I believe the JCasIterable is currently implemented as a loop which calls "process" on the analysis engines for every CAS produced by the reader and then returns the corresponding CAS. This wouldn't work with multipliers.
Can you please file an issue in the Apache Jira, preferrably with a minimal test case attached. It shouldn't be a big problem to fix this for the next release. A patch already fixing this would also work, of course ;) Cheers, -- Richard On 22.11.2013, at 08:01, Swirl <lriwsw...@gmail.com> wrote: > I have successfully used CasMultiplier to spilt up a document into segments > for further processing using SimplePipeline.runPipeline(). > I did this by wrapping the CasMultiplier and the succeeding Annotator within > a > aggregate. > > But by simply changing the usage of SimplePipeline.runPipeline() to using > JCasIterable. The code no longer runs correctly, i.e., it's returning as CAS > only the number of physical documents, instead of the segments that i > expected. > > How can I can CasMultiplier to work with a JCasIterable?
