Examples with Span edu.uci.ics.texera.api.span.Span used on opensource projects

Example 66 with Span

use of edu.uci.ics.texera.api.span.Span in project textdb by TextDB.

the class DataflowUtils method constructPhraseMatchingSpans.

/**
 * This function is used to generate the SpanList for phrase matching type in both dictionarymatcher and keywordmatcher.
 * @param attributeName
 * @param fieldValue
 * @param queryKeyword
 * @param fieldSpanList
 * @param queryTokenListWithStopwords
 * @param queryTokenList
 * @return
 */
public static List<Span> constructPhraseMatchingSpans(String attributeName, String fieldValue, String queryKeyword, List<Span> fieldSpanList, List<String> queryTokenListWithStopwords, List<String> queryTokenList) {
    List<Span> matchingResults = new ArrayList<>();
    // Sort current field's span list by token offset for later use
    Collections.sort(fieldSpanList, (span1, span2) -> span1.getTokenOffset() - span2.getTokenOffset());
    List<Integer> queryTokenOffset = new ArrayList<>();
    for (int i = 0; i < queryTokenListWithStopwords.size(); i++) {
        if (queryTokenList.contains(queryTokenListWithStopwords.get(i))) {
            queryTokenOffset.add(i);
        }
    }
    // maintains position of term being checked in
    int iter = 0;
    // spanForThisField list
    while (iter < fieldSpanList.size()) {
        if (iter > fieldSpanList.size() - queryTokenList.size()) {
            break;
        }
        // Verify if span in the spanForThisField correspond to our
        // phrase query, ie relative position offsets should be
        // similar
        // and the value should be same.
        // flag to check if a
        boolean isMismatchInSpan = false;
        // To check all the terms in query are verified
        for (int i = 0; i < queryTokenList.size() - 1; i++) {
            Span first = fieldSpanList.get(iter + i);
            Span second = fieldSpanList.get(iter + i + 1);
            if (!(second.getTokenOffset() - first.getTokenOffset() == queryTokenOffset.get(i + 1) - queryTokenOffset.get(i) && first.getValue().equalsIgnoreCase(queryTokenList.get(i)) && second.getValue().equalsIgnoreCase(queryTokenList.get(i + 1)))) {
                iter++;
                isMismatchInSpan = true;
                break;
            }
        }
        if (isMismatchInSpan) {
            continue;
        }
        int combinedSpanStartIndex = fieldSpanList.get(iter).getStart();
        int combinedSpanEndIndex = fieldSpanList.get(iter + queryTokenList.size() - 1).getEnd();
        Span combinedSpan = new Span(attributeName, combinedSpanStartIndex, combinedSpanEndIndex, queryKeyword, fieldValue.substring(combinedSpanStartIndex, combinedSpanEndIndex));
        matchingResults.add(combinedSpan);
        iter = iter + queryTokenList.size();
    }
    return matchingResults;
}

Example 67 with Span

use of edu.uci.ics.texera.api.span.Span in project textdb by TextDB.

the class WordCountOperator method computeWordCount.

private void computeWordCount() throws TexeraException {
    Tuple tuple;
    HashMap<String, Integer> wordCountMap = new HashMap<>();
    while ((tuple = this.inputOperator.getNextTuple()) != null) {
        if (addPayload) {
            tuple = new Tuple.Builder(tuple).add(SchemaConstants.PAYLOAD_ATTRIBUTE, new ListField<Span>(DataflowUtils.generatePayloadFromTuple(tuple, predicate.getLuceneAnalyzerString()))).build();
        }
        ListField<Span> payloadField = tuple.getField("payload");
        List<Span> payloadSpanList = payloadField.getValue();
        for (Span span : payloadSpanList) {
            if (span.getAttributeName().equals(predicate.getAttribute())) {
                String key = span.getValue().toLowerCase();
                wordCountMap.put(key, wordCountMap.get(key) == null ? 1 : wordCountMap.get(key) + 1);
            }
        }
    }
    sortedWordCountMap = wordCountMap.entrySet().stream().sorted((e1, e2) -> e2.getValue().compareTo(e1.getValue())).collect(Collectors.toList());
    wordCountIterator = sortedWordCountMap.iterator();
}

Example 68 with Span

use of edu.uci.ics.texera.api.span.Span in project textdb by TextDB.

the class DictionaryMatcherTest method testMultipleWordsQueryUsingPhrase.

/**
 * Scenario S-10:verifies ITuple returned by DictionaryMatcher and multiple
 * word queries using PHRASE OPERATOR
 */
@Test
public void testMultipleWordsQueryUsingPhrase() throws Exception {
    ArrayList<String> names = new ArrayList<String>(Arrays.asList("george lin lin"));
    Dictionary dictionary = new Dictionary(names);
    // create a data tuple first
    List<Span> list = new ArrayList<Span>();
    Span span = new Span("firstName", 0, 14, "george lin lin", "george lin lin");
    list.add(span);
    Attribute[] schemaAttributes = new Attribute[TestConstants.ATTRIBUTES_PEOPLE.length + 1];
    for (int count = 0; count < schemaAttributes.length - 1; count++) {
        schemaAttributes[count] = TestConstants.ATTRIBUTES_PEOPLE[count];
    }
    schemaAttributes[schemaAttributes.length - 1] = RESULTS_ATTRIBUTE;
    IField[] fields1 = { new StringField("george lin lin"), new StringField("lin clooney"), new IntegerField(43), new DoubleField(6.06), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-13-1973")), new TextField("Lin Clooney is Short and lin clooney is Angry"), new ListField<Span>(list) };
    Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
    List<Tuple> expectedResults = new ArrayList<Tuple>();
    expectedResults.add(tuple1);
    List<String> attributeNames = Arrays.asList(TestConstants.FIRST_NAME, TestConstants.LAST_NAME, TestConstants.DESCRIPTION);
    List<Tuple> returnedResults = DictionaryMatcherTestHelper.getQueryResults(PEOPLE_TABLE, dictionary, attributeNames, KeywordMatchingType.PHRASE_INDEXBASED);
    boolean contains = TestUtils.equals(expectedResults, returnedResults);
    Assert.assertTrue(contains);
}

Example 69 with Span

use of edu.uci.ics.texera.api.span.Span in project textdb by TextDB.

the class DictionaryMatcherTest method testMultipleWordQueryInTextFieldUsingScan2.

@Test
public void testMultipleWordQueryInTextFieldUsingScan2() throws Exception {
    ArrayList<String> names = new ArrayList<String>(Arrays.asList("tall fair"));
    Dictionary dictionary = new Dictionary(names);
    // create a data tuple first
    List<Span> list1 = new ArrayList<Span>();
    Span span1 = new Span("description", 0, 9, "tall fair", "Tall Fair");
    list1.add(span1);
    Attribute[] schemaAttributes = new Attribute[TestConstants.ATTRIBUTES_PEOPLE.length + 1];
    for (int count = 0; count < schemaAttributes.length - 1; count++) {
        schemaAttributes[count] = TestConstants.ATTRIBUTES_PEOPLE[count];
    }
    schemaAttributes[schemaAttributes.length - 1] = RESULTS_ATTRIBUTE;
    IField[] fields2 = { new StringField("christian john wayne"), new StringField("rock bale"), new IntegerField(42), new DoubleField(5.99), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-13-1974")), new TextField("Tall Fair"), new ListField<Span>(list1) };
    Tuple tuple2 = new Tuple(new Schema(schemaAttributes), fields2);
    List<Tuple> expectedResults = new ArrayList<Tuple>();
    expectedResults.add(tuple2);
    List<String> attributeNames = Arrays.asList(TestConstants.FIRST_NAME, TestConstants.LAST_NAME, TestConstants.DESCRIPTION);
    List<Tuple> returnedResults = DictionaryMatcherTestHelper.getQueryResults(PEOPLE_TABLE, dictionary, attributeNames, KeywordMatchingType.SUBSTRING_SCANBASED);
    boolean contains = TestUtils.equals(expectedResults, returnedResults);
    Assert.assertTrue(contains);
}

Example 70 with Span

use of edu.uci.ics.texera.api.span.Span in project textdb by TextDB.

the class DictionaryMatcherTest method testSingleWordQueryInStringFieldUsingScan.

/**
 * Scenario: verifies GetNextTuple of DictionaryMatcher and single word
 * queries in String Field using SCANOPERATOR
 */
@Test
public void testSingleWordQueryInStringFieldUsingScan() throws Exception {
    ArrayList<String> names = new ArrayList<String>(Arrays.asList("bruce"));
    Dictionary dictionary = new Dictionary(names);
    // create a data tuple first
    List<Span> list = new ArrayList<Span>();
    Span span = new Span("firstName", 0, 5, "bruce", "bruce");
    list.add(span);
    Attribute[] schemaAttributes = new Attribute[TestConstants.ATTRIBUTES_PEOPLE.length + 1];
    for (int count = 0; count < schemaAttributes.length - 1; count++) {
        schemaAttributes[count] = TestConstants.ATTRIBUTES_PEOPLE[count];
    }
    schemaAttributes[schemaAttributes.length - 1] = RESULTS_ATTRIBUTE;
    IField[] fields1 = { new StringField("bruce"), new StringField("john Lee"), new IntegerField(46), new DoubleField(5.50), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-14-1970")), new TextField("Tall Angry"), new ListField<Span>(list) };
    Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
    List<Tuple> expectedResults = new ArrayList<Tuple>();
    expectedResults.add(tuple1);
    List<String> attributeNames = Arrays.asList(TestConstants.FIRST_NAME, TestConstants.LAST_NAME, TestConstants.DESCRIPTION);
    List<Tuple> returnedResults = DictionaryMatcherTestHelper.getQueryResults(PEOPLE_TABLE, dictionary, attributeNames, KeywordMatchingType.SUBSTRING_SCANBASED);
    boolean contains = TestUtils.equals(expectedResults, returnedResults);
    Assert.assertTrue(contains);
}