Pipeline 2 Overview

The DAISY Pipeline 2 is an ongoing project to develop an open-source, cross-platform framework for the automated processing of various digital formats efficiently and economically. It intends to facilitate the production and maintenance of accessible content for people with print disabilities.

A follow-up of the DAISY Pipeline 1 project

The initial DAISY Pipeline project was started in 2006. Since then, new standards and technologies have emerged and are now being embraced in a total redesign of the DAISY Pipeline framework. These new foundations allow to reduce the maintenance cost, lower the learning curve for new adopters, and increase the interoperability with heterogeneous production workflows.

The overarching principles remain the same:

  • provide functionality to produce, maintain, and validate accessible digital formats
  • embrace good practices for the creation of quality accessible content
  • support the single source master approach where applicable
  • minimize overlap and duplication, notably via the development of reusable components

Feature Highlights

The heart of the DAISY Pipeline 2 project is a comprehensive framework for the development and execution of automated document processing:

  • Based on standard XML processing technologies

    ↪ Notably W3C recommendations like XProc (workflow description) and XSLT 2.0 (document conversion), but also XPath 2.0, OASIS XML Catalogs, etc.
    ↪ These standard technologies are platform neutral, supported by active communities, and easy to maintain.

  • Designed with a modular architecture

    ↪ Processing functionality is provided by cohesive and reusable software components, called "modules".
    ↪ The runtime framework is backed by a service-oriented and dynamic module system (OSGi).
    ↪ Modularity is the key to a better scalability and extensibility.

  • Interoperable with heterogenous production workflows

    ↪ The framework is accessible via a platform neutral RESTful API (web service).
    ↪ Processing steps can easily invoke third-party applications or even remote services.

  • Open source

    ↪ All software components maintained by the DAISY Consortium are available under a business-friendly license (LGPL)
    ↪ Reuse and integration in both open or commercial contexts is allowed and welcome.

  • Collaborative

    ↪ The project is led and maintained by the DAISY Consortium but involves several member organizations (including NIMAS/CAST, NLB, RNIB, SBS, TPB).
    ↪ This reduces the duplication of effort and ensures maximum sharing of best practices among the user community.

Concrete processing functionality is provided by a set of modules based and executed on the aforementioned framework. The primary focus of the chartered Pipeline 2 project is to support the production of accessible EPUB 3 material and the migration of XML masters to the new DAISY AI format (officially the ANSI/NISO Z39.86-201x "Authoring and Interchange", colloquially known as ZedAI). Currently availabe functionality incudes:

  • DTBook to DAISY AI
  • DAISY AI to EPUB 3
  • DAISY 2.02 to EPUB 3 (optionally with Media Overlays)

and more is in the works:

  • DTBook to EPUB 3
  • DAISY 3 to EPUB 3 (optionally with Media Overlays)
  • HTML to EPUB 3
  • Annotate EPUB 3 for TTS
  • and more...

Deployment Options

The primary entry point to the framework is a Web Service API, based on the REST prinicples. The API allows the DAISY Pipeline 2 to be called from any client application, in any programming language. The core runtime can be deployed either in "local mode" (having access to the local file system) or in "remote mode" (enabling input/output interchange with remote applications).

In addition to the programming interface, a command-line interface is provided and allows the execution of jobs from a text terminal or shell script.

In the works is also a web application that will allow the execution of jobs from a standard web browser.

More Information

For more informationm please refer to: