How do I re-read a line?

rbc has asked for the wisdom of the Perl Monks concerning the following question:

Dear Monks,

I have searched high and low and cannot find the
answer to this problem I am having.

To demonstrate here's my code:

#!/usr/bin/perl -w

use strict;

my @names;
my @items;

sub processRec {
        my $fh = shift; #ref to filehandle
        while(<$fh>) {
                chomp;
                s/\cM//;
                last if /^$/;
                last if /^\d-RECORD$/;
                push ( @names, /ITEMNAME-(.*)/);
                push ( @items, /ITEM \d-(.*)/);
        }
}

open( FH, "<data");
while(<FH>) {
        if ( /\d-RECORD/ ) { processRec(*FH); }
}
close(FH);
for my $name ( @names ) { print "name = [$name]\n"; }
for my $item ( @items ) { print "item = [$item]\n"; }
[download]

This works fine most of the time.
But here is a case where it doesn't work ...

$ cat data
1-RECORD
ITEMNAME-bob
ITEM 1-socks
ITEM 2-shoes

2-RECORD
ITEMNAME-bill
ITEM 1-hat
ITEM 2-hair piece
3-RECORD
ITEMNAME-mary
ITEM 1-roach clip

4-RECORD
ITEMNAME-jimmy
ITEM 1-peanuts
ITEM 2-tooth pick

Here's the output from that case

name = bob
name = bill
name = jimmy
item = socks
item = shoes
item = hat
item = hair piece
item = peanuts
item = tooth pick

What happened to mary?

If there was a new line after the line ...
ITEM 2-hair piece
... I wouldn't have a problem.

I suspect that I need to do something like this ...

sub processRec {
...
   if (/^\d-RECORD$/) {
      put_line_back ($_);
      last;
   };
...
}

Comment on How do I re-read a line? Download Code

Replies are listed 'Best First'.
Re: How do I re-read a line? by Mr. Muskrat (Canon) on May 01, 2002 at 21:54 UTC
Hmm, how about something like this instead? `#!/usr/bin/perl -w use strict; my @names; my @items; sub processRec { my $fh = shift; #ref to filehandle while(<$fh>) { chomp; s/\cM//; next if /^$/; next if /^\d-RECORD$/; push ( @names, /ITEMNAME-(.)/); push ( @items, /ITEM \d-(.)/); } } open( FH, "<data"); processRec(*FH); close(FH); for my $name ( @names ) { print "name = [$name]\n"; } for my $item ( @items ) { print "item = [$item]\n"; }` [download] No sense in using two while loops. Matthew Musgrove Who says that programmers can't work in the Marketing Department? Or is that who says that Marketing people can't program?	[reply] [d/l]
Re: Re: How do I re-read a line? by rbc (Curate) on May 01, 2002 at 22:00 UTC
Thanks Mr. Muskrat, Your solution works. But what if I was only able to edit the sub processRec? I am sorry for not stating that in my original post and thanks for your reply! :)	[reply]
Re: How do I re-read a line? by Kanji (Parson) on May 01, 2002 at 21:58 UTC
You can use seek to 'rewind' a file handle... `while(<$fh>) { # Make a note of the line length # in case we need to rewind $fh my $line = length; chomp; # ... if ( /^\d-RECORD$/ ) { # rewind $fh to the beginning of # the current line, so <$fh> can # re-get it. seek( $fh, 0 - $line, 1 ); last; } # ... }` [download] --k.	[reply] [d/l]
(jeffa) Re: How do I re-read a line? by jeffa (Bishop) on May 01, 2002 at 22:23 UTC
At the risk of posting code that you possibly can't use, how about a completely different approach. Instead of storing the names in one array and the items in other, how about creating a data structure that is a List of Hashes (LOH). Each hash has a 'name' key that points to a scalar, and an 'item' key that points to an array reference. Now you have a usuable data structure. The idea is to store the names and the items into the hash, and when a new record is read, simply store the hash. The problem is that you have (well, i have to - this is an invitation for a more elegant solution should someone wish to show me a new trick ;)) discard the first line and store the last record outside of the while loop: use strict; use Data::Dumper; my (@record, $discard, %temp); open(FH, 'data') or die 'file not found'; # discard the first line $discard = <FH>; while(<FH>) { chomp; s/\cM//; if (/^\d-RECORD$/) { push @record, {%temp}; undef %temp; } elsif (/ITEMNAME-(.)/) { $temp{name} = $1; } elsif (/ITEM \d-(.)/) { push @{$temp{item}}, $1; } } # the last record needs to be added outside while loop push @record, {%temp}; print Dumper \@record; # print items of last record print join(', ', @{$record[-1]{item}}), "\n"; [download] If someone knows of a more elegant way, please share. I tried using File::ReadBackwards, but then extra work is needed to reverse the items ... UPDATE: Regarding your regexes - you might want to consider changing \d to \d+ so that you can catch numbers greater than 9, and you probably should use \s or \s+ instead of a literal space. UPDATE: UPDATE: Thanks belg4mit! jeffa L-LL-L--L-LL-L--L-LL-L-- -R--R-RR-R--R-RR-R--R-RR B--B--B--B--B--B--B--B-- H---H---H---H---H---H--- (the triplet paradiddle with high-hat)	[reply] [d/l]
Re: (jeffa) Re: How do I re-read a line? by belg4mit (Prior) on May 01, 2002 at 22:42 UTC
This seems to work fine for me... `my($index, $record) = (0); while (<$fh>) { chomp; s/\cM//; if (/^(\d)-RECORD$/) { $index = $1; $record->[$index]->{items} = []; } elsif (/ITEMNAME-(.)/) { $record->[$index]->{name} = $1; } elsif (/ITEM \d-(.)/) { push @{$record->[$index]->{items}}, $1; } }` [download] UPDATE: jeffa pointed out an undef in the record set. that's due to the lack of a `0-RECORD`. Ignore it, or do `$index = $1 -1`, as you will. `-- perl -pew "s/\b;([mnst])/'$1/g"`	[reply] [d/l] [select]
Re: How do I re-read a line? by Dog and Pony (Priest) on May 01, 2002 at 21:57 UTC
I admit I didn't test this, but wouldn't `next if /^\d-RECORD$/;` [download] work in this particular case? Actually, if you also did `next if /^$/;` [download] it would seem that you wouldn't need the external check for `/^\d-RECORDS$/`. This is all based on the code I can see, of course, you may have other reasons for wanting to process each record separately, with one sub call each? Otherwise, I'd restructure my logic a bit. Update: to see what I mean, listen to the not-as-lazy-as-me Mr. Muskrat. :) Hope that helps. :) You have moved into a dark place. It is pitch black. You are likely to be eaten by a grue.	[reply] [d/l] [select]