Package org.apache.uima.analysis_engine

Examples of org.apache.uima.analysis_engine.AnalysisEngine


   * @throws ResourceInitializationException
   *
   */
  public static AnalysisEngine getAE(String filePath) throws IOException, InvalidXMLException,
          ResourceInitializationException {
    AnalysisEngine ae = null;
    // get Resource Specifier from XML file
    XMLInputSource in = new XMLInputSource(filePath);
    ResourceSpecifier specifier = UIMAFramework.getXMLParser().parseResourceSpecifier(in);

    // create AE here
View Full Code Here


   * @param parameterSettings
   * @return
   * @throws ResourceInitializationException
   */
  public static AnalysisEngine getAE(String xmlPath, Map<String, Object> parameterSettings) throws ResourceInitializationException {
    AnalysisEngine ae = null;
    try {
      XMLInputSource in = new XMLInputSource(xmlPath);

      // override descriptor's configuration parameters
      AnalysisEngineDescription desc = UIMAFramework.getXMLParser().parseAnalysisEngineDescription(in);
View Full Code Here

   * @param implementationName
   * @return
   * @throws ResourceInitializationException
   */
  public static AnalysisEngine getAEWithMockedImplementation(String xmlPath, String implementationName) throws ResourceInitializationException {
    AnalysisEngine ae = null;
    try {
      XMLInputSource in = new XMLInputSource(xmlPath);

      AnalysisEngineDescription desc = UIMAFramework.getXMLParser().parseAnalysisEngineDescription(in);
      desc.setAnnotatorImplementationName(implementationName);
View Full Code Here

        catch (InvalidXMLException e)
        {
            throw new InvocationTargetException(e);
        }
       
        AnalysisEngine textAnalysisEngine;
        try
        {
            textAnalysisEngine = UIMAFramework
                    .produceAnalysisEngine(textAnalysisEngineSpecifier);
        }
View Full Code Here

      outFile.delete();
     
      //test aggregate that uses ParallelStep
      AnalysisEngineDescription desc = UIMAFramework.getXMLParser().parseAnalysisEngineDescription(
        new XMLInputSource(JUnitExtension.getFile("TextAnalysisEngineImplTest/AggregateForParallelStepTest.xml")));
      AnalysisEngine ae = UIMAFramework.produceAnalysisEngine(desc);
      CAS cas = ae.newCAS();
      cas.setDocumentText("new test");
      ae.process(cas);
      assertEquals("new test", TestAnnotator.lastDocument);
      assertEquals("new test", TestAnnotator2.lastDocument);
      cas.reset();
     
    } catch (Exception e) {
View Full Code Here

   *
   * @param aTaeDesc
   *          description of TextAnalysisEngine to test
   */
  protected void _testProcess(AnalysisEngineDescription aTaeDesc) throws UIMAException {
    AnalysisEngine ae = UIMAFramework.produceAnalysisEngine(aTaeDesc);
    CAS tcas = ae.newCAS();

    // process(CAS,ResultSpecification)
    ResultSpecification resultSpec = new ResultSpecification_impl(tcas.getTypeSystem());
    resultSpec.addResultType("NamedEntity", true);

View Full Code Here

    _testProcessInner(ae, tcas, resultSpec, resultSpec);
  }
 
  protected void _testProcess(AnalysisEngineDescription aTaeDesc, String[] languages) throws UIMAException {
    AnalysisEngine ae = UIMAFramework.produceAnalysisEngine(aTaeDesc);
    CAS tcas = ae.newCAS();

    // process(CAS,ResultSpecification)
    ResultSpecification resultSpec = new ResultSpecification_impl(tcas.getTypeSystem());
    resultSpec.addResultType("NamedEntity", true);   // includes subtypes Person, Sentence, Place, Paragraph
                                                     // sets for lang = x-unspecified
View Full Code Here

      aggAe.collectionProcessComplete(new ProcessTrace_impl());
     
      //test that fixedFlow order is used
      File descFile = JUnitExtension.getFile("TextAnalysisEngineImplTest/AggregateForCollectionProcessCompleteTest.xml");
      AnalysisEngineDescription cpcTestDesc = UIMAFramework.getXMLParser().parseAnalysisEngineDescription(new XMLInputSource(descFile));
      AnalysisEngine cpcTestAe = UIMAFramework.produceAnalysisEngine(cpcTestDesc);
      cpcTestAe.collectionProcessComplete();
      assertEquals("One", AnnotatorForCollectionProcessCompleteTest.lastValue);
    } catch (Exception e) {
      JUnitExtension.handleException(e);
    }
  }
View Full Code Here

      // primitive
      AnalysisEngineDescription segmenterDesc = UIMAFramework.getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(JUnitExtension
                              .getFile("TextAnalysisEngineImplTest/NewlineSegmenter.xml")));
      AnalysisEngine ae = UIMAFramework.produceAnalysisEngine(segmenterDesc);
      CAS cas = ae.newCAS();
      cas.setDocumentText("Line one\nLine two\nLine three");
      CasIterator iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      CAS outCas = iter.next();
      assertEquals("Line one", outCas.getDocumentText());
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line two", outCas.getDocumentText());
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line three", outCas.getDocumentText());
      outCas.release();
      assertFalse(iter.hasNext());

      // aggregate
      AnalysisEngineDescription aggSegDesc = UIMAFramework.getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(JUnitExtension
                              .getFile("TextAnalysisEngineImplTest/AggregateWithSegmenter.xml")));
      ae = UIMAFramework.produceAnalysisEngine(aggSegDesc);
      cas = ae.newCAS();
      cas.setDocumentText("Line one\nLine two\nLine three");
      iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line one", outCas.getDocumentText());
      assertEquals("Line one", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line two", outCas.getDocumentText());
      assertEquals("Line two", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line three", outCas.getDocumentText());
      assertEquals("Line three", TestAnnotator.lastDocument);
      outCas.release();
      assertFalse(iter.hasNext());
      // Annotator should NOT get the original CAS according to the default flow
      assertEquals("Line three", TestAnnotator.lastDocument);

      // nested aggregate
      AnalysisEngineDescription nestedAggSegDesc = UIMAFramework
              .getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(
                              JUnitExtension
                                      .getFile("TextAnalysisEngineImplTest/AggregateContainingAggregateSegmenter.xml")));
      ae = UIMAFramework.produceAnalysisEngine(nestedAggSegDesc);
      cas = ae.newCAS();
      cas.setDocumentText("Line one\nLine two\nLine three");
      iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line one", outCas.getDocumentText());
      assertEquals("Line one", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line two", outCas.getDocumentText());
      assertEquals("Line two", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line three", outCas.getDocumentText());
      assertEquals("Line three", TestAnnotator.lastDocument);
      outCas.release();
      assertFalse(iter.hasNext());
      // Annotator should NOT get the original CAS according to the default flow
      assertEquals("Line three", TestAnnotator.lastDocument);

      // two segmenters
      AnalysisEngineDescription twoSegDesc = UIMAFramework.getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(JUnitExtension
                              .getFile("TextAnalysisEngineImplTest/AggregateWith2Segmenters.xml")));
      ae = UIMAFramework.produceAnalysisEngine(twoSegDesc);
      cas = ae.newCAS();
      cas.setDocumentText("One\tTwo\nThree\tFour");
      iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("One", outCas.getDocumentText());
      assertEquals("One", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Two", outCas.getDocumentText());
      assertEquals("Two", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Three", outCas.getDocumentText());
      assertEquals("Three", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Four", outCas.getDocumentText());
      assertEquals("Four", TestAnnotator.lastDocument);
      outCas.release();
      assertFalse(iter.hasNext());
      // Annotator should NOT get the original CAS according to the default flow
      assertEquals("Four", TestAnnotator.lastDocument);

      // dropping segments
      aggSegDesc = UIMAFramework.getXMLParser().parseAnalysisEngineDescription(
              new XMLInputSource(JUnitExtension
                      .getFile("TextAnalysisEngineImplTest/AggregateSegmenterForDropTest.xml")));
      ae = UIMAFramework.produceAnalysisEngine(aggSegDesc);
      cas = ae.newCAS();
      cas.setDocumentText("Line one\nDROP\nLine two\nDROP\nLine three");
      // results should be the same as the first aggregate segmenter test.
      // segmetns whose text is DROP should not be output.
      iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line one", outCas.getDocumentText());
      assertEquals("Line one", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line two", outCas.getDocumentText());
      assertEquals("Line two", TestAnnotator.lastDocument);
      outCas.release();
      assertTrue(iter.hasNext());
      outCas = iter.next();
      assertEquals("Line three", outCas.getDocumentText());
      assertEquals("Line three", TestAnnotator.lastDocument);
      outCas.release();
      assertFalse(iter.hasNext());
      // Annotator should NOT get the original CAS according to the default flow
      assertEquals("Line three", TestAnnotator.lastDocument);
     
      //with ParallelStep
      AnalysisEngineDescription desc = UIMAFramework.getXMLParser().parseAnalysisEngineDescription(
        new XMLInputSource(JUnitExtension.getFile("TextAnalysisEngineImplTest/AggregateForParallelStepCasMultiplierTest.xml")));
      ae = UIMAFramework.produceAnalysisEngine(desc);
      cas.reset();
      cas.setDocumentText("One\tTwo\nThree\tFour");
      iter = ae.processAndOutputNewCASes(cas);
      Set<String> expectedOutputs = new HashSet<String>();
      expectedOutputs.add("One");
      expectedOutputs.add("Two\nThree");
      expectedOutputs.add("Four");
      expectedOutputs.add("One\tTwo");
      expectedOutputs.add("Three\tFour");
      while (iter.hasNext()) {
        outCas = iter.next();
        assertTrue(expectedOutputs.remove(outCas.getDocumentText()));       
        outCas.release();
      }
      assertTrue(expectedOutputs.isEmpty());

     
      // test aggregate with 2 AEs sharing resource manager
      AnalysisEngineDescription aggregateSegDesc = UIMAFramework.getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(JUnitExtension
                              .getFile("TextAnalysisEngineImplTest/AggregateWithSegmenter.xml")));
     
      ResourceManager rsrcMgr = UIMAFramework.newDefaultResourceManager();
      Map<String, Object> params = new HashMap<String, Object>();
      AnalysisEngine ae1 = UIMAFramework.produceAnalysisEngine(aggregateSegDesc, rsrcMgr, params);
      AnalysisEngine ae2 = UIMAFramework.produceAnalysisEngine(aggregateSegDesc, rsrcMgr, params);
     
      // start with testing first ae
      CAS cas1 = ae1.newCAS();
      cas1.setDocumentText("Line one\nLine two\nLine three");
      CasIterator iter1 = ae1.processAndOutputNewCASes(cas1);
      assertTrue(iter1.hasNext());
      CAS outCas1 = iter1.next();
      assertEquals("Line one", outCas1.getDocumentText());
    
      // now test second ae
      CAS cas2 = ae2.newCAS();
      cas2.setDocumentText("Line one\nLine two\nLine three");
      CasIterator iter2 = ae2.processAndOutputNewCASes(cas2);
      assertTrue(iter2.hasNext());
      CAS outCas2 = iter2.next();
      assertEquals("Line one", outCas2.getDocumentText());
      outCas2.release();
      assertTrue(iter2.hasNext());
View Full Code Here

              .getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(
                              JUnitExtension
                                      .getFile("TextAnalysisEngineImplTest/AggregateWithSegmenterForErrorTest.xml")));
      AnalysisEngine ae = UIMAFramework.produceAnalysisEngine(aggSegDesc);
     
      CAS cas = ae.newCAS();
      for (int i = 0; i < 2; i++) // verify we can do this more than once
      {
        FlowControllerForErrorTest.reset();
        cas.setDocumentText("Line one\nLine two\nERROR");
        CasIterator iter = ae.processAndOutputNewCASes(cas);
        assertTrue(iter.hasNext());
        CAS outCas = iter.next();
        assertEquals("Line one", outCas.getDocumentText());
        outCas.release();
        assertTrue(iter.hasNext());
        outCas = iter.next();
        assertEquals("Line two", outCas.getDocumentText());
        outCas.release();
        try {
          UIMAFramework.getLogger().setLevel(Level.OFF)// Suppress logging of expected exception
          assertTrue(iter.hasNext());
          outCas = iter.next();
          fail(); // the above should throw an exception
        } catch (AnalysisEngineProcessException e) {
          UIMAFramework.getLogger().setLevel(Level.INFO); // Restore to apparent default of INFO
        }
        //check that FlowController was notified twice, once for the
        //segment's flow and once for the complete document's flow
        assertEquals(2, FlowControllerForErrorTest.abortedDocuments.size());
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("ERROR"));
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("Line one\nLine two\nERROR"));

        cas.reset();
      }

      // nested aggregate
      AnalysisEngineDescription nestedAggSegDesc = UIMAFramework
              .getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(
                              JUnitExtension
                                      .getFile("TextAnalysisEngineImplTest/NestedAggregateSegmenterForErrorTest.xml")));
      ae = UIMAFramework.produceAnalysisEngine(nestedAggSegDesc);
      cas = ae.newCAS();
      for (int i = 0; i < 2; i++) // verify we can do this more than once
      {
        FlowControllerForErrorTest.reset();
        cas.setDocumentText("Line one\nLine two\nERROR");
        CasIterator iter = ae.processAndOutputNewCASes(cas);
        assertTrue(iter.hasNext());
        CAS outCas = iter.next();
        assertEquals("Line one", outCas.getDocumentText());
        outCas.release();
        assertTrue(iter.hasNext());
        outCas = iter.next();
        assertEquals("Line two", outCas.getDocumentText());
        outCas.release();
        try {
          UIMAFramework.getLogger().setLevel(Level.OFF)// Suppress logging of expected exception
          assertTrue(iter.hasNext());
          outCas = iter.next();
          fail(); // the above should throw an exception
        } catch (AnalysisEngineProcessException e) {
          UIMAFramework.getLogger().setLevel(Level.INFO); // Restore to apparent default of INFO
        }
        //check that FlowController was notified three times, once for the
        //segment's flow and twice for the complete document's flow (once
        //in each aggregate)
        assertEquals(3, FlowControllerForErrorTest.abortedDocuments.size());
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("ERROR"));
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("Line one\nLine two\nERROR"));
        FlowControllerForErrorTest.abortedDocuments.remove("Line one\nLine two\nERROR");
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("Line one\nLine two\nERROR"));
       
        cas.reset();
      }

      // 2 segmenters
      AnalysisEngineDescription twoSegDesc = UIMAFramework
              .getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(
                              JUnitExtension
                                      .getFile("TextAnalysisEngineImplTest/AggregateWith2SegmentersForErrorTest.xml")));
      ae = UIMAFramework.produceAnalysisEngine(twoSegDesc);
      cas = ae.newCAS();
      for (int i = 0; i < 2; i++) // verify we can do this more than once
      {
        FlowControllerForErrorTest.abortedDocuments.clear();
        cas.setDocumentText("One\tTwo\nThree\tERROR");
        CasIterator iter = ae.processAndOutputNewCASes(cas);
        assertTrue(iter.hasNext());
        CAS outCas = iter.next();
        assertEquals("One", outCas.getDocumentText());
        outCas.release();
        assertTrue(iter.hasNext());
        outCas = iter.next();
        assertEquals("Two", outCas.getDocumentText());
        outCas.release();
        assertTrue(iter.hasNext());
        outCas = iter.next();
        assertEquals("Three", outCas.getDocumentText());
        outCas.release();
        try {
          UIMAFramework.getLogger().setLevel(Level.OFF)// Suppress logging of expected exception
          assertTrue(iter.hasNext());
          outCas = iter.next();
          fail(); // the above should throw an exception
        } catch (AnalysisEngineProcessException e) {
          UIMAFramework.getLogger().setLevel(Level.INFO); // Restore to apparent default of INFO
        }
        //check that FlowController was notified three times, once for each level of granularity
        assertEquals(3, FlowControllerForErrorTest.abortedDocuments.size());
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("ERROR"));
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("Three\tERROR"));
        assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("One\tTwo\nThree\tERROR"));
       
        cas.reset();
      }

      // segmenter that requests too many CASes
      AnalysisEngineDescription segmenterDesc = UIMAFramework.getXMLParser()
              .parseAnalysisEngineDescription(
                      new XMLInputSource(JUnitExtension
                              .getFile("TextAnalysisEngineImplTest/BadSegmenter.xml")));
      ae = UIMAFramework.produceAnalysisEngine(segmenterDesc);
      cas = ae.newCAS();
      cas.setDocumentText("Line one\nLine two\nLine three");
      CasIterator iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      CAS outCas = iter.next(); // first call OK
      outCas.release();
      assertTrue(iter.hasNext());
      // next call should fail with AnalysisEngineProcessException
      try {
        UIMAFramework.getLogger().setLevel(Level.OFF)// Suppress logging of expected exception
        iter.next();
        fail(); // should not get here
      } catch (AnalysisEngineProcessException e) {
        UIMAFramework.getLogger().setLevel(Level.INFO); // Restore to apparent default of INFO
      }
     
      // bad segmenter in an aggregate
      AnalysisEngineDescription aggWithBadSegmenterDesc = UIMAFramework.getXMLParser()
      .parseAnalysisEngineDescription(
              new XMLInputSource(JUnitExtension
                      .getFile("TextAnalysisEngineImplTest/AggregateWithBadSegmenterForErrorTest.xml")));
      ae = UIMAFramework.produceAnalysisEngine(aggWithBadSegmenterDesc);
      FlowControllerForErrorTest.reset();
      cas = ae.newCAS();
      cas.setDocumentText("Line one\nLine two\nLine three");
      iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      outCas = iter.next(); // first call OK
      outCas.release();
      assertTrue(FlowControllerForErrorTest.abortedDocuments.isEmpty());
      assertTrue(FlowControllerForErrorTest.failedAEs.isEmpty());
      // next call should fail with AnalysisEngineProcessException
      try {
        UIMAFramework.getLogger().setLevel(Level.OFF)// Suppress logging of expected exception
        if (iter.hasNext()) {
          iter.next();
        }
        fail(); // should not get here
      } catch (AnalysisEngineProcessException e) {
        UIMAFramework.getLogger().setLevel(Level.INFO); // Restore to apparent default of INFO
      }
      assertEquals(1, FlowControllerForErrorTest.abortedDocuments.size());
      assertTrue(FlowControllerForErrorTest.abortedDocuments.contains("Line one\nLine two\nLine three"));
      assertEquals(1,FlowControllerForErrorTest.failedAEs.size());
      assertTrue(FlowControllerForErrorTest.failedAEs.contains("Segmenter"));

      //configure AE to continue after error
      ae = UIMAFramework.produceAnalysisEngine(aggWithBadSegmenterDesc);
      ae.setConfigParameterValue("ContinueOnFailure", Boolean.TRUE);
      ae.reconfigure();
      FlowControllerForErrorTest.reset();

      cas.reset();
      cas.setDocumentText("Line one\nLine two\nLine three");
      iter = ae.processAndOutputNewCASes(cas);
      assertTrue(iter.hasNext());
      outCas = iter.next(); // first call OK
      outCas.release();
      assertTrue(FlowControllerForErrorTest.abortedDocuments.isEmpty());
      assertTrue(FlowControllerForErrorTest.failedAEs.isEmpty());
View Full Code Here

TOP

Related Classes of org.apache.uima.analysis_engine.AnalysisEngine

Copyright © 2018 www.massapicom. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.