Regexp to Compare partial filenames then moving to another directory perl - file

Question

I am working on a script to compare non-running files within a dir to running files from a command. I have to use Regex to strip the front half of the filenames from the dir then regex to strip the filenames out of a command which then records the unmatched names into an array.
The part I cannot figure out is how I can move the filenames from the old dir into a new directory for future deletion.
In order to move the files I will need to enclose them in wildcards, * due to the random numbers in front of the filenames and the extention.
example filenames before and after:
within dir:
13209811124300209156562070_cake_872_trucks.rts
within command:
{"file 872","cake_872_trucks.rts",running}
in #events array:
cake_872_trucks
My code:
#!/usr/bin/perl -w
use strict;
use warnings;
use File::Copy qw(move);
use Data::Dumper;
use List::Util 'max';
my $orig_dir = "/var/user/data/";
my $dest_dir = "/var/user/data/DeleteMe/";
my $dir = "/var/user/data";
opendir(DIR, $dir) or die "Could not open $dir: $!\n";
my #allfiles = readdir DIR;
close DIR;
my %files;
foreach my $allfiles(#allfiles) {
$allfiles =~ m/^(13{2}638752056463{2}635181_|1[0-9]{22}_|1[0-9]{23}_|1[0-9]{24}_|1[0-9]{25}_)([0-9a-z]{4}_8[0-9a-z]{2}_[0-9a-z]{2}[a-z][0-9a-z]0[0-9]\.rts|[a-z][0-9a-z]{3}_[0-9a-z]{4}_8[0-9a-z]{2}_[0-9a-z]{2}[a-z]{2}0[0-9]\.rts|[a-z]{2}[0-9a-z][0-9]\N[0-9a-z]\N[0-9]\N[0-9]\N[0-9a-z]{4}\N[0-9]\.rts|[a-z]{2}[0-9a-z]{2}\N{2}[0-9a-z]{2}\N{2}[0-9][0-9a-z]{2}\N[0-9]{2}\.rts|S0{2}2_86F_JATD_01ZF\.rts)$/im;
$files{$2} = [$1];
}
my #stripfiles = keys %files;
my $cmd = "*****";
my #runEvents = `$cmd`;
chomp #runEvents;
foreach my $running(#runEvents) {
$running =~ s/^\{"blah 8[0-9a-z]{2}","(?<field2>CBE1_D{3}1_8EC_J6TG0{2}\.rts|[0-9a-z]{4}_8[0-9a-z]{2}_[0-9a-z]{2}[a-z][0-9a-z]0[0-9]\.rts|[a-z]{2}[0-9a-z]{2}\N{2}[0-9a-z]{2}\N{2}[0-9][0-9a-z]{2}\N[0-9]{2}\.rts)(?:",\{239,20,93,5\},310{2},20{3},run{2}ing\}|",\{239,20,93,5\},310{2},[0-9]{2}0{3},run{2}ing\}|",\{239,20,93,5\},310{2},[0-9]{3}0{4},run{2}ing\}|",\{239,20,93,5\},3[0-9]0{2},[0-9]{2}0{4},run{2}ing\})$/$+{field2}/img;
}
my #events = grep {my $x = $_; not grep {$x =~/\Q$_/i}#runEvents}#stripfiles;
foreach my $name (#events) {
my ($randnum, $fnames) = { $files{$name}};
my $combined = $randnum . $fnames;
print "Move $file from $orig_dir to $dest_dir";
move ("$orig_dir/$files{$name}", $dest_dir)
or warn "Can't move $file: $!";
}
#print scalar(grep $_, #stripfiles), "\n";
#returned 1626
#print scalar(grep $_, #runEvents), "\n";
#returned 102
#print scalar(grep $_, #allfiles), "\n";
#returned 1906

Answer 1

Once you are parsing filenames with regex there is no reason not to be able to capture all parts so that you can later reconstitute needed parts of the filename.
I assume that that overly long (and incomplete) regex does what it is meant to.
I am not sure how the files to move relate to the original files in #allfiles, since those are fetched from /var/user/data while your moving attempt uses /home/user/RunBackup. So code snippets below are more generic.
If what gets moved are precisely the files from #allfiles then just keep the file name
my %files;
foreach my $oldfile (#allfiles) {
$oldfile =~ m/...(...).../; # your regex, but capture the name
$files{$1} = $oldfile;
}
where by /...(...).../ I mean to indicate that you use your regex, but to which you add parenthesis around the part of the pattern that matches the name itself.
Then you can later retrieve the filename from the "name" of interest (cake_872_trucks).
If, however, the filename components may be needed to patch a different (while related) filename then capture and store the individual components
my %files;
foreach my $oldfile (#allfiles) {
$oldfile =~ m/(...)(...)(...)/; # your regex, just with capture groups
$files{$2} = [$1, $3]; # add to %files: name => [number, ext]
}
The regex only matches (why change names in #allfiles with s///?), and captures.
The first set of parenthesis captures that long leading factor (number) into $1, the second one gets the name (cake_872_trucks) into $2, and the third one has the extension, in $3.
So you end up with a hash with keys that are names of interest, with their values being arrayrefs with all other needed components of the filename. Please adjust as needed as I don't know what that regex does and may have missed some parts.
Now once you go through #events you can rebuild the name
use File::Copy qw(move);
foreach my $name (#events) {
my ($num, $ext) = #{ $files{$name} };
my $file = $num . $name . $ext;
say "Move $file from $orig_dir to $dest_dir";
move("$orig_dir/$file", $dest_dir) or warn "Can't move $file: $!";
}
But if the files to move are indeed from #allfiles (as would be the case in this example) then use the first version above to store filenames as values in %files and now retrieve them
foreach my $name (#events) {
move ("$orig_dir/$files{$name}", $dest_dir)
or warn "Can't move $file: $!";
}
I use the core module File::Copy, instead of going out to the system for the move command.
You can also rebuild the name by going through the directory again, now with names of interest on hand. But that'd be very expensive since you have to try to match every name in #events for every file read in the directory (O(mn) complexity).
What you asked about can be accomplished with glob (and note File::Glob's version)
my #files = glob "$dir/*${name}*";
but you'd have to do this for every $name -- a huge and needless waste of resources.
If that regex really must spell out specific numbers, here is a way to organize it for easier digestion (and debugging!): break it into reasonable parts, with a separate variable for each.
Ideally each part of alternation would be one variable
my $p1 = qr/.../;
my $p2 = qr/.../;
...
my $re_alt = join '|', $p1, $p2, ...;
my $re_other = qr/.../;
$var =~ m/^($re_alt)($re_other)(.*)$/; # adjust anchors, captures, etc
where the qr operator builds a regex pattern.
Adjust those capturing parenthesis, anchors, etc to your actual needs. Breaking it up so that the regex is sensibly split into variables will go a long way for readability, and thus correctness.
Assuming that there is a good reason to seek those specific numbers in filenames, this is also a good way to document any such fixed factors.

Answer 2

I guess you need something like this:
my $path = '/home/user/RunBackup/';
my #files = map {$path."*$_*"} #events;
system(join " ", "mv", #files, "/home/user/RunBackup/files/");
If there are lots of files you might need to move them one by one:
system(join " ", "mv", $_, "/home/user/RunBackup/files/") for #files;