Hi,

On Dec 7, 2007 10:50 AM, Chavan, Girish <[EMAIL PROTECTED]> wrote:
> Hi All,
>
> I am a newbie with UIMA, and some of my terminologies might be a bit off
> here. Please bear with me.

Looks to me like you got the terminology right.

> Here's my question:
> I am going through the code in 'DocumentAnalyzer.java' and I see that it
> creates an Aggregate AE which contains the user specified AE and a CAS
> consumer to the end of it. However, the CPM also seems to have a
> addCasConsumer() method. Why would you not add the CasConsumer to the
> CPM instead of tagging it at the end of the user's AE? The comment talks
> about how this is done to support a user AE which might have a CAS
> Multiplier.
> When a CasConsumer is added to a CPM, does it only see CASes that were
> output by the CollectionReader?, and not all CASes coming out of the
> user's AE?
>

Yes, that's right.  It's currently a limitation that a CasConsumer
won't see CASes that are created by a CasMultiplier within the user's
AE.  This is explained here:
http://incubator.apache.org/uima/downloads/releaseDocs/2.2.0-incubating/docs/html/tutorials_and_users_guides/tutorials_and_users_guides.html#ugr.tug.cm.using_cm_in_cpe

If you need a CAS Consumer that processes CASes created within an
aggregate AnalysisEngine, then you need to add the CasMultiplier to
that aggregate AnalysisEngine, and not to the CPE.

-Adam

Reply via email to