New locking semantics introduced to be able to cope with multithreaded processing. This mostly affects server-like apps integrating the indexer, not as much IDEs.
IndexContext new methods:
IteratorSearchResult/IteratorResultSet new methods:
Others:
Notes: if an app integrating Indexer Core did NOT tamper with indexingContexts directly (by using one of it's ctx.getIndexWriter(), ctx.getIndexReader() and ctx.getIndexSearcher()), it should almost not notice anything. The only change needed is to explicitly close non-consumed IteratorSearchResults. Otherwise, you have to adapt and manually fiddle with locking of contexts, or use proper API calls. In future, these methods and direct IndexingContext use within integrating applications will be highly discouraged!