using File::Find for reading files in order

Ashes.cfg has asked for the wisdom of the Perl Monks concerning the following question:

hello All, Its a long time I have worked on Perl(an yr).I have a simple question. I am using File::Find module to traverse through my directory and access my file

sub messages {

  my $dir = shift or return;

  my @messages;
 find(sub { next if /^\.+$/ || -d $File::Find::name; push @messages, $
+File::Find::name }, $dir);

  return @messages;
}
[download]

The Output I get is really strange..Consider I have 100 files(numbered 1 thru 100)...somehow the way the files are accessed is

File 1
File 10
File 100
File 2
File 20
File 21
.
. and so on
[download]

I think I am making a mistake in the regex.Holy Monks...Kindly help me.... The Rest of my code is working correctly. Regards Ashes

Comment on using File::Find for reading files in order Select or Download Code

Replies are listed 'Best First'.
Re: using File::Find for reading files in order by ysth (Canon) on Feb 04, 2009 at 06:50 UTC
Use preprocess to tell find what order to process each directory in (untested): `find( { preprocess => sub { sort { $a <=> $b \|\| $a cmp $b } @_ }, wanted => sub { next if /^\.+$/...} }, $dir );` [download] -- Online Fortune Cookie Search Office Space merchandise	[reply] [d/l]
Re^2: using File::Find for reading files in order by grinder (Bishop) on Feb 04, 2009 at 07:27 UTC
Either that, or if there are many files that will be skipped by the regexp, sort the results just before returning them. `return sort {$a <=> $b} @messages;` [download] • another intruder with the mooring in the heart of the Perl	[reply] [d/l]
Re^2: using File::Find for reading files in order by Ashes.cfg (Initiate) on Feb 04, 2009 at 06:57 UTC
Tht worked wonders.. Thanks a lot ysth. I think I would be more on perlmonks in coming months, since I might have to script a lot in coming weeks	[reply]
Re: using File::Find for reading files in order by jwkrahn (Abbot) on Feb 04, 2009 at 10:23 UTC
If you had warnings enabled then you would have received the message `Exiting subroutine via next at YOUR_PROGRAM line N.` You need to use return to exit from a subroutine instead.	[reply] [d/l]
Re: using File::Find for reading files in order by leocharre (Priest) on Feb 04, 2009 at 14:37 UTC
I'm just curious here.. Why do you think File 1 File 10 File 100 File 2 File 20 File 21 Is out of order? Or strange at all? It's obviously sorted alphanum. You expected to sort by comparison, by.. is two greater than, 100. It's not greater ammount, but 100 does come before 2 alphanumerically. Just wanted to point that out- because this is a bug in the coder, not the code. If you want to be able to sort files like these as you expect, one thing to do is name them appropriately... # rename ' ' ' 00' File\ ?; rename ' ' ' 0' File\ ?? # ls -1 File\ * File 001 File 002 File 010 File 020 File 021 File 100	[reply]
Re: using File::Find for reading files in order by hbm (Hermit) on Feb 04, 2009 at 13:50 UTC
Another tip: In the find sub, `$_` is equivalent to `File::Find::name`, so you could do: `find sub { push @messages, $_ if (-d $_ && !/^\.+$/); }, $dir` [download]	[reply] [d/l]
Re^2: using File::Find for reading files in order by merlyn (Sage) on Feb 04, 2009 at 15:14 UTC
In the find sub, $_ is equivalent to File::Find::name Nope. $_ holds the name relative to the current directory at the time of the subroutine call, which in the default case has been chdir'ed to the location of the file. But `$File::Find::name` is always relative to the original directory. So they aren't the same. Definitely different uses. You can use $_ inside the subroutine, but if you want to save the name for later, you want to use $File::Find::name. And in fact, it'd be wrong to use $File::Find::name within the subroutine, because your relative name would have been applied twice. Ouch. -- Randal L. Schwartz, Perl hacker	[reply] [d/l]
Re^3: using File::Find for reading files in order by hbm (Hermit) on Feb 04, 2009 at 16:37 UTC
merlyn, Thanks for clarifying. I was hesitant to say "equivalent"; even to say anything other than "you can shorten your code..." And now I've reread the doc, a Good Thing. I've reread the OP too, and I think the claimed output is misleading. Surely it is more like the following? (In which case my suggestion can be stricken, as it would only yield the files.) `/path/to/File 1 /path/to/File 10 /path/to/File 100 /path/to/File 2 /path/to/File 20 ...` [download] Whew.	[reply] [d/l]