Getting Started with Ferret

Ferret is an automatic tool, meaning that, once started, it responds automatically to your interactions with the Eclipse environment. For example, selecting an item from the Outline view causes Ferret to begin ferreting through the system to answer a set of queries for that particular item. The results of the queries are then listed in the Ferret view using an Eclipse tree view.

Ferret can also be explicitly invoked for items using:

• the keyboard shortcut F9, or
• through selecting the Query with Ferret from pop-up context menus, or
• through Navigate → Query with Ferret.

Configuration Jump Start

Ferret is only operational once the Ferret view has been opened. This view can either be manually opened using Window → Show View, or selecting Query with Ferret from an object's context menu as described above.

Ferret first requires some simple configuration to describe its query parameters before it can begin querying. This is accomplished using the Ferret Sphere Configuration Wizard which will pop-up on the first query, or which can be explicitly invoked using the Configure Spheres... from the Ferret view's view menu (). The configuration process is described in more detail elsewhere, but for now, simply select (1) the Java static program queries and (2) click on Finish.

With such a configuration, Ferret performs like a normal Java search tool.

Ferret Walk Through

Let's take a closer look at the Ferret view. This view displays the results of a set of queries applicable to the query elements.

This image highlights four parts:

1. This area displays a textual description of the current query element (or elements). This example features queries for the Java class EmpiricalFormula.
2. The Clustering/Fidelity column provides information about the clusterings of the results of a query. In this example, the results of instantiators has already been clustered by the declaring package of the results.
3. We can see the two clusters based on the declared package of the query results. The cluster org.manumission.chem can still be clustered in three different ways.
4. The Clustering/Fidelity column also indicates the fidelity of the result to the original query element(s).

The Ferret Toolbar

• : this toggle button supresses Ferret's automatic querying. Queries may still be made explicitly using the keyboard shortcut F9 or through the menus.
• : expand all top-level queries by one level.
• : collapse the tree to show only the top-level queries.
• : re-open original query element, or select from previously queried-for elements. This button is useful for returning to the start after some exploration. If there are multiple elements, then you will be prompted which parent to open.
• : a useful shortcut to open the containing element of the current query element. If there are multiple elements, then you will be prompted which parent to open.

The Ferret Query Results

Description

A description of the element.

Category

Queries are grouped by category.

#E

The total number of elements for either the query or the current cluster.

Clustering/Fidelity

The currently selected clustering, or the number of possible clusterings. For query results, this column displays the fidelity of the result.

When items correspond to elements in the workbench, they can opened by either double-clicking or using the Open from the item's context menu. The context menu can be used to open the item's container or querying on the item's container (when found), and other operations (when supported).

The results of a query can also serve as the context for further queries, displayed in the context of the tree.

Query Categories

Queries are sorted by their category. There are currently four query categories:

declarations

Queries about relations involving the declarations of elements.

inter-class

Queries about relations between different classes.

intra-class

Queries about relations within a class.

hierarchical

Queries about relations involving type inheritance.