alemela / trifid-ld

Lightweight Linked Data Server and Proxy

Trifid-LD - Lightweight Linked Data Server and Proxy

Trifid-LD provides a lightweight and easy way to access Linked Data URIs via HTTP. In the Linked Data world this is often called dereferencing. Trifid-LD is inspired by Pubby and written in (server side) JavaScript.

Features:

• Provides a Linked Data interface to SPARQL protocol servers
• Provides an in-memory store suitable for smaller data sets and testing environments
• Provides a simple HTML interface showing the data available about each resource
• Support for multiple HTML templates based on namespace patterns
• HTML view is providing embedded JSON-LD, which is rendered by client-side JavaScript
• Takes care of content-negotiation
• Runs well behind HTTP reverse proxies like Varnish
• Provides a SPARQL proxy and YASGUI as web frontend
• Does not do 303 redirects because we don't like the extra round-trip

Requirements:

• A SPARQL endpoint (accessed via HTTP, needs to support JSON-LD for HTML view)
• Or for development some triples in a local file. In this case the built-in in-memory store can be used.

Trifid-LD supports all content-types provided by the SPARQL endpoint and does not do additional format conversion.

Installation

Trifid-LD is a Node.js based application. To install and run it you either have to have Node.js or IO.js installed or you can use Docker an build a docker image containing Trifid-LD and all its dependencies.

Installing with Node.js or IO.js

Make sure you have either Node.js or IO.js installed. Once this is done clone the Github repository and run

npm install

to install all module dependencies.

Trifid-LD is using Bunyan for logging. To pretty print log output on the console you might want to install it globally using

npm install -g bunyan

To start the server execute

npm start | bunyan

Note that it is not recommended to run Node applications on well-known ports (< 1024). You should use a reverse proxy like Varnish instead.

Installing/Using with Docker

To use Trifid-LD with this method you only need to have Docker installed, see https://docs.docker.com/installation/ for installation intructions for your platform.

Once Docker is inatlled clone the Github repository and run

docker build -t trifid .

This creates an image named trifid that you can execute with

docker run -p 8080:8080 trifid 

Once it is started you can access for example http://localhost:8080/data/person/sheldon-cooper.

Using it

Trifid-LD comes with two configurations

Default configuration

The default configuration config.js uses the LDP module with an in memory store and a sample dataset with characters from The Big Bang Theory.

During the init process the graphs are splitted, which may take some time for bigger graphs. For development the LDP module is useful but it should be replaced by a SPARQL endpoint in production environments.

Using a HTTP SPARQL endpoint

For production systems we recommend data access via the SPARQL 1.1 Protocol interface.

config.sparql.js is already prepared to work with the sample Fuseki configuration in data/scripts/. The folder contains scripts to run a Fuseki server with the same example data used in the LDP module configuration, adjust this according to your needs.

If you run Trifid-LD behind a reverse proxy, the proxy must set the X-Forwarded-Host and X-Forwarded-Port header fields. This constructed URI is used in the SPARQL DESCRIBE query which is also defined in the configuration. If you want to use another query you need to adjust the buildQuery method.

SPARQL Proxy

The SPARQL web proxy is by default listening on /sparql. If you want to adjust that alter the configuration in the sparqlProxy section. Note that SPARQL is currently not supported by the in-memory store.

Folder Structure

data/public

Static files (JavaScript, CSS, images) for HTML rendering. js/render-ld.js contains the JSON-LD graph render code.

data/scripts

Scripts for external applications (Fuseki).

data/templates

HTML templates for graph rendering and errors.

Support

Issues & feature requests should be reported on Github.

Pull requests are very welcome.

HTTPRange-14 or Ceci n'est pas une pipe

Trifid-LD does not care about HTTPRange-14 and neither should you. We consider the extra 303 redirect round-trip a waste of precious response time. You get back what you asked for in content-negotiation.

We are aware that the URI and the representation are not the same thing but if we ever have to care about it, we would implement it using the Content-Location header field.

License

Copyright 2015 Zazuko GmbH

Trifid-LD is licensed under the Apache License, Version 2.0. Please see LICENSE and NOTICE for details.