Shlomi Fish


File::Find::Object - An object oriented File::Find replacement


    use File::Find::Object;
    my $tree = File::Find::Object->new({}, @targets);

    while (my $r = $tree->next()) {
        print $r ."\n";


File::Find::Object does the same job as File::Find but works like an object and with an iterator. As File::Find is not object oriented, one cannot perform multiple searches in the same application. The second problem of File::Find is its file processing: after starting its main loop, one cannot easily wait for another event and so get the next result.

With File::Find::Object you can get the next file by calling the next() function, but setting a callback is still possible.



    my $ffo = File::Find::Object->new( { options }, @targets);

Create a new File::Find::Object object. @targets is the list of directories or files which the object should explore.



Boolean - returns the directory content before the directory itself.


Boolean - doesn't continue on filesystems different than the parent.

Boolean - follow symlinks when they point to a directory.

You can safely set this option to true as File::Find::Object does not follow the link if it detects a loop.


Function reference - should point to a function returning TRUE or FALSE. This function is called with the filename to filter, if the function return FALSE, the file is skipped.


Function reference - should point to a function, which would be called each time a new file is returned. The function is called with the current filename as an argument.


Returns the next file found by the File::Find::Object. It returns undef once the scan is completed.


Returns the current filename found by the File::Find::Object object, i.e: the last value returned by next().


Like next() only returns the result as a convenient File::Find::Object::Result object. $ff->next() is equivalent to $ff->next_obj()->path().


Like item() only returns the result as a convenient File::Find::Object::Result object. $ff->item() is equivalent to $ff->item_obj()->path().


Sets the children to traverse to from the current node. Useful for pruning items to traverse.

Accepts a single array reference.


    $ff->set_traverse_to([ grep { ! /\A\./ } @{ $ff->get_traverse_to }]);


Prunes the current directory. Equivalent to $ff->set_traverse_to([]).

[@children] = $ff->get_traverse_to()

Retrieves the children that will be traversed to. Returns a single array reference.

(Example under set_traverse_to).

[@files] = $ff->get_current_node_files_list()

Gets all the files that appear in the current directory. This value remains constant for every node, even after traversal or calls to set_traverse_to() and is useful to use as the basis of the argument for set_traverse_to().

Returns a single array reference.


No bugs are known, but it doesn't mean there aren't any.


There's an article about this module in the Perl Advent Calendar of 2006:

File::Find is the core module for traversing files in perl, which has several limitations.

File::Next, File::Find::Iterator, File::Walker and the unmaintained File::FTS are alternatives to this module.


Copyright (C) 2005, 2006 by Olivier Thauvin

This package is free software; you can redistribute it and/or modify it under the following terms:

1. The GNU General Public License Version 2.0 -

2. The Artistic License Version 2.0 -

3. At your option - any later version of either or both of these licenses.