filter out, or you could use a different glob, like $table_[0-9]* if there's always an _ then a digit after the table name.
In perl you could do it something like the following though:
opendir( DIR, '/var/lib/mysql/sfsnort/' );
my #files = sort grep { /$table_\d/ } readdir( DIR );
closedir( DIR );
$files[$#files] =~ /(^[^.]+)/;
my $value = $1;