Table of Contents

• Motivation
• Supported browsers
• Adding TRILL on SWISH to your application
• Technology
• Source code

Motivation

TRILL on SWISH allows the user to write probabilistic Description Logic theories and compute the probability of queries with just a web browser. Moreover, it features all the functionalities of SWISH

SWISH is inspired by web-based tools such as JSFIDDLE in that it allows you to develop a program and then share it with others on the Web. Besides sharing basic Prolog code fragments, we see several opportunities for deploying SWISH:

• SWISH can be associated with a dataset, either in pure Prolog, RDF (ClioPatria) or a traditional relational database through ODBC and CQL. In this setting it can facilitate collaboratively management of increasingly complex queries against this dataset. Query results can be downloaded as CSV and queries can be executed through an HTTP API. See the File/Help ... for details.
• SWISH can be embedded in Prolog course material. See Learn Prolog Now! for a prototype. The menu File/Help ... explains how SWISH can be preloaded with source and example queries.
• SWISH can be used as a shared management and development interface for web services. The ClioPatria version allows for executing arbitrary management commands through the web interface, provided the user is logged in with administrative rights.

Supported browsers

SWISH works fine recent versions of in Safari, Firefox, Chrome and Internet Explorer (tested IE 11).

Adding TRILL on SWISH to your application

TRILL on SWISH can be added easily to any SWI-Prolog web application. For example, it is available as an extension pack for ClioPatria, the SWI-Prolog RDF framework. It can be added to ClioPatria using the command below, providing an alternative for SPARQL for exploring the database (requires SWI-Prolog 7.1.32 or later).

?- cpack_install(trill_on_swish).

Technology

SWISH is implemented as a set of jQuery plugins, using Bootstrap for menus, modal windows and various styling aspects. SWISH uses CodeMirror for editing the Prolog source and query. Queries are executed by sending both the source text (left editor) and query (bottom-right editor) to the server, which creates a Pengine (Prolog Engine) to execute the query.

Pengines have many other applications, such as supporting Prolog RPC (Remote Procedure Calls, which allows you to execute a query fully transparent, including backtracking, on a remote server. An interesting aspect is that they provide a generic query language on top of an arbitrary Prolog program that can be used from JavaScript. A demo is available

Source code

The source code for TRILL on SWISH is available on GitHub.

The source code for SWISH is available on GitHub.