Technical Overview

Here's a quick overview of how the parts of the SHARE notification service work together.

The main harvester processor is called scrAPI, and the code can be found here: https://github.com/fabianvf/scrapi

The Code Behind the Notification Service

The SHARE Notification Service has two main hubs: scrAPI, and the OSF.

scrAPI

ScrAPI is the "brains" of the SHARE Notification service - it coordinates all of the harvesters, makes sure that each task is run when it is scheduled to run, and stores all of the raw metadata for later use.

All of the code for scrAPI is located the scrAPI github repository.

Once scrAPI gathers all of the metadata from each source, it then sends it to the OSF for further processing, and ultimate notification distribution.

The OSF

The output and outside access portions of the SHARE Notification Service works together with the Open Science Framework, and takes advantage of the robust structure, scalability, and security that the OSF offers.

The Open Science Framework provides the scaffolding for the many sources of output for the notification service, including the RSS and atom feeds, resourcesync, and pubsubhubbub updates. While it is under active development, all of the code for the notification service is located off of the main branch.

Here's a link to the branch of the OSF that contains all of the code for the notification service. The code for the notification service right now is classified as an OSF Application - and so all of the routes and functions are located under the applications addon section.

The SHARE notification service is currently an OSF application with the application ID 6qajn - which is why the SHARE rss feed, raw json, and resourcesync access points have that guid in their urls.

For example, the RSS feed (and other notifications like it) are generated using elasticsearch queries of the notifications database.