When I started writing conciliator, I focused solely on implementing the main URL endpoint for reconciliation, since that’s what I needed at the time. There are actually several APIs that work together in OpenRefine: the Preview, Suggest, and recently added Data Extension APIs provide functionality that complements Reconciliation. And there may be more in the future as OpenRefine continues to evolve.
My code didn’t extend very easily, so I’ve rewritten a ton of stuff. This is currently in the master branch and it’s running on http://refine.codefork.com. I will release 3.0 probably sometime in January.
Most users don’t need to do anything. Out of the box, it should work as the previous version did. If you modified the conciliator.properties file, you should look at the changes in that file.
Here are the changes under the hood so far:
– Spring controllers and components are now used more effectively for better separation of concerns, less manual plumbing, better extensibility and maintainability.
– The classes representing data in/out for the various APIs are more fully fleshed out.
– The conciliator.properties file allows for less configurability than before, but I don’t know how useful that ever was, really.
– Tests have been rewritten for more “real world” coverage.
– Custom cache code has been replaced with Ehcache.
– Requires Java 8.
With this scaffolding in place, I can start actually implementing Data Extension for specific data sources. I plan to start with OpenLibrary as it provides the richest data.
Thanks to everyone who contributed bug reports and suggestions! I’m amazed and gratified that this software gets used as much as it does.