Examples with DataFlowException edu.uci.ics.textdb.api.exception.DataFlowException used on opensource projects

Example 6 with DataFlowException

use of edu.uci.ics.textdb.api.exception.DataFlowException in project textdb by TextDB.

the class FuzzyTokenMatcher method processOneInputTuple.

@Override
public Tuple processOneInputTuple(Tuple inputTuple) throws TextDBException {
    ListField<Span> payloadField = inputTuple.getField(SchemaConstants.PAYLOAD);
    List<Span> payload = payloadField.getValue();
    List<Span> relevantSpans = filterRelevantSpans(payload);
    List<Span> matchResults = new ArrayList<>();
    /*
         * The source operator returns spans even for those fields which did not
         * satisfy the threshold criterion. So if two attributes A,B have 10 and
         * 5 matching tokens, and we set threshold to 10, the number of spans
         * returned is 15. So we need to filter those 5 spans for attribute B.
         */
    for (String attributeName : this.predicate.getAttributeNames()) {
        AttributeType attributeType = this.inputSchema.getAttribute(attributeName).getAttributeType();
        // types other than TEXT and STRING: throw Exception for now
        if (attributeType != AttributeType.TEXT && attributeType != AttributeType.STRING) {
            throw new DataFlowException("FuzzyTokenMatcher: Fields other than TEXT or STRING are not supported");
        }
        List<Span> fieldSpans = relevantSpans.stream().filter(span -> span.getAttributeName().equals(attributeName)).filter(span -> predicate.getQueryTokens().contains(span.getKey())).collect(Collectors.toList());
        if (fieldSpans.size() >= predicate.getThreshold()) {
            matchResults.addAll(fieldSpans);
        }
    }
    if (matchResults.isEmpty()) {
        return null;
    }
    ListField<Span> spanListField = inputTuple.getField(predicate.getSpanListName());
    List<Span> spanList = spanListField.getValue();
    spanList.addAll(matchResults);
    return inputTuple;
}

Example 7 with DataFlowException

use of edu.uci.ics.textdb.api.exception.DataFlowException in project textdb by TextDB.

the class FuzzyTokenMatcherSourceOperator method createLuceneQueryObject.

public static Query createLuceneQueryObject(FuzzyTokenPredicate predicate) throws DataFlowException {
    try {
        /*
             * By default the boolean query takes 1024 # of clauses as the max
             * limit. Since our input query has no limitaion on the number of
             * tokens, we have to put a check.
             */
        if (predicate.getThreshold() > 1024)
            BooleanQuery.setMaxClauseCount(predicate.getThreshold() + 1);
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.setMinimumNumberShouldMatch(predicate.getThreshold());
        MultiFieldQueryParser qp = new MultiFieldQueryParser(predicate.getAttributeNames().stream().toArray(String[]::new), LuceneAnalyzerConstants.getLuceneAnalyzer(predicate.getLuceneAnalyzerStr()));
        for (String s : predicate.getQueryTokens()) {
            builder.add(qp.parse(s), Occur.SHOULD);
        }
        return builder.build();
    } catch (ParseException e) {
        throw new DataFlowException(e);
    }
}

Example 8 with DataFlowException

use of edu.uci.ics.textdb.api.exception.DataFlowException in project textdb by TextDB.

the class Join method close.

@Override
public void close() throws TextDBException {
    if (cursor == CLOSED) {
        return;
    }
    try {
        innerOperator.close();
        outerOperator.close();
    } catch (Exception e) {
        throw new DataFlowException(e.getMessage(), e);
    }
    // Set the inner tuple list back to null on close.
    innerTupleList = null;
    innerTupleListCursor = 0;
    cursor = CLOSED;
}

Example 9 with DataFlowException

use of edu.uci.ics.textdb.api.exception.DataFlowException in project textdb by TextDB.

the class NlpSentimentOperator method open.

@Override
public void open() throws TextDBException {
    if (cursor != CLOSED) {
        return;
    }
    if (inputOperator == null) {
        throw new DataFlowException(ErrorMessages.INPUT_OPERATOR_NOT_SPECIFIED);
    }
    inputOperator.open();
    Schema inputSchema = inputOperator.getOutputSchema();
    // check if input schema is present
    if (!inputSchema.containsField(predicate.getInputAttributeName())) {
        throw new RuntimeException(String.format("input attribute %s is not in the input schema %s", predicate.getInputAttributeName(), inputSchema.getAttributeNames()));
    }
    // check if attribute type is valid
    AttributeType inputAttributeType = inputSchema.getAttribute(predicate.getInputAttributeName()).getAttributeType();
    boolean isValidType = inputAttributeType.equals(AttributeType.STRING) || inputAttributeType.equals(AttributeType.TEXT);
    if (!isValidType) {
        throw new RuntimeException(String.format("input attribute %s must have type String or Text, its actual type is %s", predicate.getInputAttributeName(), inputAttributeType));
    }
    // generate output schema by transforming the input schema
    outputSchema = transformSchema(inputOperator.getOutputSchema());
    cursor = OPENED;
    // setup NLP sentiment analysis pipeline
    Properties props = new Properties();
    props.setProperty("annotators", "tokenize, ssplit, parse, sentiment");
    sentimentPipeline = new StanfordCoreNLP(props);
}

Example 10 with DataFlowException

use of edu.uci.ics.textdb.api.exception.DataFlowException in project textdb by TextDB.

the class NlpSplitOperator method open.

@Override
public void open() throws TextDBException {
    if (cursor != CLOSED) {
        return;
    }
    if (inputOperator == null) {
        throw new DataFlowException(ErrorMessages.INPUT_OPERATOR_NOT_SPECIFIED);
    }
    inputOperator.open();
    Schema inputSchema = inputOperator.getOutputSchema();
    // check if input schema is present
    if (!inputSchema.containsField(predicate.getInputAttributeName())) {
        throw new DataFlowException(String.format("input attribute %s is not in the input schema %s", predicate.getInputAttributeName(), inputSchema.getAttributeNames()));
    }
    // check if attribute type is valid
    AttributeType inputAttributeType = inputSchema.getAttribute(predicate.getInputAttributeName()).getAttributeType();
    boolean isValidType = inputAttributeType.equals(AttributeType.STRING) || inputAttributeType.equals(AttributeType.TEXT);
    if (!isValidType) {
        throw new DataFlowException(String.format("input attribute %s must have type String or Text, its actual type is %s", predicate.getInputAttributeName(), inputAttributeType));
    }
    // generate output schema by transforming the input schema based on what output format
    // is chosen (OneToOne vs. OneToMany)
    outputSchema = transformSchema(inputOperator.getOutputSchema());
    cursor = OPENED;
}