Hi Peter,
Thanks for the information. So as I understand it from your post the
following 2 pipelines are equivalent:
a)

   public static AnalysisEngineDescription getTokenProcessingPipeline()
throws ResourceInitializationException, MalformedURLException {

      AggregateBuilder builder = new AggregateBuilder();

      builder.add( SimpleSegmentAnnotator.createAnnotatorDescription() );

      builder.add( SentenceDetector.createAnnotatorDescription() );

      builder.add(ClearNLPDependencyParserAE.createAnnotatorDescription());

      builder.add(
ContextDependentTokenizerAnnotator.createAnnotatorDescription() );

      builder.add( POSTagger.createAnnotatorDescription() );

      return builder.createAggregateDescription();

   }
b)

   public static AnalysisEngineDescription getTokenProcessingPipeline()
throws ResourceInitializationException, MalformedURLException {

      AggregateBuilder builder = new AggregateBuilder();

      builder.add( SimpleSegmentAnnotator.createAnnotatorDescription() );

      builder.add( SentenceDetector.createAnnotatorDescription() );

      builder.add( TokenizerAnnotatorPTB.createAnnotatorDescription() );

      builder.add( LvgAnnotator.createAnnotatorDescription() );

      builder.add(
ContextDependentTokenizerAnnotator.createAnnotatorDescription() );

      builder.add( POSTagger.createAnnotatorDescription() );

      return builder.createAggregateDescription();

   }


The difference between the 2 is that a) has


      builder.add(ClearNLPDependencyParserAE.createAnnotatorDescription());



      instead of


      builder.add( TokenizerAnnotatorPTB.createAnnotatorDescription() );

      builder.add( LvgAnnotator.createAnnotatorDescription() );


      in b)


Is that correctly understood?

On Thu, Aug 19, 2021 at 7:53 PM Peter Abramowitsch <pabramowit...@gmail.com>
wrote:

> RE the Dependency Parser, it is part of the release.
> Just use the PiperCreator app and look for it in the list of modules.  It
> will tell you it's prerequisites, and therefore where in your Piper file it
> belongs.
>
> addDescription ClearNLPDependencyParserAE
>
> - Peter
>
> On Thu, Aug 19, 2021 at 2:39 AM Benjamin hansen <
> benjaminkakke...@gmail.com>
> wrote:
>
> > Thanks for the input.
> > So I am trying to use
> > the ClinicalPipelineFactory.getTokenProcessingPipeline which includes the
> > LVG which seems to be having the issue. But you say i can just replace
> the
> > LVG part with the DependencyParser in that pipeline? Can you elaborate a
> > bit on where I can find the DependencyParser?
> >
> >
> > On Wed, Aug 18, 2021 at 5:56 PM Peter Abramowitsch <
> > pabramowit...@gmail.com>
> > wrote:
> >
> > > Hi Benjamin,
> > >
> > > If what you're looking for are the lemmas of tokens, just use the
> > > DependencyParser instead of LVG
> > > I had another problem with LVG as well, but I think that it might be
> > simply
> > > that all the needed resources are not being copied into the right
> place.
> > > This was done without the lvg
> > >
> > > {
> > >       "_type": "ConllDependencyNode",
> > >       "sofa": 1,
> > >       "begin": 3,
> > >       "end": 10,
> > >       "id": 2,
> > >       "form": "decided",
> > >       "lemma": "decide",
> > >       "cpostag": "VBD",
> > >       "postag": "VBD",
> > >       "feats": "_",
> > >       "head": 137,
> > >       "deprel": "root",
> > >       "pdeprel": "_"
> > >     }
> > >
> > > On Wed, Aug 18, 2021 at 7:29 AM Finan, Sean <
> > > sean.fi...@childrens.harvard.edu> wrote:
> > >
> > > > Hi Benjamin,
> > > >
> > > > My first question is: what pipeline are you trying to run?
> > > >
> > > > My second question is: Do you really need to use LVG?
> > > >
> > > > Sean
> > > > ________________________________________
> > > > From: Benjamin hansen <benjaminkakke...@gmail.com>
> > > > Sent: Wednesday, August 18, 2021 3:07 AM
> > > > To: dev@ctakes.apache.org
> > > > Subject: An exception occured while executing the Java class. URI is
> > not
> > > > hierarchical [EXTERNAL]
> > > >
> > > > * External Email - Caution *
> > > >
> > > >
> > > > While working at a simple pipeline example I got this error:
> > > >
> > > > *java.lang.IllegalArgumentException*: *URI is not hierarchical*
> > > >
> > > >     *at* java.io.File.<init> (*File.java:420*)
> > > >
> > > >     *at* org.apache.ctakes.lvg.resource.LvgCmdApiResourceImpl.load (
> > > > *LvgCmdApiResourceImpl.java:65*)
> > > >
> > > >
> > > > I found that this issue has already been reported 4 years ago here
> > > >
> > > >
> > >
> >
> https://urldefense.com/v3/__https://issues.apache.org/jira/browse/CTAKES-445__;!!NZvER7FxgEiBAiR_!9HBmXkq30TUdwnSpuHc8_7iEVkoMAiJ3p_rSTXE5d90TARHEdioOMNukOUaL6eB5CboTRBMsOYI$
> > > >
> > > >
> > > > I am on MacOS which the workaround patched proposed in that thread
> does
> > > not
> > > > fix... And like the last comment in the thread says - the patch
> likely
> > > also
> > > > does not work on linux.
> > > >
> > > >
> > > > This seems to be quite a serious bug since both mac and linux would
> be
> > > > serious development and production platforms for ctakes users.
> > > >
> > > >
> > > > Is there no fix for this after 4 years?
> > > >
> > >
> >
>

Reply via email to