Creating web services with SWI-Prolog is a breeze. This how-to is intended as your first bluffer's guide. All examples come with fully operational source code.

• Hello World
• Development support for web servers
• Hello World in HTML
• Rule definition facilities (html//1 rules)
• Processing HTTP parameters
• How do I upload a file, preserving the filename?
• Adding style (CSS and simple JavaScript) to your pages
• Serving plain files
• Making your web server scalable

Deploying Prolog-based web servers

• Making the server accessible at port 80
• Running the server under VNC
• Running the server from Unix (x)inetd
• Running the server from Unix init
• Running the server from Linux upstart

This is work in progress! If you want to contribute or comment, send an e-mail to me or to the mailinglist.

Tutorials

Here is a tutorial with exercises by Anne Ogborn.

Here is a sample implementation of a JavaScript client communicating with a SWI-Prolog server, using the newer ibrary(http/http_server) API.

Example code

We have two servers available with complete source. One is the server that runs http://www.swi-prolog.org, which can be downloaded through GIT from the repository plweb.git. The other is ClioPatria. Although ClioPatria targets semantic web (RDF) applications, it is also useful as a starting point for general purpose web servers.

Here is a very incomplete selection of functionality that you can find in these servers:

• SWI-Prolog.org
  • Call CGI scripts (http_cgi.pl)
  • Autocompletion (autocomplete.pl)
  • PlDoc based wiki processing (wiki.pl, wiki_edit.pl)
  • Stress testing (directory stress)
  • Logfile handling (directory logs)
• ClioPatria
  • User management, OpenID
  • Extension packs (CPACK)
  • Modular server configuration
  • Server statistics
  • Handle messages from Prolog
  • Lots of RDF stuff

Plans (todo)

• AJAX interaction (YUI, simple table, tree, N-Queens code?)
• PWP
• Logfiles and processing thereof
• Stress-testing by replaying logfiles