Understanding Greenstone Digital Library Interfaces: Reader & Librarian

When a library decides to put its collections online, it needs software that can handle two very different groups of people. On one side are the librarians who build and organise the collection. On the other are the readers who simply want to find and use the material. Greenstone, the open-source digital library software developed by the New Zealand Digital Library Project at the University of Waikato in cooperation with UNESCO and the Human Info NGO, solves this neatly by offering two separate interfaces. Understanding how each one works is the key to running a successful digital library, so let’s break them down.

The two faces of Greenstone 🔗

Greenstone is not a digital library in itself. It is a tool for building digital libraries. To do this, it provides two separate interactive interfaces: the Reader interface and the Librarian interface. The distinction is simple. End users access the digital library through the Reader interface, which runs inside a web browser. Library staff build and manage collections through the Librarian interface, a separate desktop application.

This separation matters. The person searching for a document does not need to see metadata editors, plugin settings, or indexing options. The person building the collection needs all of those tools but does not want them cluttering the public-facing pages. By splitting the work into two interfaces, Greenstone keeps each experience clean and focused.

The librarian’s interface 🔗

The Greenstone Librarian Interface, usually shortened to GLI, is a Java-based graphical user interface. It is the tool that librarians, archivists, and other digital collection managers use to develop their collections. In the early days, Greenstone offered three ways to build collections: a command line, a web interface, and the GLI. Over time the GLI became the most popular method because it makes a complex process simple and visual.

The strength of the GLI is that it walks you through collection building in a logical sequence. The interface is organised into panels, and you move through them roughly in order as you develop a collection. The main panels are Gather, Enrich, Design, Create, and Format.

Gather: bringing documents together 🔗

The Gather panel is where collection building begins. You select files from your computer or local network and simply drag them into the collection area. Greenstone can also download material from external sources using standard protocols such as the Open Archives Initiative (OAI), Z39.50, and the web itself. A helpful detail is that when you organise documents into folders, metadata assigned at the folder level is inherited by the files inside it, which saves a great deal of repetitive work.

Enrich: adding metadata 🔗

Once documents are gathered, the Enrich panel lets you add metadata. Metadata can be added to individual documents, to several documents at once, or to entire folders. Greenstone comes with predefined metadata sets, including Dublin Core, RFC 1807, and government locator standards such as the New Zealand and Australian Government Locator Services. If these do not fit your needs, you can use the built-in metadata set editor (GEMS) to design your own.

Design: shaping search and browse 🔗

The Design panel is the most technical part of the GLI. Here you decide how documents will be processed and how readers will interact with the collection. This is where you choose your search indexes, partition indexes, and browsing classifiers. Plugins tell Greenstone how to process each type of file, since every document must be handled by a plugin. The decisions you make in this panel directly shape what the reader will eventually see and be able to do.

Create and Format: building and presenting 🔗

The Create panel is where you actually build the collection by clicking a single button. Greenstone then processes all the documents, applies the plugins, and constructs the indexes. The Format panel controls how the collection looks and behaves. A useful point to remember is that changes in the Format panel take effect immediately, but if you change anything in the Gather, Enrich, or Design panels, you must rebuild the collection for the changes to appear.

Building collections remotely 🔗

The GLI is flexible about where the collection actually lives. Through its client-server arrangement, librarians can use the interface to manage Greenstone installations on a remote server. The GLI Client works almost identically to the regular interface, with only minor differences such as the absence of CD-ROM writing options. This makes it possible for a librarian to work on collections hosted elsewhere without installing the full system locally.

The reader’s interface 🔗

The Reader interface is what the public sees. It operates entirely within a web browser, so end users need nothing more than an internet connection to browse and search the library. Everything that the librarian designed in the GLI is presented here in a clean, accessible form.

The reader’s experience centres on two activities: searching for specific items and browsing through organised lists. In Greenstone 3, the web interface also includes built-in library management facilities that become available once a user logs in, which is an improvement over the earlier version where such management lived only in the GLI.

Searching the collection 🔗

Greenstone lets you build search indexes from both the full text of documents and the metadata associated with them. A reader can perform a plain search by typing a word or phrase into a search box. For more precise results, Greenstone supports advanced searching, including Boolean queries that combine terms with operators like AND, OR, and NOT.

The kind of searching available depends on the indexer chosen during the Design stage. Greenstone offers three indexing tools: MG, MGPP, and Lucene. MG is the original indexer and handles section-level indexing. MGPP performs word-level indexing, which enables phrase, proximity, and fielded searching. Lucene, developed by the Apache Software Foundation, offers similar capabilities with added features such as wildcard and range searching, and it supports incremental building. Advanced Boolean searching is switched on through the “Enable Advanced Searching” option in the Search Types section of the Design panel.

Browsing the collection 🔗

Not every reader arrives knowing exactly what they want. Browsing serves those who prefer to explore. In Greenstone, browsing is handled by browsing classifiers, which generate lists from the metadata attached to documents. Readers can browse by author, title, date, subject, and other fields.

Greenstone also supports hierarchical browsing, where documents are organised into nested categories that the reader can drill down through. For example, a subject hierarchy classifier lets readers explore broad topics first and then narrow down to specific areas. This kind of structured, hierarchical navigation is particularly useful for large collections where a flat list would be overwhelming.

Customisation and multilingual support 🔗

One of Greenstone’s notable strengths is how much it can be tailored to a particular library’s needs. The look and behaviour of the digital library are not fixed, and several technologies work together to make this flexibility possible.

Tailoring the look and behaviour 🔗

CSS (Cascading Style Sheets) controls the visual design of the web interface. Librarians can change fonts, colours, and layout so that the digital library matches their institution’s branding. Java servlets play a central role in Greenstone 3, which is itself written in Java and built on web technologies such as XML Transforms (XSLT). Perl scripts are used throughout Greenstone to process text and automate server-side tasks, and they can be extended to handle custom workflows such as batch importing or generating reports.

Serving many languages 🔗

Greenstone is Unicode-compliant, which means it can build, search, and browse documents in any Unicode-supported language. The interface itself can be presented in many languages, and recent development has focused heavily on improving multilingual support, including better handling of non-ASCII filenames and character-based segmentation for languages such as Chinese, Japanese, and Korean.

This capability is especially valuable in a multilingual setting. A library can offer interfaces and content in regional languages alongside English, making the collection genuinely accessible to communities that speak different languages. For institutions serving diverse user groups, this ability to localise the library is a major practical advantage.

Why the two-interface design works 🔗

The real elegance of Greenstone lies in how the two interfaces connect. The librarian makes every important decision in the GLI: which documents to include, what metadata to attach, which search indexes to build, and how readers will browse. All of that work is then quietly delivered to the reader through a simple web interface. The reader never has to understand plugins or classifiers. They just search and browse.

This division of labour is what allows a single piece of open-source software to serve both a small local collection and a large institutional archive. The complexity stays where it belongs, with the people building the library, while the simplicity is reserved for the people using it.

What do you think? If you were setting up a digital library for your own institution, which feature would matter most to you: the precision of advanced Boolean searching or the accessibility of a multilingual browsing interface? And how might the needs of your particular readers change the way you design the collection in the GLI?

References

1. https://www.greenstone.org/factsheet
2. https://wiki.greenstone.org/doku.php?id=en:beginner:building_collections
3. https://ebooks.inflibnet.ac.in/lisp8/chapter/case-studies-greenstone-digital-library-software/
4. https://files.greenstone.org/tutorial/gs2-current/en/all_tutorials.html
5. http://what-when-how.com/how-to-build-a-digital-library/distributed-operation-digital-library/
6. https://wiki.greenstone.org/doku.php?id=old:more_about_indexing
7. https://www.lispower.in/2021/04/green-stone-digital-library-software-note.html