Context Navigation

Papyrological Navigator

one component of IDPSoftware

The PN is "a customized search engine ... capable of retrieving information from multiple related sites." It is a custom web application, prototyped at Columbia that is intended to replace the current production applications for APIS and DDBDP, and to provide access to the full content of HGV records. It is therefore envisioned as being capable of searching both metadata and texts, and displaying metadata records, texts and images pulled from all three source datasets. Links to other data sources, such as Trismegistos?, are planned for the near term. Content expansion, to incorporate other full datasets, is envisioned further down the road. The PN employs the portal metaphor.

Background on the PN, from Columbia.

Software and language components:

Lots of Java:

Apache Tomcat
Apache Jetspeed-2 portlet container
- portlet spec
- Velocity templates
- Apache derby supports the portlet container, but is not really part of the app
- Some javax.naming implementations to bridge the bits of a request served by the container to the portlet app
Lucene
- For the PN, really just indexing code. The DDb has lots of fancy tokenizers, etc.
eRez/FSI images serving software, and the flash plugin FSI licenses to view images
- a web-app that serves up XML config files for the Flash plugin based on the APIS id of the document in question

Modules/Portlets

Metadata Search Portlet

http://apptest.cul.columbia.edu:8082/navigator/portal/default-page.psml

This is the only component currently linked from the top level at papyri.info.

Image Portlet

The Image Portlet provides for the online viewing and limited manipulation of APIS-hosted images via the PN. This portlet is supported server-side by licensed proprietary software: eRez/FSI image server and Flash Plugin FSI. Prior development has taken advantage of institutional licenses at Columbia for these. Columbia's licenses will continue to cover production needs through July 2010. Development work at NYU may require access to separate licenses, and when NYU takes over production responsibilities in 2010, licenses will definitely be required.

Text Portlet

Experimental text search interface (not linked from main public page yet): http://appdev.cul.columbia.edu:8082/ddbdp-nav/search

Statistics regarding content indexed for same (also not public yet): http://appdev.cul.columbia.edu:8082/ddbdp-nav/stats

The Text Portlet (a.k.a. the DDbDP app) supports rendering and manipulation of DDbDP-originated papyrus texts in their original languages. No proprietary software is involved. The application takes advantage of SRU in its query interface. The back-end for search right now is actually based on receiving a valid CQL query ( http://www.loc.gov/standards/sru/specs/cql.html ) rather than the input from any particular HTML form per se. This means that the number of clauses is just a matter of the HTML having a way to accommodate the appropriate number of inputs.

Translation Portlet

??????? APIS translations. HGV translations.

Getting data into the PN

Workflow for refreshing data is still pretty much nonexistent, although automating it is an [IDP1] deliverable for June/July 2008. As of now, basically 3 huge XML files (apis data, hgv data, and a composite file, called 'aggregated' (http://epiduke.cch.kcl.ac.uk/aggregated/), that has a mapped subset of the previously mentioned) are indexed by Lucene. At some point in the near future will probably move to putting a harvestable interface to APIS data up instead of relying on the Big Xml File. In any case, that's the stuff that goes into the PN as of now.

Specific PN Software Development Tasks

Interface Assessment: "conduct usability testing in order to identify areas of possible improvement"
Enhancements to new (Greek) lemmatized and proximity searching
the refinement of combined metadata and text searching functionality
mechanisms for automatic ingestion of updated EpiDoc / UNICODE texts and metadata from the DDbDP and HGV
implementation of basic EpiDoc support for APIS translations
investigation and, if feasible, implementation of XML-based interoperability with non-APIS digital projects like the German Papyrus Portal and databases involved in Trismegistos

Documents:

In the APIS Phase 6 proposal (pdf), see:
- Section 5.1: Interface Assessment, PDF document pages 21-22
- Section 5.2.1: Papyrological Navigator (papyri.info), PDF document pages 22-23.
- Appendix D: APIS/Papyrological Navigator Work Plan Details, PDF document pages 39-45

Source code

Not yet imported into the IDP SVN repository.

Related Milestones

milestone:"PN Code Handover" (see details at CodeHandover)

Related tickets (query)

Download in other formats:

Plain Text