Trippi RDF SPI implementation using SPARQL Update over HTTP

[Daniel Davis]

We have published trippi-sparql to Github https://github.com/Smithsonian/trippi-sparql with documentation on the Smithsonian Confluence site https://confluence.si.edu/display/RISSC/Trippi-SPARQL  This library was developed by Adam Soroka to fill a need for those sites that must stay on Fedora 3.x notably Islandora which we use.  This library permits you to replace Mulgara with a SPARQL 1.1 compliant semantic triplestore.  We are testing with JENA/Fuseki but it should work properly with many other products.  We also hope this will provide a bridge for subsequent development particularly with Fedora 4.

We plan to integrate this with Islandora 7.x.1.7 over the next two months.

Key Features

• Enables the use of any SPARQL 1.1 compliant triplestore (tested with Apache Jena/Fuseki)
• Supports all the features of Trippi used by Fedora 3. x (testing with Fedora 3.8.1)
  • Compliant with the Trippi Triplestore API
  • Query, Insert, and Delete
  • RISearch including its HTML pages (though we recommend using a UI like the one provided by Fuseki)
• Can be installed external to the Fedora WAR using any SPARQL 1.1 compliant HTTP endpoints (tested with Fuseki)
• Available as drop-in JAR and available pre-embedded into the Fedora WAR
• May be configured to use developer-specified graphs (or the default graph) providing support for full expressivity
  • Fedora 3.x prefers the conventional <#ri> to identify the graph it uses
  • Fedora 3.x also prefers the use of a set of convenience namespaces
• Fully supports the traditional Fedora 3. x rebuilder
• Published in Maven Central for source builds
• Apache 2 license

----

Daniel Davis
Technical Manager, Office of Research Information Services
Office of the CIO, Smithsonian Institution

[Melissa Anez]

This is really great, Dan! Any chance you could add it to Islandora Awesome for further discovery? https://github.com/Islandora-Labs/islandora_awesome

Thanks

- Melissa

[Nick Ruest]

Hi Dan,

Just want to take the time to thank you and Adam Soroka very much for
this work!

York University Digital Library replaced Mulgara with Blazegraph over
the weekend, and things are running a lot smoother now. 7.7 million
triples in Mulgara is a bad time :-(

One item to note, is depending on your setup, you might have to tweak
some settings in the fedora.fcfg if you run into an error like this
while running the stock indexer:
https://gist.github.com/ruebot/ad7979b5e09a07a2592bf632f366905f

We just dialled these[1] settings back a lot, and things went a lot more
smoothly.

Also worth noting, is the hot indexer[2] that Adam Soroka created, which
is an alternative to the stock indexer. This indexer went through 935k
objects in just over two hours, whereas the stock indexer brought
everything over in around 9hrs.

Andy Wagner, thanks for sharing the Tuque patches!

Thanks again everyone! This is the type of open source community
collaboration I love seeing :-)

-nruest

[1]
https://gist.github.com/ruebot/ee4dd1e47eaa06a1b9782b44f8170a1a#file-fedora-fcfg-L548-L559
[2] https://github.com/ajs6f/fcrepo3-rdf-extractor.git

On 2016-06-06 12:40 PM, Daniel Davis wrote:
> We have published trippi-sparql to Github
> https://github.com/Smithsonian/trippi-sparql with documentation on the
> Smithsonian Confluence
> site https://confluence.si.edu/display/RISSC/Trippi-SPARQL This library
> was developed by Adam Soroka to fill a need for those sites that must
> stay on Fedora 3.x notably Islandora which we use. This library permits
> you to replace Mulgara with a SPARQL 1.1 compliant semantic triplestore.
> We are testing with JENA/Fuseki but it should work properly with many
> other products. We also hope this will provide a bridge for subsequent
> development particularly with Fedora 4.
>
> We plan to integrate this with Islandora 7.x.1.7 over the next two months.
>
>
> Key Features
>

> * Enables the use of any SPARQL 1.1 compliant triplestore (tested
> with Apache Jena/Fuseki <https://jena.apache.org/>)
> * Supports all the features of Trippi used by Fedora 3. x (testing
> with Fedora 3.8.1)
> o Compliant with the Trippi Triplestore API
> o Query, Insert, and Delete
> o RISearch including its HTML pages (though we recommend using a

> UI like the one provided by Fuseki)

> * Can be installed external to the Fedora WAR using any SPARQL 1.1

> compliant HTTP endpoints (tested with Fuseki)

> * Available as drop-in JAR and available pre-embedded into the Fedora WAR
> * May be configured to use developer-specified graphs (or the default

> graph) providing support for full expressivity

> o Fedora 3.x prefers the conventional |<#ri>| to identify the
> graph it uses
> o Fedora 3.x also prefers the use of a set of convenience namespaces
> * Fully supports the traditional Fedora 3. x rebuilder
> * Published in Maven Central for source builds
> * Apache 2 license

>
>
> ----
>
> Daniel Davis Technical Manager, Office of Research Information
> Services Office of the CIO, Smithsonian Institution
>

> --
> For more information about using this group, please read our Listserv
> Guidelines: http://islandora.ca/content/welcome-islandora-listserv
> ---
> You received this message because you are subscribed to the Google
> Groups "islandora" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to islandora+...@googlegroups.com
> <mailto:islandora+...@googlegroups.com>.
> Visit this group at https://groups.google.com/group/islandora.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/islandora/6ab5b5a3-bbc0-46d9-8739-4f274db32953%40googlegroups.com
> <https://groups.google.com/d/msgid/islandora/6ab5b5a3-bbc0-46d9-8739-4f274db32953%40googlegroups.com?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout.

[soro...@gmail.com]

Thanks for the hat-tip, Nick!

I should note that the hot indexer to which Nick refers below is (1) also funded by the Smithsonian Institution, but (2) not yet officially released and neither directly supported nor guaranteed to be production-ready. Nick was generously lending his time to play the role of guinea pig to get it there. I hope there will be a not-too-far-in-the-future announcement in this channel and elsewhere when that tool is "ready for prime time".

---
A. Soroka
The University of Virginia Library

> To unsubscribe from this group and stop receiving emails from it, send an email to islandora+...@googlegroups.com.

> Visit this group at https://groups.google.com/group/islandora.

> To view this discussion on the web visit https://groups.google.com/d/msgid/islandora/e7b45e73-c937-e857-e6f9-64660cda892a%40gmail.com.