On 28 Nov 2003, at 10:47, Martin Holz wrote:
You have access to the content, but must use getContent or getContentBytes as Martin described. These stream input into arrays for multiple access. This certainly can become a problem when content is BIG. But, what else could one do?
You could use the stream and set a new one in the interceptor. Its mostly a documentation issue for pre interceptors. post interceptors are more of a problem.
What about a "transation" interceptor? In that case, we could hook the lucene indexer on the "transaction commit" event, then list that URI for asynchronous indexing (so that the interceptor isn't blocked by the lucene indexer which could take a while).
Thoughts?
-- Stefano.
smime.p7s
Description: S/MIME cryptographic signature
