Currently, VIVO depends on 5 different 3rd-party products to deal with data in JSON format. These are:
Glassfish parser (javax.json.Json)
SourceForge.net parser (net.sf.json)
JSON.org parser (org.json)
Graham notes that we must keep two of these, at least for now: For TPF, we have a dependency on GSON, but for Jena 3, that has a dependency on Jackson.
Tackle the parser usages in this order:
1) ORCiD client (which uses the Glassfish parser) - this should be the easiest to address. We also need to transition to vivo-project, but we can discuss the requirements of that separately.
2) Use of the sourceforge.net parser - this might be the largest amount of work, but it would also allow us to cut out a few other dependencies that we don’t otherwise need.
3) The embedded org.json parser source.