Preferences (and Project Properties)

Last revised 11/17/2020. Contact us.

Table of Contents

• Introduction
• SADL Preferences by Category
  • General SADL Settings
  • Graphing Settings
  • Inference and Querying Settings
  • Metrics Settings
  • Translation Settings
  • Content Assist Settings

Introduction

Preferences are user-settable parameters that control the way the SADL environment behaves during editing and during inference. In the Eclipse environment, preferences can be set at the workspace level or at the project level. Workspace preferences are set from the Windows dropdown menu.

To set preferences in Eclipse at the project level, select the desired project and request Project Properties, either from the Project dropdown menu or from the project context menu in Project Explorer.

The SADL preference page in Eclipse looks like this.

By default everything is disabled and the link in the upper-right takes the user to the workspace preferences. To enable project-specific preferences, check the checkbox with the label "Enable project specific settings". Note that the project preferences are initially set to the workspace preferences.

If SADL in the left-side navigation menu of project properties is expanded, Reasoner Preferences becomes visible as a menu choice and, if selected, allows a reasoner and an associated translator to be specified for the project. Note that Reasoner Preferences can only be specified at the project level. Reasoner Preferences are NOT Eclipse preferences and are sometimes referred to as project properties. These are stored in the project's OwlModels folder in the configuration.rdf file. This file will carry the Reasoner Preferences forward to the Knowledge as a Service (KaaS) environment.

By clicking on a reasoner (not on the checkbox in front of the reasoner, but on the name of the reasoner) the reasoner is highlighted and the Edit button is made active. Clicking the Edit button takes one to the reasoner-specific settings available for the particular reasoner selected. For example, the default Jena reasoner has the following settings.

Configuration settings for a specific reasoner are discussed in the documentation for that reasoner.

SADL Preferences by Category

General SADL Settings

In the Eclipse environment, General SADL Settings include the following.

• Base URI: this should be a valid URI (starting with "http://") and will be used if content assistance is requested after the uri keyword. The name of the SADL file will be appended to create the inserted URI for the model.
• Saved OWL model format:  this preference specifies the serialization format to be used for the OWL models translated from the SADL models. All but the last, Jena TDB, are formats for OWL files saved in the OwlModels folder. Jena TDB is a binary triple store format and, if selected, will result in a triple store of this type being created under the OwlModels folder. The triple store will contain a named graph for each SADL model in the project.
• Show import model list as: if content assistance is requested after the import keyword, this preference determines whether the resulting list will display model namespaces or SADL file names for selection.
• Show prefixes for imported concepts only when needed for disambiguation -- if checked, lists of model concepts shown as a result of a request for content assistance will only show a prefix before the name if the name would otherwise be ambiguous, i.e., there are multiple models in the project that define concepts with that name.
• Show Timing Information (Build, Reasoning) -- if checked, the amount of time required to build or do reasoning will be displayed in the console window.
• Interpret Date 10/11/2012 as: this preference controls whether dates using this format are interpreted as month first or day first.
• Check for ambiguous names --  if checked, names will be checked to see if they are ambiguous, meaning that more than one namespace known to the current model contains a concept with that name.

Graphing Settings

• Graph renderer package and class: specifies the implementation of the IGraphVisualizer Java Interface class to be used to render graphs. The value shown is the renderer distributed with the SADL IDE and requires download of GraphViz, see Visualization.
• Include Implicit Elements in Graph -- if checked, concepts from the implicit models SadlImplicitModel.sadl, SadlBuiltinFunctions.sadl, SadlListModel.owl, and SadlBaseModel.owl will be included in rendered graphs.
• Include Implicit Element Instances in Graph -- if checked, instances of concepts from the implicit models SadlImplicitModel.sadl, SadlBuiltinFunctions.sadl, SadlListModel.owl, and SadlBaseModel.owl will be included in rendered graphs.

Inference and Querying Settings

• Validate before Testing -- if checked AND if the specified reasoner is capable of doing model validation, the current model and imported models will be validated before doing inference.
• Show Namespaces in Query Results -- if checked, concepts found in query results will include namespaces (will be shown as complete URIs).
• Disable Deep Validation of Model -- for reasoners capable of doing validation, checking this preference disables deep validation, thus improving performance.
• Tabular data importer package and class: SADL includes the capability to provide a mapping to import tabular data into an OWL model. Importers must implement the ITabularDataImporter Java Interface class. The importer shown is for comma-separated-values (CSV) data.

Metrics Settings

• Enable Metrics Collection -- if checked, metrics will be generated as models are being built.
• Generate metrics report during project clean/build -- if checked, a metrics report will be generated (provided valid queries are provided) during clean/build.
• File containing metric queries: identifies a file containing SPARQL queries to be used to generate metrics reports. The Browse button can be used to locate the query file.

Note: metrics generation and reporting is not completely implemented in SADL at present.

Translation Settings

• Use indefinite and definite articles in validation and translation -- if checked, rules and queries can be written without explicit variables. Rather, an implicit variable is created by an indefinite article in front of a class name and references to the variable are written with a definite article. Ordinals are used if multiple variables of the same type are needed. See Definite and Indefinite Articles.
• Find and expand missing patterns in translation -- if checked, graph patterns in rules and queries will be checked to see if they are incomplete but unambiguous. If so, they will be expanded to complete graph patterns in translation. See Implied and Expanded Properties.
• Ignore Unitted Quantities (treat as numeric only) during translation -- if checked, numbers followed by units will be treated as if they were numbers only, rather than being translated into an instance of UnittedQuanitity.
• Translate multiple-class domain or range as union class (owl:unionOf) -- if checked, domains and ranges that are unions of multiple classes will be translated to explicit owl:unionOf constructs. Otherwise they will be translated to multiple <property> rdfs:domain <domain-class> statement. Similarly for range.

Type Checking Settings

• Check for cardinality of property on specific domain -- if checked, properties for a given domain class, or a subclass thereof, will be expected to have a cardinality restriction on that class for that property. If not a warning will be issued.
• Property range specification required -- if checked, the lack of a specified range for a property will be considered an error.
• Type checking issues as warning only -- if checked, any type checking issues will be reported as warnings and not errors. This allows OWL models to be build even if type errors are reported.

Content Assist Settings

• Filter SADL implicit model -- if checked, content assistance will not include any concepts from the SadlImplicitModel.