use strict;
use warnings;

# Create a test file
open outFile, ">", "file1.txt";
print outFile <<end;
rabbit 1
hen    0
pig    2
cow    2
sheep  5
end
close outFile;

#Sample code
open inFile, "<", "file1.txt" or die "Unable to read file1.txt. $!";
my @lines = <inFile>;
close inFile;

my %animals;

# build a hash keyed by animal with the count as value
map {/(\w+)\s+(\d+)/; $animals{$1}=$2} @lines;

# Production code would use file2 rather than DATA
while (<DATA>)
  {
  /(\w+)/; #Assume animal is always the first thing on the line

  # Bump animal count - creates new entry if new animal
  ++$animals{$1} if defined $1;
  }

open outFile, ">", "file1.txt";
for (@lines)
  {
  /(\w+)/;
  printf outFile "%-6s %d\n", $1, $animals{$1};
  $animals{$1} = 0; # Tag as printed
  }

#add any new animals to the end of the file
for (keys %animals)
  {
  next if ! $animals{$_};
  printf outFile "%-6s %d\n", $_, $animals{$_};
  }

close outFile;

__DATA__
rabbit feed 10kg used on 21/10/2005 by consumer1
cow feed 100kg used on 11/11/2005 by consumer1
sheep feed 50kg used on 24/11/2005 by consumer3
[download]

Creates a file containing:

rabbit 2
hen    0
pig    2
cow    3
sheep  6
[download]

This works a treat , thanks.
I see from the multiple answers that there are many ways of approaching this problem. I guess that's the beauty of Perl ?

This is my bet at it:

use warnings;
use strict;

my %fed;
open my $fd, '<', 'file2' or die "open: $!";
$fed{$1}++ while <$fd> =~ m/^(\w+)\s/;
close $fd;

open $fd, '<', 'file1' or die "open: $!";
open my $fd2, '>', '/tmp/file1.tmp' or die "open: $!";  ## FIXME: tmpf
+ile
while (<$fd>) {
  m/^(\w+)\s+(\d+)$/;
  printf $fd2 "$1\t%d\n", exists $fed{$1} ? $2+1 : $2;
}
close $fd2; close $fd;
rename '/tmp/file1.tmp', 'file1' or warn "rename: $!";
[download]

I've tried editing file1 inplace using $^I and <> but didn't get anything.

You can doi it like this:

use warnings;
use strict;

# Create a test file
open outFile, ">", "file1.txt";
print outFile <<end;
rabbit 1
hen    0
pig    2
cow    2
sheep  5
end
close outFile;

my %fed;
$fed{$1}++ while <DATA> =~ m/^(\w+)\s/;

local @ARGV = ('file1.txt');
local $^I = '.bak';
while (<>) {
  m/^(\w+)\s+(\d+)$/;
  printf "$1\t%d\n", exists $fed{$1} ? $2+1 : $2;
}

__DATA__
rabbit feed 10kg used on 21/10/2005 by consumer1
cow feed 100kg used on 11/11/2005 by consumer1
sheep feed 50kg used on 24/11/2005 by consumer3
[download]

file1.txt contains:

rabbit    2
hen    0
pig    2
cow    3
sheep    6
[download]

---

DWIM is Perl's answer to Gödel

My error was:

local $^I = '.bak';
while (<>) {
  m/^(\w+)\s+(\d+)$/;
  $2++;                ## this
[download]

I guess that would work if I used Tie::File or something like that. Thank you!

--
David Serrano

If I wanted to change the logic of the above code so all values not in $fed{1} were incremented by 1 instead , how would I do this ?
I tried doing this

local @ARGV = ('file1.txt');
local $^I = '.bak';
while (<>) {
  m/^(\w+)\s+(\d+)$/;
  printf "$1\t%d\n", not exists $fed{$1} ? $2+1 : 0;
}
[download]

This only works the first time it's run, after that everything that isn't in $fed{1} is always only set to 1.
I was trying to get it to increment all other values by 1 each time and set the $fed{1} values to 0

If I wanted to change the logic of the above code so all values not in $fed{1} were incremented by 1 instead , how would I do this ?
I tried doing this


local @ARGV = ('file1.txt');
local $^I = '.bak';
while (<>) {
  m/^(\w+)\s+(\d+)$/;
  printf "$1\t%d\n", not exists $fed{$1} ? $2+1 : 0;
}
[download]

This only works the first time it's run, after that everything that isn't in $fed{1} is always only set to 1. I was trying to get it to increment all other values by 1 each time and set the $fed{1} values to 0
I think my logic is screwed up somewhere ?

A simple approach :

use strict ;
use warnings ; 

my %initial = () ;
my %seen = () ;  # increment if animal is seen 
my $new_value ; 



# Load initial values
open FILE, '<', 'file1.txt' or die 'Failed to open file1' ; 
while ( my $line = <FILE> ) {     
    my ($animal, $count ) = split " ", $line ;
    $initial{$animal}=$count ;
    }
close FILE ;


# Open file to look for new animals
open UPDATE , '<', 'file2.txt' or die 'Failed to open file2' ; 
while ( my $line = <UPDATE> ) {
    # We assume a special pattern form your example file
    my $animal = $1 if $line =~ m/^(\w+) feed/ ;
    $seen{$animal}++ ;
    }
close UPDATE ;


# Open result file for writing
open FILE, '>','file1.txt' or die 'Failed to open file1 for writing';

foreach ( sort keys %initial) {

#Deal with the three possible case : 
#new animal, previously existing but unseen, or existing and seen

    if ( defined $seen{$_} and not defined $initial{$_} ) { 
    
            $new_value = 1
    
    } elsif (  not defined $seen{$_} and defined $initial{$_} ) { 
    
            $new_value = $initial{$_} 
    
    } else { 
            $new_value = $initial{$_} + 1  
    
    } 
    
    print FILE "$_ $new_value\n" ;
    
    
}

close FILE ;
[download]