Regex renaming

ArifS has asked for the wisdom of the Perl Monks concerning the following question:

Here is the code I am working with-

while (my $file = <DATA>) {
    $file =~ /(\d+)(lc|pp)/i or next;
    my $posA = uc "$1$2-";
#    $file =~ /(sh|lic)/i or next;
#    my $posB = uc $1;
    $file =~ /(800|2621|3524|6509|7206)(s-|r-)/i or next;
    my $dv = uc "$1$2";
    $file =~ /([a-z]+)(\d+)/i or next;
    my $loc = uc "$1$2";
    print "$posA" or die;
#    print "$posB" or die;
    print "$dv" or die;
    print "$loc" or die;
    print "\n";
}

__DATA__
1lc-2621s-RT02-9FL
2pp-3524-7FL
PP_800c_-RT1-TX
10lc-7206-3FL-SW02
LC-6509r-SW1
[download]

I am trying to get the following-

01LC-2621S-RT02-09FL-STATE
02PP-3524#-DEVICE#-07FL-STATE
#PP-800C-RT01-#FL-TX
10LC-7206#-SW02-03FL-NY
#LC-6509R-DEVICE#-#FL-GA

Device# if missing RT#, FW#, or SW# etc. STATE if missing 2 digits state info like, NY, FL etc. #FL, if missing floor#.

Please let me know

Comment on Regex renaming Download Code

Replies are listed 'Best First'.

Re: Regex renaming
by GotToBTru (Prior) on Oct 09, 2014 at 19:51 UTC

It's unclear what you are trying to do. In what case should STATE be appended? What are the rules for handling the initial digit? When should # be inserted?

1 Peter 4:10

[reply]

Re^2: Regex renaming

by ArifS (Beadle) on Oct 10, 2014 at 13:40 UTC

Thank you for your suggestion. I have updated what I am looking for in the original post.

[reply]

Re: Regex renaming
by AnomalousMonk (Archbishop) on Oct 09, 2014 at 20:34 UTC

Further to GotToBTru's comments on the obscurity of the string transformations you're trying to achieve, one more thing puzzles me. There are several statements of the form
print "$part_of_final_string" or next;
that will output an incompletely transformed string and then go on to process the next input string if the print I/O operation fails. In this unlikely (and probably catastrophic) event, don't you really want to die with an appropriate error message rather than trying to press on? Just curious...

[reply]
[d/l]

Re^2: Regex renaming

by ArifS (Beadle) on Oct 10, 2014 at 13:41 UTC

That's a good suggestion. I have updated as you suggested.

[reply]

Re: Regex renaming
by Athanasius (Archbishop) on Oct 11, 2014 at 07:30 UTC

Hello ArifS,

It looks like you data lines are composed of fields separated by hyphens. So I would begin with this kind of structure:

while (my $file = <DATA>)
{
    chomp $file;
    my @old = split /-/, $file;
    my @new;

    for my $old_field (@old)
    {
        # analyse   $old_field
        # construct $new_field
        push @new,  $new_field;
    }

    print join('-', @new), "\n";
}
[download]

This is conceptual only. The mechanics are complicated by the requirement that fields do not appear in a fixed order within the data. For example, in the line:

1lc-2621s-RT02-9FL
[download]

the “device” field RT02 preceeds the “floor” field 9FL, whereas in the line:

10lc-7206-3FL-SW02
[download]

the order of these two fields is reversed. To handle this, you need a more flexible implementation. Here is one such approach: