If there are interesting use cases for this I would like to know more about it.
Technically it doesn't seem hard to implement, but would you not use
the native IndexSearcher anyway to make use of this? I mean, would
there be a practical use case to set the collector and then let the
search results be handled by Hibernate Search, or are you going to
handle the collected results with custom code anyway?
Also we have plans to set a custom collector to perform some nice
performance optimization tricks: I was thinking to leverage the
advantage of FieldCaches by default, so you would loose this advantage
by setting a custom one if yours doesn't use the same caches.
2010/11/27 Amin Mohammed-Coleman <aminmc(a)gmail.com>:
Apologies up front if this isn't the right place for this. Are there any plans to
allow Collectors to be used within Hibernate Search? I had an interesting requirement
which meant I couldn't use pagination and as result I had to use a custom collector
using IndexSearcher. My use case essentially meant I had to do a search and collect some
specific data for each matching document and return the entire result set to the next
process in the workflow.
As users have direct access to the underlying directory I guess using native lucene for
the use of collectors is ok but I guess it would be nice to essentially set an optional
collector in the FullTextQuery.
Again apologies up front if this is implemented already or I should post my questions on
hibernate-dev mailing list