9.4

Edit

ML SDK-related questions

A local extract runner failed on NonSequenceBaseAnnotator with NumberFromatException.

Symptoms

A local extract failed with an exception like:

Caused by: org.apache.uima.resource.ResourceInitializationException: Initialization of annotator class "com.workfusion.nlp.uima.annotator.NonSequenceBasedFeatureExtractionAnnotator" failed.  
...
Caused by: java.lang.NumberFormatException: \uddha
    at com.google.gson.stream.JsonReader.readEscapeCharacter(JsonReader.java:1527)

Possible causes

The path to the model inside the strat classpath from the folder with the name starting from the u letter.

During extraction, the absolute path to the model is created using the first \ symbol, and the JSON parser will recognize \u\[first five symbols\] as an escaped special Unicode symbol.

Solution

  1. Check that the model extract runner has all / slash symbols in the path variables, not \.
  2. If Step 1 is passed and you still have an error, try to put your model and documents into another folder that does not start from u, for example, c://model/.
← PreviousNext →