Re^2: restrict use of regex module in script

[pmpmmpmp]

Thank you for your answers. OK, so this was very simple once I tried it out on a very small piece of code instead of what I was working on. I did not realize that you could just stick {} pretty much anywhere you want. Example below, in case anyone is interested.

#!/usr/bin/perl
use strict;
use warnings;

use re::engine::TRE max_cost => 5;

my $seq1 = "TTTTACGAGAGAATATGTTAGGTGAAGGAACCTCTATCTGAGAGAAAAA";
my $seq2 = "TTTTGAGCTCGTTGTCGATCCGAGGTACTTTTGAATCCGCAGTTTCTTG";

if ("A pearl is a hard object produced ..." =~ /\(Perl\)/i)
    {
     print "$1\n";
    }


if (($seq2 =~ /GTTGTTCGATCCAGGTAC/) && ($seq1 =~ /ACGAGAGATAGATGA/))
    {
     print "Have a match\n";
    }

This will print 'pearl' and "Have a match".

#!/usr/bin/perl
use strict;
use warnings;

my $seq1 = "TTTTACGAGAGAATATGTTAGGTGAAGGAACCTCTATCTGAGAGAAAAA";
my $seq2 = "TTTTGAGCTCGTTGTCGATCCGAGGTACTTTTGAATCCGCAGTTTCTTG";

if ("A pearl is a hard object produced ..." =~ /\(Perl\)/i)
   {
    print "$1\n";
   }

 {
  use re::engine::TRE max_cost => 5;

  if (($seq2 =~ /GTTGTTCGATCCAGGTAC/) && ($seq1 =~ /ACGAGAGATAGATGA/))
      {
        print "Have a match\n";
      }
  }

This will only print out "Have a match".

[stevieb]

"I did not realize that you could just stick {} pretty much anywhere you want."

That's called a "block", and they can't quite go anywhere, but pretty close.

Blocks are how we limit the "scope" of something. Anything defined within that block is limited to that scope. This is where my and local for example come in to play. Here's a couple of examples:

use warnings;
use strict;

my $thing = 123;

{
    my $thing = 456;
}

print $thing; # 123

open my $fh, '<', 'file.txt' or die $!;

{
    local $/; # slurp in a file
    my $contents = <$fh>
    ...
}

# now we go back to reading a file line-by-line again

while (<$fh>){
   ...
}

Subroutines (functions/methods) are blocks as well, and some functions even take blocks as parameters (or operate in "block mode" so it appears to take a block as a param):

my @list = map {$_ => 1} @other_list;
my @filtered_list = grep {$_ =~ /blah/} @some_list;

...etc.

Please see AnomalousMonk's post here for how my above example is actually quite broken if taken literally. Haste took rank, but we have people who correct hasty things thankfully. I should have added an "untested" warning or better.

[AnomalousMonk]

{
    local $/; # slurp in a file
    my $contents = <$fh>
    ...
}

# now we go back to reading a file line-by-line again

while (<$fh>){
   ...
}

Parenthetic note: in the quoted example code, the  $fh file handle would become "exhausted" after the "slurp" read to  $contents and the subsequent while-loop read of the handle would produce nothing. A seek operation is needed to reset the logical file position back to the beginning of the file (update: or at least to somewhere other than the end) so that subsequent reads may be successful. Try the following code without the seek statement:

c:\@Work\Perl\monks\pmpmmpmp>perl -wMstrict -le
"open my $fh, '<', 'junk.txt' or die $!;
 ;;
 my $contents = do { local $/;  <$fh>; };
 print qq{[[$contents]]};
 ;;
 seek $fh, 0, 0;
 print 'now we go back to reading a file line-by-line again';
 while (<$fh>){
   print qq{<<$_>>};
   }
"
[[line 1 fee fie foe
line 2 wibble wobble
line 3 blah yada
]]
now we go back to reading a file line-by-line again
<<line 1 fee fie foe
>>
<<line 2 wibble wobble
>>
<<line 3 blah yada
>>

[stevieb]

Indeed. I was going to create a second handle for the latter example, but forgot :)

update: I've updated my original reply to the OP to point to your update, which technically is extremely informative./update

[BillKSmith]

When a special variable is intended to apply to only one operation, I prefer to make that explicit by limiting the scope even more.

my $contents = do{local $/; <$fh>};

[pmpmmpmp]

Thank you stevieb and Bill for your examples and explanation. They have been very helpful. ( I like the seek function ! )