Thanks. Here are some more specific tips:
You can specify all engines in the call to runPipeline - no need for the
AggregateBuilder
unless you need to do sofa mappings.
SimplePipeline.runPipeline(reader, preparationEngine, whitespaceEngine,
casConsumer));
Parameter constants typically begin with "PARAM_" instead of ending in
"_PARAM". That makes a difference if you ever plan to use the
uimafit-maven-plugin to automatically generate descriptors from your AEs,
because it uses prefixes to detect parameter name constants.
uimaFIT should be able to automatically coerce single values into multi-valued
parameters. So it should be possible to write this
AnalysisEngineFactory.createEngineDescription(WhitespaceTokenizer.class,
"SofaNames", SimpleParserAE.SOFA_NAME_TEXT_ONLY);
Cheers,
-- Richard
On 23.01.2014, at 14:45, Luca Foppiano <[email protected]> wrote:
> On Thu, Jan 23, 2014 at 3:13 PM, Richard Eckart de Castilho
> <[email protected]>wrote:
>
>> Hi,
>>
>> Hi Richard,
>
>
>> can you provide the full code for your sample pipeline? I think that would
>> make it easier to help.
>>
>
> Sure, is located here: https://github.com/lfoppiano/uima-fit-sample-pipeline
>
>
>> With the present information, I can only give some general advice.
>>
>> [...]
>
>>
>> I would recommend using the CAS/CasUtil only if you want to implement a
>> generic component that can be configured to work with different types. If
>> your component is fixed to a certain type system, then using the
>> JCas/JCasUtil is much more convenient.
>>
>
> Thanks a lot for your input, in fact it shed some light1 around type
> systems.
>
> Regards
> --
> Luca Foppiano
>
> Software Engineer
> +31615253280
> [email protected]
> www.foppiano.org
