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

Example 71 with Span

use of edu.uci.ics.textdb.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 72 with Span

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

the class DictionaryMatcherTest method testMatchingWithLimit.

public void testMatchingWithLimit() throws Exception {
    ArrayList<String> word = new ArrayList<String>(Arrays.asList("angry"));
    Dictionary dictionary = new Dictionary(word);
    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;
    Span span1 = new Span("description", 5, 10, "angry", "Angry", 1);
    Span span2 = new Span("description", 6, 11, "angry", "Angry", 1);
    Span span3 = new Span("description", 40, 45, "angry", "Angry", 8);
    Span span4 = new Span("description", 6, 11, "angry", "angry", 1);
    List<Span> list1 = new ArrayList<>();
    list1.add(span1);
    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<>(list1) };
    List<Span> list2 = new ArrayList<>();
    list2.add(span2);
    IField[] fields2 = { new StringField("brad lie angelina"), new StringField("pitt"), new IntegerField(44), new DoubleField(6.10), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-12-1972")), new TextField("White Angry"), new ListField<>(list2) };
    List<Span> list3 = new ArrayList<>();
    list3.add(span3);
    IField[] fields3 = { 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<>(list3) };
    List<Span> list4 = new ArrayList<>();
    list4.add(span4);
    IField[] fields4 = { new StringField("Mary brown"), new StringField("Lake Forest"), new IntegerField(42), new DoubleField(5.99), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-13-1974")), new TextField("Short angry"), new ListField<>(list4) };
    Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
    Tuple tuple2 = new Tuple(new Schema(schemaAttributes), fields2);
    Tuple tuple3 = new Tuple(new Schema(schemaAttributes), fields3);
    Tuple tuple4 = new Tuple(new Schema(schemaAttributes), fields4);
    List<String> attributeNames = Arrays.asList(TestConstants.FIRST_NAME, TestConstants.LAST_NAME, TestConstants.DESCRIPTION);
    List<Tuple> expectedList = new ArrayList<>();
    List<Tuple> resultList = DictionaryMatcherTestHelper.getQueryResults(PEOPLE_TABLE, dictionary, attributeNames, KeywordMatchingType.PHRASE_INDEXBASED, 3, 0);
    expectedList.add(tuple1);
    expectedList.add(tuple2);
    expectedList.add(tuple3);
    expectedList.add(tuple4);
    Assert.assertEquals(expectedList.size(), 4);
    Assert.assertEquals(resultList.size(), 3);
    Assert.assertTrue(expectedList.containsAll(resultList));
}

Example 73 with Span

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

the class DictionaryMatcherTest method testSingleWordQueryInStringFieldUsingPhrase.

/**
     * Scenario: verifies GetNextTuple of DictionaryMatcher and multiple word
     * queries in String Field using PHRASEOPERATOR
     */
@Test
public void testSingleWordQueryInStringFieldUsingPhrase() throws Exception {
    ArrayList<String> names = new ArrayList<String>(Arrays.asList("john Lee", "bruce"));
    Dictionary dictionary = new Dictionary(names);
    // create a data tuple first
    List<Span> list1 = new ArrayList<Span>();
    List<Span> list2 = new ArrayList<Span>();
    Span span1 = new Span("lastName", 0, 8, "john Lee", "john Lee");
    Span span2 = new Span("firstName", 0, 5, "bruce", "bruce");
    list1.add(span1);
    list2.add(span2);
    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>(list1) };
    Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
    IField[] fields2 = { 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>(list2) };
    Tuple tuple2 = new Tuple(new Schema(schemaAttributes), fields2);
    List<Tuple> expectedResults = new ArrayList<Tuple>();
    expectedResults.add(tuple1);
    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.PHRASE_INDEXBASED);
    boolean contains = TestUtils.equals(expectedResults, returnedResults);
    Assert.assertTrue(contains);
}

Example 74 with Span

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

the class SimilarityJoinPredicate method mergeTuples.

private Tuple mergeTuples(Tuple innerTuple, Tuple outerTuple, Schema outputSchema, List<Span> mergeSpanList) {
    List<IField> resultFields = new ArrayList<>();
    for (String attrName : outputSchema.getAttributeNames()) {
        // generate a new _ID field for this tuple
        if (attrName.equals(SchemaConstants._ID)) {
            IDField newID = new IDField(UUID.randomUUID().toString());
            resultFields.add(newID);
        // use the generated spanList
        } else if (attrName.equals(SchemaConstants.SPAN_LIST)) {
            resultFields.add(new ListField<Span>(mergeSpanList));
        // put the payload of two tuples together
        } else if (attrName.equals(SchemaConstants.PAYLOAD)) {
            ListField<Span> innerPayloadField = innerTuple.getField(SchemaConstants.PAYLOAD);
            List<Span> innerPayload = innerPayloadField.getValue();
            ListField<Span> outerPayloadField = outerTuple.getField(SchemaConstants.PAYLOAD);
            List<Span> outerPayload = outerPayloadField.getValue();
            List<Span> resultPayload = new ArrayList<>();
            resultPayload.addAll(innerPayload.stream().map(span -> addFieldPrefix(span, INNER_PREFIX)).collect(Collectors.toList()));
            resultPayload.addAll(outerPayload.stream().map(span -> addFieldPrefix(span, "outer_")).collect(Collectors.toList()));
        // add other fields from inner/outer tuples
        } else {
            if (attrName.startsWith(INNER_PREFIX)) {
                resultFields.add(innerTuple.getField(attrName.substring(INNER_PREFIX.length())));
            } else if (attrName.startsWith(OUTER_PREFIX)) {
                resultFields.add(outerTuple.getField(attrName.substring(OUTER_PREFIX.length())));
            }
        }
    }
    return new Tuple(outputSchema, resultFields.stream().toArray(IField[]::new));
}

Example 75 with Span

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

the class SimilarityJoinPredicate method joinTuples.

@Override
public Tuple joinTuples(Tuple innerTuple, Tuple outerTuple, Schema outputSchema) throws DataFlowException {
    if (similarityThreshold == 0) {
        return null;
    }
    // get the span list only with the joinAttributeName
    ListField<Span> innerSpanListField = innerTuple.getField(SchemaConstants.SPAN_LIST);
    List<Span> innerRelevantSpanList = innerSpanListField.getValue().stream().filter(span -> span.getAttributeName().equals(innerJoinAttrName)).collect(Collectors.toList());
    ListField<Span> outerSpanListField = outerTuple.getField(SchemaConstants.SPAN_LIST);
    List<Span> outerRelevantSpanList = outerSpanListField.getValue().stream().filter(span -> span.getAttributeName().equals(outerJoinAttrName)).collect(Collectors.toList());
    // get a set of span's values (since multiple spans may have the same value)
    Set<String> innerSpanValueSet = innerRelevantSpanList.stream().map(span -> span.getValue()).collect(Collectors.toSet());
    Set<String> outerSpanValueSet = outerRelevantSpanList.stream().map(span -> span.getValue()).collect(Collectors.toSet());
    // compute the result value set using the similarity function
    Set<String> resultValueSet = new HashSet<>();
    for (String innerString : innerSpanValueSet) {
        for (String outerString : outerSpanValueSet) {
            if (this.similarityFunc.calculateSimilarity(innerString, outerString) >= this.similarityThreshold) {
                resultValueSet.add(innerString);
                resultValueSet.add(outerString);
            }
        }
    }
    // return null if none of them are similar
    if (resultValueSet.isEmpty()) {
        return null;
    }
    // generate the result spans
    List<Span> resultSpans = new ArrayList<>();
    for (Span span : innerRelevantSpanList) {
        if (resultValueSet.contains(span.getValue())) {
            resultSpans.add(addFieldPrefix(span, INNER_PREFIX));
        }
    }
    for (Span span : outerRelevantSpanList) {
        if (resultValueSet.contains(span.getValue())) {
            resultSpans.add(addFieldPrefix(span, OUTER_PREFIX));
        }
    }
    return mergeTuples(innerTuple, outerTuple, outputSchema, resultSpans);
}