Bolt on where a match is not found to a print script

[Anonymous Monk]

The script works very well to match and print each time the number exists on a line in one or more of the files. My question is that I would like to be able to identify when there is no match at all for that number in any of the files (i.e. id didn't find the match in any of the files and im letting you know). To clarify, I still need to match and print every time the number is matched in every file, not just if it was matched once. also the line output where it matches is critical to be printed. ultimately this is just to show if the number was not matched anywhere in any of the files.

 #!/perl/bin/perl

use strict;
use warnings;

my @files = <c:/perl64/myfiles/*>;

foreach my $file ( @files ) {

    open my $file_h, '<', $file or die "Can't open $file: $!";

    while ( <$file_h> ) {

        print "$file $_" if /\b1203\b/;
        print "$file $_" if /\b1204\b/;
        print "$file $_" if /\b1207\b/;
    }
}

[hippo]

Keep a tally:

#!/perl/bin/perl

use strict;
use warnings;

my @files = <c:/perl64/myfiles/*>;
my $total = 0;

foreach my $file ( @files ) {

    open my $file_h, '<', $file or die "Can't open $file: $!";

    while ( <$file_h> ) {

        if (/\b120[347]\b/) {

            print "$file $_";
            $total++;
        }
    }
}
print "No files matched.\n" unless $total;

(Edited for typos)

[QM]

I think they wanted to check for each number separately. But this is most of the way there -- I'd just add a hash for the match, something like $matched{$1} = 1, and check that at the end.

[hippo]

Hmmm. Yes, you could well be right about each one requiring a separate flag, although that's not how I originally read the requirement. At least our anonymous brother has 2 marginally differing approaches to choose between. The whole thing does have something of an XY ring to it. Perhaps we'll find out.

[QM]

Something like this (untested):

#!/usr/bin/env perl

use strict;
use warnings;

my @files = <c:/perl64/myfiles/*>;

# record matching regexes
our %matched;

my @nums = ('1203', '1204', '1207');
my $regex = '\b(' . join('|', @nums) . ')\b'; # update: fixed + to .

for my $file ( @files ) {

    open my $file_h, '<', $file or die "Can't open $file: $!";

    while ( <$file_h> ) {
        if (my ($match) = m/$regex/) {
            $matched{$match} = 1;
            print "$file $_";
        }
    }
}

# Check all nums have been seen
for my $num (@nums) {
    if (not exists($matched{$num})) {
        print "$num not found\n";
    }
}

I would probably remove the @files and pass them on the command line (changing the for loop to while (<>)).

You might also do something more generic around the numbers, and allow those to be passed on the command line as well -- but this will take some command line parsing, or one of the GetOpt modules.

Update: Fixed capture problem, as noted by AnomalousMonk.

[AnomalousMonk]

my @nums = ('1203', '1204', '1207');
my $regex = '\b(?:' + join('|', @nums) + ')\b';

No way is provided for  if (my ($match) = m/$regex/) { ... } to capture anything (and thus be true). Also, only a single match per line is assumed. (Also  + (addition) is used where  . (concatenation) is intended.)

I would suggest something along the lines of

c:\@Work\Perl\monks>perl -wMstrict -MData::Dump -le
"my @nums = ('1203', '1204', '1207');
 my $regex = '\b(?:' . join('|', @nums) . ')\b';
 print qq{'$regex'};
 ;;
 my @matches = 'w 1207 x 1203 y 9999 z' =~ m{ $regex }xmsg;
 dd \@matches;
"
'\b(?:1203|1204|1207)\b'
[1207, 1203]

with the loop being (also untested):

while ( <$file_h> ) {
  if (my @matches = m/$regex/g) {
    ++$matched{$_} for @matches;
    print "$file $_";
  }
}

Update: Changed code example to include 9999 group.

Update 2: Shorter, IMHO sweeter:

c:\@Work\Perl\monks>perl -wMstrict -MData::Dump -le
"my @nums = qw(1203 1204 1207 1111);
 my ($regex) = map qr{ \b (?: $_) \b }xms, join ' | ', @nums;
 print $regex;
 ;;
 my %seen;
 for ('w 1207 x 1203 y 9999 z',
      'w 11203 x 12033 y 112033 z',
      'w 1207 x 1203 y 9999 z 1207 zz',
     ) {
   print qq{>$_<} if map ++$seen{$_}, m{ $regex }xmsg;
   }
 dd \%seen;
 ;;
 my $not_seen = join ' ', grep !$seen{$_}, @nums;
 print 'num(s) not seen: ', $not_seen || '(none)';
"
(?msx-i: \b (?: 1203 | 1204 | 1207 | 1111) \b )
>w 1207 x 1203 y 9999 z<
>w 1207 x 1203 y 9999 z 1207 zz<
{ 1203 => 2, 1207 => 3 }

num(s) not seen: 1204 1111

[QM]

Ah, yes, I started down one path, then changed tacks.

I believe if you change this line:

my $regex = '\b(?:' . join('|', @nums) . ')\b'; # update: fixed concat
+enation issue

to this:

my $regex = '\b(' . join('|', @nums) . ')\b'; # update: fixed concaten
+ation issue

it does the job.

For instance, from the debugger (which, note, doesn't play well with my):

DB<1> $_ = "oh blah dee"
DB<2> ($x) = m/(blah)/
DB<3> x $x
0  'blah'

Update: Fixed concatenation issue

[shmem]

This is how I would do it:

# *perl*
use strict; use warnings;
use Getopt::Std;

our ($opt_n, $opt_g);
getopt('n');

if (! $opt_n || ! $opt_g) {
    print "usage: $0 -n numberlist -g globpattern\n"
        . "  where numberlist are numbers separated by commas,\n"
        . "  e.g. -n 1203,0815,22222\n";
    exit;
}
my %matches;
$matches{$_} = 0 for split /,/, $opt_n;
$opt_n =~ s/,/|/g;
my $re = qr{\b($opt_n)\b};

@ARGV = <$opt_g>; # set @ARGV to found files 

while(<>) {
    my @m;
    @m = $_ =~ /$re/g and @matches{@m} = (1) x @m and print "$ARGV $_"
+;
}

my @nomatch = grep { ! $matches{$_} } keys %matches;
if(@nomatch) {
    print "$_ not matched\n" for @nomatch;
}

C:\some\path>perl match.pl -n 1203,1204,1207 -g c:/perl64/myfiles/*

Of course, TIMTOWTDI (there is more than one way to do it)