It uses a hash of arrays.

#!/bin/perl5

use strict;
use warnings;
use Data::Dumper;

my %data;
my $file;

while (my $line = <DATA>){
  if ($line =~ /([^.]+\.csv)/){
    $file = $1;
    $data{$file} = [];
  }
  elsif ($line =~ /(-?[\d.]+)/){
    push @{$data{$file}}, $1;
  }
}

for my $file (keys %data){
  print "file: $file\n";
  # open $file
  print "data:\n";
  print "\t$_\n" for @{$data{$file}};
  print "\n";
  # print data to file
}

#print Dumper \%data;

__DATA__
filename =

without012.csv


pred =

   -0.0105
    0.0645
    0.1989
    0.0151
    0.0289
    0.0400
    0.5313


filename =

without013.csv


pred =

   -0.0080
    0.0663
    0.1911
    0.0075
    0.0288
    0.0302
    0.5234


filename =

without014.csv


pred =

   -0.0023
    0.0595
    0.2582
    0.0323
    0.0285
    0.0578
    0.5807
[download]

output:

---------- Capture Output ----------
> "C:\Perl\bin\perl.exe" _new.pl
file: without012.csv
data:
    -0.0105
    0.0645
    0.1989
    0.0151
    0.0289
    0.0400
    0.5313

file: without013.csv
data:
    -0.0080
    0.0663
    0.1911
    0.0075
    0.0288
    0.0302
    0.5234

file: without014.csv
data:
    -0.0023
    0.0595
    0.2582
    0.0323
    0.0285
    0.0578
    0.5807


> Terminated with exit code 0.
[download]

Hope this helps

Actually with a few small modifications your starter code as shown in the OP is able to do what you expected from it (showing you the file names):

#!/usr/bin/perl

$/ = '';   # read input by paragraphs

@file = <>;

for($i = 1; $i < @file; $i+=4)  # from field 1 (2nd field); and += for
+ proper loop
{

    print "$file[$i]";
}
[download]

Edit: Oops, it's $/ = '';  # not undef

perl -ne 'if(/(\w+\.csv)/){close $fh if defined $fh;open $fh,">$1"} el
+sif(/([0-9.-]+)/){print $fh "$1\n"}' file.txt
[download]

open FILE, "<file.txt";           # Open the file
while( <FILE> ) {                 # read a line
    if ( /(\w+\.csv)/ ) {         # If the line is a filename...
        close $fh if defined $fh; # close previous
        open $fh,">$1";           # and open new for writing
    } elsif ( /([0-9.-]+)/ ) {    # If the line is a number...
        print $fh "$1\n";         # write it
    }
}
close $fh;                        # close the files opened
close FILE;
[download]

I would be more inclined to check for the result of open even in an one liner. If I want to be lazy, I'd leave out the close instead, because perl takes care of those.

Hm, I'd separate these into "chunks" that are each a file, and worry those apart later. Something like this:

open INFILE, '<', 'infile.txt' or die("Can't read: $!");

# each 'line' is a chunk with 'filename =' delimiting.
local $/ = 'filename =';

while (<INFILE>) {
   s/^s\+//s; # remove starting whitespace.
   
   # use a regex capture -- slow but simple.
   my ($filename, $data) = ($1, $2) if m/^(\w+)\s+pred=\s+(.*?)\s+$/s;

   open my $OUT, '>', $filname or die("Can't write $filename: $!");
   foreach ( split(/\s+|\n/, $data) ) {
      next unless m/^[-.0-9][0-9.]+/; #skip any empty or malformed lin
+es
      print $OUT $_;
   }
}
[download]

Untested, but the principle is sound and somewhat reusable.

The program you might be looking for might be this :

perl -ne '/(\S+\.csv)/ and open X, ">$1" and next; next if /=/ or /^\s
+*$/; s/^\s+//; print X' inputfile
[download]

Of course, you don't want to do it like that :)

The key things to keep in mind are triggering an output destination swap when you see a filename, and triggering an output print when you see output you intend to store. Nothing else matters.

You don't need an array of files. You don't even need to hold it in memory for any operations.

If the same filename could come up more than once, use ">>" instead of ">" to open the files.

First thing to notice is you forgot to initially open the original text file. So whether or not the count of line numbers is right, there is no data to work with.

Cheers, Sören

He's using <> so the original text file is implicitly opened if given on the cmd line. I guess you already knew, but you may have just overlooked it.

To be sure, just take his code and modify it like thus:

#!/usr/bin/perl -l

my @file = <>;
print 0+@file;
[download]

Indeed

$ ./bar.pl data.txt 
46
$ wc data.txt 
 46  36 346 data.txt
[download]