Object Hierarchy:

Gtk.TreeModelFilter Gtk.TreeModelFilter Gtk.TreeModelFilter GLib.Object GLib.Object GLib.Object->Gtk.TreeModelFilter Gtk.TreeDragSource Gtk.TreeDragSource Gtk.TreeDragSource->Gtk.TreeModelFilter Gtk.TreeModel Gtk.TreeModel Gtk.TreeModel->Gtk.TreeModelFilter


[ CCode ( type_id = "gtk_tree_model_filter_get_type ()" ) ]
public class TreeModelFilter : Object, TreeDragSource, TreeModel

A TreeModelFilter is a tree model which wraps another tree model, and can do the following things:

  • Filter specific rows, based on data from a “visible column”, a column storing booleans indicating whether the row should be filtered or not, or based on the return value of a “visible function”, which gets a model, iter and user_data and returns a boolean indicating whether the row should be filtered or not.
  • Modify the “appearance” of the model, using a modify function. This is extremely powerful and allows for just changing some values and also for creating a completely different model based on the given child model.
  • Set a different root node, also known as a “virtual root”. You can pass in a TreePath indicating the root node for the filter at construction time.

The basic API is similar to TreeModelSort. For an example on its usage, see the section on TreeModelSort.

When using TreeModelFilter, it is important to realize that TreeModelFilter maintains an internal cache of all nodes which are visible in its clients. The cache is likely to be a subtree of the tree exposed by the child model. TreeModelFilter will not cache the entire child model when unnecessary to not compromise the caching mechanism that is exposed by the reference counting scheme. If the child model implements reference counting, unnecessary signals may not be emitted because of reference counting rule 3, see the TreeModel documentation. (Note that e.g. TreeStore does not implement reference counting and will always emit all signals, even when the receiving node is not visible).

Because of this, limitations for possible visible functions do apply. In general, visible functions should only use data or properties from the node for which the visibility state must be determined, its siblings or its parents. Usually, having a dependency on the state of any child node is not possible, unless references are taken on these explicitly. When no such reference exists, no signals may be received for these child nodes (see reference couting rule number 3 in the TreeModel section).

Determining the visibility state of a given node based on the state of its child nodes is a frequently occurring use case. Therefore, TreeModelFilter explicitly supports this. For example, when a node does not have any children, you might not want the node to be visible. As soon as the first row is added to the node’s child level (or the last row removed), the node’s visibility should be updated.

This introduces a dependency from the node on its child nodes. In order to accommodate this, TreeModelFilter must make sure the necessary signals are received from the child model. This is achieved by building, for all nodes which are exposed as visible nodes to TreeModelFilter's clients, the child level (if any) and take a reference on the first node in this level. Furthermore, for every row-inserted, row-changed or row-deleted signal (also these which were not handled because the node was not cached), TreeModelFilter will check if the visibility state of any parent node has changed.

Beware, however, that this explicit support is limited to these two cases. For example, if you want a node to be visible only if two nodes in a child’s child level (2 levels deeper) are visible, you are on your own. In this case, either rely on TreeStore to emit all signals because it does not implement reference counting, or for models that do implement reference counting, obtain references on these child levels yourself.

Namespace: Gtk
Package: gtk+-3.0



Creation methods:


Inherited Members:

All known members inherited from interface Gtk.TreeDragSource