File::Find Question

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

I have a directory structure, for example:

/nfs/export/netflow/intermediate/
/nfs/export/netflow/intermediate/200101-200102
/nfs/export/netflow/intermediate/200103-200104
/nfs/export/netflow/intermediate/questionable-200101-200102
/nfs/export/netflow/intermediate/questionable-200103-200104
/nfs/export/netflow/intermediate/current@
/nfs/export/netflow/intermediate/previous@
[download]

I want the code to look for any directory(s) beginning with "questionable", and any directory(s) matching the date-range convention of yyyymm-yyyymm. I am using File::Find to recurse through the top-level directory. Here's my problem: I need to process files contained the directories

/nfs/export/netflow/intermediate/200101-200102
/nfs/export/netflow/intermediate/200103-200104
[download]

differently than the files contained in:

/nfs/export/netflow/intermediate/questionable-200101-200102
/nfs/export/netflow/intermediate/questionable-200103-200104
[download]

HOWEVER, the files contained in both the "questionable" directories and the "yyyymm-yyyymm" directories are named using the same convention.

File names in either directory will be, for example:
bgp-nexthop.chicago4-cr8.20010307.gz or unknown-prefix.chicago4-cr8.20010307.gz

I know that File::Find will ignore any symlinks by default, which is what I want. But, I'm not sure how to pass separate functions to files, based upon what directory they reside in.
Here's some code:

#!/usr/local/bin/perl -w
use strict;
use File::Find;

my $intdir = "/nfs/export/netflow/intermediate";

find(\&get_good_int, $intdir);

sub get_good_int {
    chomp;
    next if ( $_ =~ /^unknown/);
    my @list = split;
    foreach my $file (@list) {
         print "My filename is: $file\n";
    }
}
[download]

This prints the contents of "intermediate", without recursion. Obviously, the subroutine, &get_good_int doesn't do any thing useful yet, but I want to make sure that I'm grabbing the correct files, first. From reading other posts re: File::Find, I think that it is wise to mention that I am on a machine running 5.003_26.

Comment on File::Find Question Select or Download Code

Replies are listed 'Best First'.
Re: File::Find Question by repson (Chaplain) on Mar 14, 2001 at 10:58 UTC
What you want to start with is perlman:-X, which describes a series of functions for testing file/directory types. These can be used with $_ which will be set to the current filename, and the directory will be chdired to the appropriate place to use it. $File::Find::name (the full path) and $_ can then be used with common string operations to find if the file/directory name suits your conditions. Here is a possible starting point I put together: `find( sub { if (-d $_) { if (/^questionable/) { dir_func_1($File::Find::name); } elsif (/-(\d{6}-\d{6})$/) { dir_func_2($_,$1); } # Uncomment following line to prevent furthur recursion. # $File::Find::prune = 1; } else { # you may want furthur tests before assuming good file file_func($_); } }, $intdir);` [download] Updated: Changed chmod to chdir, stupid me	[reply] [d/l]
Re: File::Find Question by buckaduck (Chaplain) on Mar 14, 2001 at 20:02 UTC
There's a few bits of your sample code that don't seem right: There's no need to chomp the filename. (harmless though) It's odd that you should specify $_ in the pattern match, since you appear to understand how to use it as the default for the split command later. (also harmless) Why are you splitting the filename on whitespace? The variable $_ contains only one filename, because the subroutine is being called once for each file. It looks like you think that $_ contains a list of filenames... Even so, I don't see why your code wouldn't recurse. It just doesn't do what you want. Your test sub could be written more simply: `sub get_good_int { next if /^unknown/; print "My filename is: $_\n"; }` [download] Something closer to being useful might be this: `use File::Find; use File::Basename; find (\get_good_int, '/nfs/export/netflow/intermediate'); sub get_good_int { next if /^unknown/; # If the directory path includes this type of dir if ($File::Find::dir =~ m#/questionable-\d{6}-\d{6}#) { myfunc1($_); } # If it contains this other type of dir elsif ($File::Find::dir =~ m#/\d{6}-\d{6}#) { myfunc2($_); } }` [download] buckaduck	[reply] [d/l] [select]