Thanks. This is for sure a reason to consider HK2 which we already use.
I struggle a bit with un-winding the current class-dependency graph of Solr to
figure out where it makes most sense to start a POC. I think perhaps injecting
singleton CoreContainer could be a nice start. Do you know of a to
+1 to experimenting with DI!
One thing that should probably be incorporated into the SIP and
resolved at some level is that Solr does have some "non-homegrown" DI
currently. Jersey, the JAX-RS framework we're using to serve our v2
APIs, allows some minimal DI in the resource classes that it
insta
+1 on this SIP and a POC would make sense. I think it could be
valuable to update the Motivation section with the anticipated value
provided by this SIP, i.e. reduced code maintenance, removing brittle
home-grown interfaces, etc. Most of my IOC/DI experience is with
Spring and Dagger seems like a b
This is the discuss thread for SIP-19
JIRA: https://issues.apache.org/jira/browse/SOLR-16998
SIP link:
https://cwiki.apache.org/confluence/display/SOLR/SIP-19+Adopt+JSR-330+dependency+injection
Use a standardized dependency injection in Solr instead of our home grown
@SolrCoreAware, @ResourceLo