well, i guess you could make a small subroutine that returns an 11-line block out of the file. not taking into account any bad records, here's some pretty straight-forward code:
sub returnrec {
my ($i,$record);
for ($i = 0; $i < 11; $i++) {
$record .= <FILEHANDLE>;
}
return $record;
}
then you can say something like:
my @elements = split /\;/, returnrec();
--twerq | [reply]
[d/l]
[select] |
my $recordSet = 11;
open (FILE, $file) or die $!;
my @list = <FILE>;
close (FILE);
my @datasets = ();
while (my @subList = splice( @list, 0, $recordSet) ){
push (@datasets, join("\n", @subList);
}
undef @list;
So, now you've got your datasets in the list @datasets...
| [reply]
[d/l] |
I think this is the best answer in the thread, but the join is a bad idea. You've already got newlines in there, so, so later someone will need to come along and do a "split /\n\n/, @datasets;". Assuming that the newlines are not needed, I would chomp them and use array references.
my $recordSet = 11;
open FILE, "< $file" or die "Cannot open $file for reading: $!";
chomp ( my @list = <FILE> );
close FILE;
my @datasets = ();
while (my @subList = splice( @list, 0, $recordSet) ){
push @datasets, \@subList;
}
With this, you now have an array or array refs, each containing precisely one record and no worries about adding extra newlines. Remember to always use Perl's data structures when possible (and reasonable, of course).
Cheers,
Ovid
Join the Perlmonks Setiathome Group or just click on the the link and check out our stats. | [reply]
[d/l] |
or just take it into an LDIF-like datastructure:
my $recordSet = 11;
open FILE, "< $file"
or die "Cannot open $file for reading: $!";
chomp ( my @list = <FILE> );
close FILE;
my @datasets = ();
while (my @subList = splice( @list, 0, $recordSet) ){
my %hash = ();
foreach (@subList){
my ($key, $value) = split(/\s*\:\s*/, $_, 2);
if (exists $hash{$key} ){
push (@{ $hash{$key} }, $value);
}
else {
$hash{$key} = $value;
}
push @datasets, \%hash;
# if unique key exists, better use a Hash of Hashes
# instead of array of hashes
}
}
}
| [reply]
[d/l] |
$/=undef; # Slurps all data in file.
$_=<DATA>; # Put it in $_, convenient for regex.
my $lines=0; # Line count $. might suffice and then the pre increme
+nt can be dropped.
my $records=11; # Number of lines in a record.
s/\n/(',',"\n")[++$line%$records==0]/eg;# replaces all \n with commas
+except the recordth.
print $_;
--
Brother Frankus.
¤ | [reply]
[d/l]
[select] |
#!/usr/bin/perl
use strict;
use warnings;
use constant RECORD_SIZE => 11;
my (@record, @lines);
open FOO, 'test.plx' or die;
while (@lines = grep defined, map scalar <FOO>, 1 .. RECORD_SIZE) {
if (@lines == 11) {
chomp @lines;
# flatten into a single array ...
@record = map split(':'), @lines;
# ... or build an array of arrays
# @record = map [ split ':' ], @lines;
}
else {
# what to do if line count is not a multiple of RECORD_SIZE?
}
}
| [reply]
[d/l] |
Without fully understanding your input data, simply setting $/ (which, in English, literally means INPUT_RECORD_SEPARATOR) to your separator (:?) could do exactly what you want.
my @data;
{
local $/ = ":"; # or "\n:\n" for : on an empty line
@data = <DATA>;
}
This would put each "record" (set of lines) into individual elements of @data. It'd be up to you then to split on newlines or whatever, and to eliminate the trailing : on most records. | [reply]
[d/l]
[select] |