CODEWITHSUNDEEP
perl
user1336997
user1336997

Reputation: 105

prevent over writing data in hash

I have data.txt with data as shown below. I have around hundred names with 3 constant ids 100, 200, 300 and each have 32 data values.

NAME: xyz
ID: 100
DATA: 10 15 99 13 ...
ID: 200
DATA: 23 45 78 90..
ID: 300
DATA: 45 67 89 56
NAME: abc
ID: 100
DATA: 2 4 787 8..
ID: 200
DATA: 12 14 17..
ID: 300
DATA: 45 34 22..

I need to write this data to another file which looks like

xyz_100, xyz_200,xyz_300,abc_100,...
10     , 23     , 45     ,2
15     , 45     ,67       ,4

I build a hash to store the values, but right now my code overwrites the first two entries and store the last entry. How can I save the first two entries too, please let me know if I can simplify the code.

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

my @avar_names;
my %record;

local $/ = '';
open my $fh, '<', 'datalog.dat' or die "failed: $!";
while (<$fh>) {
  chomp;
   my %avar;
   while (/^([A-Z:]+):\s*(.*)/mg) {
      $avar{$1} = $2;
   }
   my $avar_name = "$avar{NAME}_$avar{ID}";

   push @avar_names, $avar_name;
   $record{$avar_name} = $avar{DATA};

use Data::Dumper;
print Dumper \%record;
}

Upvotes: 2

Views: 94

Answers (2)

D&#233;j&#224; vu
D&#233;j&#224; vu

Reputation: 28840

There is a consistency problem: the local line makes the first while process the whole NAME/ID data as a whole (then, thanks to the while(//mg) loop), but the code as it is expects a line by line processing.

I suggest simplicity: process the file lines one by one, for each one, identify a "NAME/ID" couple (ID could be DATA) thanks to a regex on the current line.

In order for that to work, %avar needs not to be local to the while on <$fh>, since it has to keep track of the last NAME/ID reference it got from the last lines of the file.

The @avar_names array is not used.

Finally, you want to print the resulting records at the end (I supposed), but that is your choice.

This is your program from which I tried to change only the things mentioned above:

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

my @avar_names;
my %record;

#local $/ = ''; ## commented out
open my $fh, '<', 'datalog.dat' or die "failed: $!";

my %avar; ## moved here

while (<$fh>) {
   chomp;
   if (m/^([A-Z]+):\s*(.*)/) {  ## if, not a while
      $avar{$1} = $2;

      ## ensure we have at least our 3 keys, process only if we met DATA
      if (exists($avar{NAME}) && exists($avar{ID}) && exists($avar{DATA}) && $1 eq 'DATA') { 
         my $avar_name = "$avar{NAME}_$avar{ID}";

         push @avar_names, $avar_name;  ## not used
         $record{$avar_name} = $avar{DATA};
      }
   }
}

use Data::Dumper;
print Dumper \%record;

Output:

$VAR1 = {
          'xyz_100' => '10 15 99 13 ...',
          'xyz_300' => '45 67 89 56',
          'abc_200' => '12 14 17..',
          'abc_300' => '45 34 22..',
          'abc_100' => '2 4 787 8..',
          'xyz_200' => '23 45 78 90..'
        };

Upvotes: 2

perreal
perreal

Reputation: 97958

You need something like this:

my @avars;
# ... some code
while (<$fh>) {
  # more code
  my %curr;
  while (/^([A-Z:]+):\s*(.*)/mg) {
    $curr->{$1} = $2;
  }
  if (%curr) {
     push @avars, \%curr;
     my $avar_name = "$curr{NAME}_$curr{ID}";
     push @avar_names, $avar_name;
     $record{$avar_name} = $avar{DATA};
  }
  # ....

Upvotes: 1

Related Questions