The simplest explanation for what you're seeing would be that the type system is somehow not loaded. The annotator requests a type object from the the type system, and doesn't check that it's not null. The type object is then handed to createFS(), which barfs on a null type. To verify, put a break point on line 165 of MarkupHandler. attributeType must not be null. If it's null, that probably means the type system is not loaded. Did you change the descriptors? Any error messages on startup, in the UIMA log?
(That code is using a mixture of JCas and plain old CAS, which I don't quite understand. That should have no bearing on your issue, however.) --Thilo On 2/16/2010 19:39, Roland Cornelissen wrote: > Hi, > >> version of the code , > I use > UIMA 2.30 > Tika Annotator from the UIMA-Annotator-addons package (2.3.0) > Tika 0.6 >> and send us the the full stack trace. > java.lang.NullPointerException > at org.apache.uima.cas.impl.CASImpl.createFS(CASImpl.java:474) > at > org.apache.uima.tika.MarkupHandler.populateCAS(MarkupHandler.java:165) > at > org.apache.uima.tika.TIKAWrapper.populateCASfromURL(TIKAWrapper.java:105) > at > org.apache.uima.tika.FileSystemCollectionReader.getNext(FileSystemCollectionReader.java:99) > at > org.apache.uima.collection.impl.cpm.engine.ArtifactProducer.readNext(ArtifactProducer.java:494) > at > org.apache.uima.collection.impl.cpm.engine.ArtifactProducer.run(ArtifactProducer.java:711) > java.lang.NullPointerException > at org.apache.uima.cas.impl.CASImpl.createFS(CASImpl.java:474) > at > org.apache.uima.tika.MarkupHandler.populateCAS(MarkupHandler.java:165) > at > org.apache.uima.tika.TIKAWrapper.populateCASfromURL(TIKAWrapper.java:105) > at > org.apache.uima.tika.FileSystemCollectionReader.getNext(FileSystemCollectionReader.java:99) > at > org.apache.uima.collection.impl.cpm.engine.ArtifactProducer.readNext(ArtifactProducer.java:494) > at > org.apache.uima.collection.impl.cpm.engine.ArtifactProducer.run(ArtifactProducer.java:711) > uima.tcas.DocumentAnnotation [Dolphin] AdditionalInfo=31 SortOrder=1 > Timestamp=2010,2,16,13,42,48 ViewMode=1 > org.apache.uima.tika.MarkupAnnotation [Dolphin] AdditionalInfo=31 > SortOrder=1 Timestamp=2010,2,16,13,42,48 ViewMode=1 > org.apache.uima.tika.MarkupAnnotation > org.apache.uima.tika.SourceDocumentAnnotation > org.apache.uima.tika.MarkupAnnotation > org.apache.uima.tika.MarkupAnnotation [Dolphin] AdditionalInfo=31 > SortOrder=1 Timestamp=2010,2,16,13,42,48 ViewMode=1 > > > I have simple testsetup where output is writen to an annotation writer: in > this case Tika reads 3 html pages, errors on the first 2 and passes the > annotations from the last (?). Last lines of the stack trace are the printed > annotations. > > I hope this is better info. > > Roland > >
