[ 
https://issues.apache.org/jira/browse/OPENNLP-1369?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17799335#comment-17799335
 ] 

Martin Wiesner commented on OPENNLP-1369:
-----------------------------------------

As reported "generator.featuregen" is missing in the serialized 
pt-pos-perceptron.bin file. Stack trace looks like this:
 
{quote}java.lang.NullPointerException: Cannot read the array length because "b" 
is null
 
at java.base/java.io.FilterOutputStream.write(FilterOutputStream.java:108)
at 
opennlp.tools.util.model.ByteArraySerializer.serialize(ByteArraySerializer.java:36)
at 
opennlp.tools.util.model.ByteArraySerializer.serialize(ByteArraySerializer.java:27)
at opennlp.tools.util.model.BaseModel.serialize(BaseModel.java:615)
at 
opennlp.tools.util.model.POSModelSerializer.serialize(POSModelSerializer.java:62)
at 
opennlp.tools.util.model.POSModelSerializer.serialize(POSModelSerializer.java:33)
at opennlp.tools.util.model.BaseModel.serialize(BaseModel.java:615)
at 
opennlp.tools.namefind.TokenNameFinderModelTest.testNERWithPOSModel(TokenNameFinderModelTest.java:102){quote}

> NPE when serializing a TokenNameFinder model trained with 
> POSTaggerNameFeatureGeneratorFactory
> ----------------------------------------------------------------------------------------------
>
>                 Key: OPENNLP-1369
>                 URL: https://issues.apache.org/jira/browse/OPENNLP-1369
>             Project: OpenNLP
>          Issue Type: Bug
>    Affects Versions: 2.0.0, 2.1.0, 2.2.0, 2.3.0
>            Reporter: Lucas Avanço
>            Priority: Major
>
> There is an issue of null pointer when trying to serialize a TokenNameFinder 
> model trained using as one of feature generators 
> POSTaggerNameFeatureGeneratorFactory.
> In this method  
> opennlp.tools.util.model.{*}BaseModel#serialize{*}(java.io.OutputStream) 
> there is a for which iterates over the artifactMap of the POS-model that must 
> be serialized together, but a NPE occurs because *generator.featuregen* is 
> not found inside of *pt-pos-perceptron.zip*
>  
> <generator 
> class="opennlp.tools.util.featuregen.WindowFeatureGeneratorFactory">
>     <int name="prevLength">2</int>
>     <int name="nextLength">2</int>
>     <generator 
> class="opennlp.tools.util.featuregen.POSTaggerNameFeatureGeneratorFactory">
>     <str name="model">pt-pos-perceptron.zip</str>
>     </generator>
>   </generator>
>  
> This is the POS-tagger model used:
> http://opennlp.sourceforge.net/models-1.5/pt-pos-perceptron.bin
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to