Question

I’m not getting the result expect from the following code:

#!/bash/perl
use strict;
use warnings;

my @file;
my $file2;

open (IN, "+</home/opmeitle/labs-perl/numbers");
@file = <IN>;
seek IN,0,0;

my %change = ( 80.928 => "85.950", 320.000  => "380.500");
my $changekey = join "|", keys %change;

foreach $file2 (@file){
$file2 =~ s/($changekey)/$change{$1}/g;
print IN $file2;}
close IN;

This is the contents of /home/opmeitle/labs-perl/numbers:

80.928
320.000

Here is the output:

85.950
380.500.000

Here is the result I desire:

85.950
380.500

I appreciate your answers.

luis.

but, looking this example, change in file numbers 80.928 and 320.000 for “Hola, mi nombre es Luis y vivo en Argentina” in code my %change = ( nombre => “name”, mi => “my”); this is result “Hola, my name es Luis y vivo en Argentina a ” in the end , an word the more! because?

Answer 1

Your first problem is in . (dot)
. (dot) – in regular expression means “any symbol”.

Your second problem in 320.000 — this is the number and it’s exactly equals to 320

320.000 =~ s/320/380.500/g; => 380.500.000

I suppose that solution may be in changing type of hash keys from numbers to strings
and escape all dots '.' => '\.'

To support my point of view, I wrote small script:

#!/usr/bin/perl

my %change = ( 80.928 => "85.950", 320.000  => "380.500");
my $changekey = join "|", keys %change;

print $changekey;

And voila, it produces output: 320|80.928

But when %change hash is written as my %change = ( "80.928" => "85.950", "320.000" => "380.500");, output would be 320.000|80.928

And the third problem, you open file in read-write mode, but when resulting file has less size then source, at the end would be rubbish. To avoid this, you must use truncate, or open file in readonly mode, read it, close, and after that open it in writeonly mode.