Examples with FeatureStructure org.apache.uima.cas.FeatureStructure used on opensource projects

Example 21 with FeatureStructure

use of org.apache.uima.cas.FeatureStructure in project webanno by webanno.

the class WebAnnoTsv3WriterTestBase method makeChainHead.

private static void makeChainHead(Type aType, AnnotationFS first) {
    CAS cas = first.getCAS();
    FeatureStructure h = cas.createFS(aType);
    FSUtil.setFeature(h, "first", first);
    cas.addFsToIndexes(h);
}

Example 22 with FeatureStructure

use of org.apache.uima.cas.FeatureStructure in project webanno by webanno.

the class WebAnnoTsv3WriterTestBase method testZeroLengthSlotFeature2.

@Test
public void testZeroLengthSlotFeature2() throws Exception {
    JCas jcas = makeJCasOneSentence();
    CAS cas = jcas.getCas();
    List<Token> tokens = new ArrayList<>(select(jcas, Token.class));
    Token t1 = tokens.get(0);
    Token t2 = tokens.get(1);
    Token t3 = tokens.get(2);
    Type type = cas.getTypeSystem().getType("webanno.custom.SimpleSpan");
    AnnotationFS s2 = cas.createAnnotation(type, t2.getBegin(), t3.getEnd());
    cas.addFsToIndexes(s2);
    AnnotationFS s3 = cas.createAnnotation(type, t3.getEnd(), t3.getEnd());
    cas.addFsToIndexes(s3);
    FeatureStructure link1 = makeLinkFS(jcas, "p1", s2);
    FeatureStructure link2 = makeLinkFS(jcas, "p2", s3);
    makeLinkHostFS(jcas, t1.getBegin(), t1.getEnd(), link1, link2);
    writeAndAssertEquals(jcas, WebannoTsv3Writer.PARAM_SLOT_FEATS, asList("webanno.custom.SimpleLinkHost:links"), WebannoTsv3Writer.PARAM_SPAN_LAYERS, asList("webanno.custom.SimpleSpan", "webanno.custom.SimpleLinkHost"), WebannoTsv3Writer.PARAM_LINK_TYPES, asList("webanno.custom.LinkType"), WebannoTsv3Writer.PARAM_SLOT_TARGETS, asList("webanno.custom.SimpleSpan"));
}

Example 23 with FeatureStructure

use of org.apache.uima.cas.FeatureStructure in project webanno by webanno.

the class WebAnnoTsv3WriterTestBase method testMultiTokenStackedSlotFeature.

@Test
public void testMultiTokenStackedSlotFeature() throws Exception {
    JCas jcas = makeJCasOneSentence();
    CAS cas = jcas.getCas();
    List<Token> tokens = new ArrayList<>(select(jcas, Token.class));
    Token t1 = tokens.get(0);
    Token t2 = tokens.get(1);
    Token t3 = tokens.get(2);
    Type type = cas.getTypeSystem().getType("webanno.custom.SimpleSpan");
    AnnotationFS s2 = cas.createAnnotation(type, t2.getBegin(), t3.getEnd());
    cas.addFsToIndexes(s2);
    AnnotationFS s3 = cas.createAnnotation(type, t2.getBegin(), t3.getEnd());
    cas.addFsToIndexes(s3);
    FeatureStructure link1 = makeLinkFS(jcas, "p1", s2);
    FeatureStructure link2 = makeLinkFS(jcas, "p2", s3);
    makeLinkHostFS(jcas, t1.getBegin(), t1.getEnd(), link1, link2);
    writeAndAssertEquals(jcas, WebannoTsv3Writer.PARAM_SLOT_FEATS, asList("webanno.custom.SimpleLinkHost:links"), WebannoTsv3Writer.PARAM_SPAN_LAYERS, asList("webanno.custom.SimpleSpan", "webanno.custom.SimpleLinkHost"), WebannoTsv3Writer.PARAM_LINK_TYPES, asList("webanno.custom.LinkType"), WebannoTsv3Writer.PARAM_SLOT_TARGETS, asList("webanno.custom.SimpleSpan"));
}

Example 24 with FeatureStructure

use of org.apache.uima.cas.FeatureStructure in project webanno by webanno.

the class WebAnnoTsv3WriterTestBase method testSimpleSameRoleSlotFeature.

@Test
public void testSimpleSameRoleSlotFeature() throws Exception {
    JCas jcas = makeJCasOneSentence();
    CAS cas = jcas.getCas();
    List<Token> tokens = new ArrayList<>(select(jcas, Token.class));
    Token t1 = tokens.get(0);
    Token t2 = tokens.get(1);
    Token t3 = tokens.get(2);
    Type type = cas.getTypeSystem().getType("webanno.custom.SimpleSpan");
    AnnotationFS s2 = cas.createAnnotation(type, t2.getBegin(), t2.getEnd());
    cas.addFsToIndexes(s2);
    AnnotationFS s3 = cas.createAnnotation(type, t3.getBegin(), t3.getEnd());
    cas.addFsToIndexes(s3);
    FeatureStructure link1 = makeLinkFS(jcas, "p1", s2);
    FeatureStructure link2 = makeLinkFS(jcas, "p1", s3);
    makeLinkHostFS(jcas, t1.getBegin(), t1.getEnd(), link1, link2);
    writeAndAssertEquals(jcas, WebannoTsv3Writer.PARAM_SLOT_FEATS, asList("webanno.custom.SimpleLinkHost:links"), WebannoTsv3Writer.PARAM_SPAN_LAYERS, asList("webanno.custom.SimpleSpan", "webanno.custom.SimpleLinkHost"), WebannoTsv3Writer.PARAM_LINK_TYPES, asList("webanno.custom.LinkType"), WebannoTsv3Writer.PARAM_SLOT_TARGETS, asList("webanno.custom.SimpleSpan"));
}

Example 25 with FeatureStructure

use of org.apache.uima.cas.FeatureStructure in project webanno by webanno.

the class Tsv3XDeserializer method readContent.

private void readContent(LineNumberReader aIn, TsvDocument aDoc) throws IOException {
    StringBuilder text = new StringBuilder();
    State prevState = State.INTER_SENTENCE_SPACE;
    State state = State.INTER_SENTENCE_SPACE;
    StringBuilder sentenceText = new StringBuilder();
    TsvSentence prevSentence = null;
    TsvSentence sentence = null;
    TsvToken token = null;
    String line = aIn.readLine();
    while (!State.END.equals(state)) {
        // These variables are only used in TOKEN and SUBTOKEN states.
        String[] fields = null;
        String id = null;
        String[] offsets = null;
        int begin = -1;
        int end = -1;
        // Determine the status of the current line
        if (startsWith(line, PREFIX_TEXT)) {
            state = State.SENTENCE;
        } else if (line == null) {
            state = State.END;
        } else if (isEmpty(line)) {
            state = State.INTER_SENTENCE_SPACE;
        } else {
            fields = splitPreserveAllTokens(line, FIELD_SEPARATOR);
            // Get token metadata
            id = fields[0];
            offsets = split(fields[1], "-");
            begin = Integer.valueOf(offsets[0]);
            end = Integer.valueOf(offsets[1]);
            // TOKEN or SUBTOKEN?
            if (id.contains(".")) {
                state = State.SUBTOKEN;
            } else {
                state = State.TOKEN;
            }
        }
        // Assert that the order of information in the file is correct
        switch(prevState) {
            case INTER_SENTENCE_SPACE:
                if (!State.SENTENCE.equals(state)) {
                    throw new IOException("Line " + aIn.getLineNumber() + ": Expected sentence header but got [" + state + "]");
                }
                break;
            case SENTENCE:
                if (!(State.SENTENCE.equals(state) || State.TOKEN.equals(state))) {
                    throw new IOException("Line " + aIn.getLineNumber() + ": Expected sentence header or token but got [" + state + "]");
                }
                break;
            case TOKEN:
            case SUBTOKEN:
                if (!(State.INTER_SENTENCE_SPACE.equals(state) || State.END.equals(state) || State.TOKEN.equals(state) || State.SUBTOKEN.equals(state))) {
                    throw new IOException("Line " + aIn.getLineNumber() + ": Expected token, sub-token or sentence break but got [" + state + "]");
                }
                break;
        }
        // Do the actual parsing
        switch(state) {
            case END:
            case INTER_SENTENCE_SPACE:
                // End of sentence action
                // The -1 here is to account for the tailing line break
                sentence.getUimaSentence().setEnd(text.length() - 1);
                sentence.getUimaSentence().addToIndexes();
                prevSentence = sentence;
                sentence = null;
                break;
            case TOKEN:
                // End of sentence header action
                if (State.SENTENCE.equals(prevState)) {
                    // last sentence!
                    if (text.length() > begin) {
                        assert text.length() == begin + 1;
                        assert text.charAt(text.length() - 1) == LINE_BREAK;
                        text.setLength(text.length() - 1);
                    }
                    // the gap.
                    if (text.length() < begin) {
                        text.append(repeat(' ', begin - text.length()));
                    }
                    assert text.length() == begin;
                    assert sentence == null;
                    Sentence uimaSentence = new Sentence(aDoc.getJCas());
                    uimaSentence.setBegin(text.length());
                    sentence = aDoc.createSentence(uimaSentence);
                    text.append(sentenceText);
                    sentenceText.setLength(0);
                }
                // Token parsing action
                Token uimaToken = new Token(aDoc.getJCas(), begin, end);
                uimaToken.addToIndexes();
                token = sentence.createToken(uimaToken);
                // Read annotations from the columns
                parseAnnotations(aDoc, sentence, token, fields);
                break;
            case SUBTOKEN:
                // Read annotations from the columns
                TsvSubToken subToken = token.createSubToken(begin, end);
                parseAnnotations(aDoc, sentence, subToken, fields);
                break;
            case SENTENCE:
                // Header parsing action
                String textFragment = substringAfter(line, "=");
                textFragment = unescapeText(aDoc.getFormatHeader(), textFragment);
                sentenceText.append(textFragment);
                sentenceText.append(LINE_BREAK);
                break;
        }
        prevState = state;
        line = aIn.readLine();
    }
    aDoc.getJCas().setDocumentText(text.toString());
    // After all data has been read, we also add the annotations with disambiguation ID to
    // the CAS indexes. This ensures we only add them after their final begin/end offsets
    // have been determined since most of these annotations are actually multi-token
    // annotations.
    CAS cas = aDoc.getJCas().getCas();
    Set<FeatureStructure> fses = new LinkedHashSet<>();
    for (TsvSentence s : aDoc.getSentences()) {
        for (TsvToken t : s.getTokens()) {
            for (Type type : t.getUimaTypes()) {
                fses.addAll(t.getUimaAnnotations(type));
            }
            for (TsvSubToken st : t.getSubTokens()) {
                for (Type type : st.getUimaTypes()) {
                    fses.addAll(st.getUimaAnnotations(type));
                }
            }
        }
    }
    fses.forEach(cas::addFsToIndexes);
}