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

Example 56 with Span

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

the class DictionaryMatcherTest method testSingleWordQueryInStringFieldUsingPhraseChinese.

/**
     * Scenario: verifies GetNextTuple of DictionaryMatcher and multiple word
     * queries in String Field using PHRASE_INDEXBASED in Chinese.
     */
@Test
public void testSingleWordQueryInStringFieldUsingPhraseChinese() throws Exception {
    ArrayList<String> names = new ArrayList<String>(Arrays.asList("长孙", "无忌"));
    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, 2, "长孙", "长孙");
    Span span2 = new Span("firstName", 0, 2, "无忌", "无忌");
    list1.add(span1);
    list2.add(span2);
    Attribute[] schemaAttributes = new Attribute[TestConstantsChinese.ATTRIBUTES_PEOPLE.length + 1];
    for (int count = 0; count < schemaAttributes.length - 1; count++) {
        schemaAttributes[count] = TestConstantsChinese.ATTRIBUTES_PEOPLE[count];
    }
    schemaAttributes[schemaAttributes.length - 1] = RESULTS_ATTRIBUTE;
    IField[] fields1 = { new StringField("无忌"), new StringField("长孙"), new IntegerField(46), new DoubleField(5.50), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-14-1970")), new TextField("北京大学电气工程学院"), new ListField<Span>(list1) };
    Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
    IField[] fields2 = { new StringField("无忌"), new StringField("长孙"), new IntegerField(46), new DoubleField(5.50), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("01-14-1970")), new TextField("北京大学电气工程学院"), 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(TestConstantsChinese.FIRST_NAME, TestConstantsChinese.LAST_NAME, TestConstantsChinese.DESCRIPTION);
    List<Tuple> returnedResults = DictionaryMatcherTestHelper.getQueryResults(CHINESE_TABLE, dictionary, attributeNames, KeywordMatchingType.PHRASE_INDEXBASED);
    boolean contains = TestUtils.equals(expectedResults, returnedResults);
    Assert.assertTrue(contains);
}

Example 57 with Span

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

the class DataReader method buildPayloadFromTermVector.

private ArrayList<Span> buildPayloadFromTermVector(List<IField> fields, int docID) throws IOException {
    ArrayList<Span> payloadSpanList = new ArrayList<>();
    for (Attribute attr : inputSchema.getAttributes()) {
        String attributeName = attr.getAttributeName();
        AttributeType attributeType = attr.getAttributeType();
        // payload.
        if (attributeType != AttributeType.TEXT) {
            continue;
        }
        String fieldValue = fields.get(inputSchema.getIndex(attributeName)).getValue().toString();
        Terms termVector = luceneIndexReader.getTermVector(docID, attributeName);
        if (termVector == null) {
            continue;
        }
        TermsEnum termsEnum = termVector.iterator();
        PostingsEnum termPostings = null;
        // go through document terms
        while ((termsEnum.next()) != null) {
            termPostings = termsEnum.postings(termPostings, PostingsEnum.ALL);
            if (termPostings.nextDoc() == DocIdSetIterator.NO_MORE_DOCS) {
                continue;
            }
            // for each term, go through its postings
            for (int i = 0; i < termPostings.freq(); i++) {
                // nextPosition needs to be called first
                int tokenPosition = termPostings.nextPosition();
                int charStart = termPostings.startOffset();
                int charEnd = termPostings.endOffset();
                String analyzedTermStr = termsEnum.term().utf8ToString();
                String originalTermStr = fieldValue.substring(charStart, charEnd);
                Span span = new Span(attributeName, charStart, charEnd, analyzedTermStr, originalTermStr, tokenPosition);
                payloadSpanList.add(span);
            }
        }
    }
    return payloadSpanList;
}

Example 58 with Span

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

the class DictionaryMatcherPerformanceTest method match.

/**
     * This function does match for a dictionary
     */
public static void match(ArrayList<String> queryList, KeywordMatchingType opType, String luceneAnalyzerStr, String tableName) throws Exception {
    List<String> attributeNames = Arrays.asList(MedlineIndexWriter.ABSTRACT);
    Dictionary dictionary = new Dictionary(queryList);
    DictionarySourcePredicate dictionarySourcePredicate = new DictionarySourcePredicate(dictionary, attributeNames, luceneAnalyzerStr, opType, tableName, null);
    DictionaryMatcherSourceOperator dictionaryMatcher = new DictionaryMatcherSourceOperator(dictionarySourcePredicate);
    long startMatchTime = System.currentTimeMillis();
    dictionaryMatcher.open();
    Tuple nextTuple = null;
    int counter = 0;
    while ((nextTuple = dictionaryMatcher.getNextTuple()) != null) {
        ListField<Span> spanListField = nextTuple.getField(SchemaConstants.SPAN_LIST);
        List<Span> spanList = spanListField.getValue();
        counter += spanList.size();
    }
    dictionaryMatcher.close();
    long endMatchTime = System.currentTimeMillis();
    matchTime = (endMatchTime - startMatchTime) / 1000.0;
    resultCount = counter;
}

Example 59 with Span

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

the class RegexMatcherPerformanceTest method matchRegex.

/*
     *         This function does match for a list of regex queries
     */
public static void matchRegex(List<String> regexes, String tableName) throws TextDBException, IOException {
    List<String> attributeNames = Arrays.asList(MedlineIndexWriter.ABSTRACT);
    for (String regex : regexes) {
        // analyzer should generate grams all in lower case to build a lower
        // case index.
        RegexSourcePredicate predicate = new RegexSourcePredicate(regex, attributeNames, tableName, null);
        RegexMatcherSourceOperator regexSource = new RegexMatcherSourceOperator(predicate);
        long startMatchTime = System.currentTimeMillis();
        regexSource.open();
        int counter = 0;
        Tuple nextTuple = null;
        while ((nextTuple = regexSource.getNextTuple()) != null) {
            ListField<Span> spanListField = nextTuple.getField(SchemaConstants.SPAN_LIST);
            List<Span> spanList = spanListField.getValue();
            counter += spanList.size();
        }
        regexSource.close();
        long endMatchTime = System.currentTimeMillis();
        double matchTime = (endMatchTime - startMatchTime) / 1000.0;
        totalMatchingTime += matchTime;
        totalRegexResultCount += counter;
    }
}

Example 60 with Span

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

the class ExcelSinkTest method writeSampleExcelFile.

/**
     * Create two tuples, write into a excel file. Need to manually delete the generated file.
     * @throws ParseException
     */
@Test
public void writeSampleExcelFile() throws Exception {
    ArrayList<String> attributeNames = new ArrayList<>();
    attributeNames.add(TestConstants.FIRST_NAME);
    attributeNames.add(TestConstants.LAST_NAME);
    attributeNames.add(TestConstants.DESCRIPTION);
    // Prepare the expected result list
    List<Span> list = new ArrayList<>();
    Span span1 = new Span("firstName", 0, 5, "bruce", "bruce");
    Span span2 = new Span("lastnName", 0, 5, "jacki", "jacki");
    list.add(span1);
    list.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] = SchemaConstants.SPAN_LIST_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<>(list) };
    IField[] fields2 = { new StringField("test"), new StringField("jackie chan"), new IntegerField(0), new DoubleField(6.0), new DateField(new SimpleDateFormat("MM-dd-yyyy").parse("09-18-1994")), new TextField("Angry Bird"), new ListField<>(list) };
    Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
    Tuple tuple2 = new Tuple(new Schema(schemaAttributes), fields2);
    IOperator inputOperator = Mockito.mock(IOperator.class);
    Mockito.when(inputOperator.getOutputSchema()).thenReturn(new Schema(schemaAttributes)).thenReturn(null);
    Mockito.when(inputOperator.getNextTuple()).thenReturn(tuple1).thenReturn(tuple2).thenReturn(null);
    excelSink = new ExcelSink(new ExcelSinkPredicate());
    excelSink.setInputOperator(inputOperator);
    excelSink.open();
    excelSink.collectAllTuples();
    excelSink.close();
    Files.deleteIfExists(Paths.get(excelSink.getFilePath()));
}