NAME

       TM::Tree - Topic Maps, trait for induced tree retrieval

SYNOPSIS

         use TM::Materialized::AsTMa;
         my $tm = new TM::Materialized::AsTMa (file => 'old_testament.atm');
         Class::Trait->apply ( $tm => 'TM::Tree' );
         $tm->sync_in;

         # old-testament pedigree
         print Dumper $tm->tree (lid   => 'adam',
                                 type  => 'has-parent',
                                 arole => 'parent',
                                 brole => 'child' );

         # new-testament inverse pedigree
         print Dumper $tm->tree (lid   => 'gw-bush',
                                 type  => 'has-parent',
                                 arole => 'child',
                                 brole => 'parent' );

DESCRIPTION

       Obviously, topic maps can carry information which is tree structured. A family pedigree is
       a typical example of it; associations having a particular type, particular roles and you
       can derive a tree structure from that.

       This is exactly what this operator does: it takes one topic map basis and primed with a
       starting node, an association type and two roles a tree will be returned to the caller.

INTERFACE

   Methods
       tree
           $treeref = $tm->tree (
                            $start_topic,
                            $type_topic,
                            $role_topic,
                            $role_topic,
                            [ depth => $integer ])

           $treeref = $tm->tree_x (
                            $start_topic,
                            $type_topic,
                            $role_topic,
                            $role_topic,
                            [ depth => $integer ])

           This function will analyze the topic map and will detect all maplets of the given type
           (direct and indirect ones) having the specified roles. Starting from the start topic
           it will so find other topics playing the brole. Those will be used as a next starting
           point, and so forth.

           To avoid the tree to become too big, you can impose an optional limit. Loops are
           detected.

           Every output tree node contains following fields:

           "lid":
               the lid of the node

           "children":
               a list reference of child nodes, there is no specific sort order

           "children*":
               Note: This is currently deactivated.

               for convenience this list reference contains all children, grand-children, grand-
               grand children.... of this node (this list is neither sorted nor unique out of
               performance considerations).

           The version "tree_x" does not honor subclassing of roles and type (but "tree" does).
           This means that is can be considerably faster, especially if you use it for taxonomy
           stuff with "isa" and "is-subclass-of".

       taxonomy
           $treeref = $tm->taxonomy ([ $start_lid ])

           This function is a specialization of "tree", in that it looks at a particular
           association type ("is-subclass-of") and the appropriate roles ("superclass",
           "subclass"). Obviously the result is a tree holding all subtypes.

           The only optional parameter is a toplet "lid"; that becomes the starting point of the
           tree. If that parameter is missing, "thing" is assumed.

SEE ALSO

       TM

COPYRIGHT AND LICENSE

       Copyright 200[3-6] by Robert Barta, <drrho@cpan.org>

       This library is free software; you can redistribute it and/or modify it under the same
       terms as Perl itself.