use of edu.uci.ics.textdb.api.field.TextField in project textdb by TextDB.
the class KeywordPhraseTest method testWordInMultipleFieldsQueryWithStopWords1.
/**
* Verifies: Query with Stop Words match corresponding phrases in the
* document
*
* @throws Exception
*/
@Test
public void testWordInMultipleFieldsQueryWithStopWords1() throws Exception {
// Prepare Query
String query = "lin and and angry";
ArrayList<String> attributeNames = new ArrayList<>();
attributeNames.add(TestConstants.FIRST_NAME);
attributeNames.add(TestConstants.LAST_NAME);
attributeNames.add(TestConstants.DESCRIPTION);
// Prepare expected result list
List<Span> list = new ArrayList<>();
Span span1 = new Span("description", 25, 45, "lin and and angry", "lin clooney is Angry");
list.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] = new Attribute(RESULTS, AttributeType.LIST);
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<>(list) };
Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields1);
List<Tuple> expectedResultList = new ArrayList<>();
expectedResultList.add(tuple1);
// Perform Query
List<Tuple> resultList = KeywordTestHelper.getQueryResults(PEOPLE_TABLE, query, attributeNames, phrase);
// Perform Check
boolean contains = TestUtils.equals(expectedResultList, resultList);
Assert.assertTrue(contains);
}
use of edu.uci.ics.textdb.api.field.TextField in project textdb by TextDB.
the class KeywordPhraseTest method testWordInMultipleFieldsQueryWithStopWords5.
/**
* Verifies: Query with Stop Words match corresponding phrases in the
* document Used to cause exception sometimes if there is a space between
* words
*
* @throws Exception
* with Medline data
*/
@Test
public void testWordInMultipleFieldsQueryWithStopWords5() throws Exception {
// Prepare Query
String query = "gain weight";
ArrayList<String> attributeNames = new ArrayList<>();
attributeNames.add(keywordTestConstants.ABSTRACT);
// Prepare expected result list
List<Span> list = new ArrayList<>();
Span span1 = new Span(keywordTestConstants.ABSTRACT, 26, 37, "gain weight", "gain weight");
list.add(span1);
Attribute[] schemaAttributes = new Attribute[keywordTestConstants.ATTRIBUTES_MEDLINE.length + 1];
for (int count = 0; count < schemaAttributes.length - 1; count++) {
schemaAttributes[count] = keywordTestConstants.ATTRIBUTES_MEDLINE[count];
}
schemaAttributes[schemaAttributes.length - 1] = new Attribute(RESULTS, AttributeType.LIST);
IField[] fields = { new IntegerField(4566015), new TextField(""), new TextField("Significance of milk pH in newborn infants."), new TextField("V C Harrison, G Peat"), new StringField("4-5839 Dec 2, 1972"), new TextField("British medical journal"), new TextField(""), new TextField("Infant Nutritional Physiological Phenomena, Infant, Newborn, Milk"), new TextField("Bottle-fed infants do not gain weight as rapidly as breast-fed babies during the first week of life. This " + "weight lag can be corrected by the addition of a small amount of alkali (sodium bicarbonate or trometamol) to " + "the feeds. The alkali corrects the acidity of cow's milk which now assumes some of the properties of human breast " + "milk. It has a bacteriostatic effect on specific Escherichia coli in vitro, and in infants it produces a stool with" + " a preponderance of lactobacilli over E. coli organisms. When alkali is removed from the milk there is a decrease in" + " the weight of an infant and the stools contain excessive numbers of E. coli bacteria.A pH-corrected milk appears to" + " be more physiological than unaltered cow's milk and may provide some protection against gastroenteritis in early " + "life. Its bacteriostatic effect on specific E. coli may be of practical significance in feed preparations where " + "terminal sterilization and refrigeration are not available. The study was conducted during the week after birth, and " + "no conclusions are derived for older infants. The long-term effects of trometamol are unknown. No recommendation can " + "be given for the addition of sodium bicarbonate to milks containing a higher content of sodium."), new DoubleField(0.667832788), new ListField<>(list) };
Tuple tuple1 = new Tuple(new Schema(schemaAttributes), fields);
List<Tuple> expectedResultList = new ArrayList<>();
expectedResultList.add(tuple1);
List<Tuple> results = KeywordTestHelper.getQueryResults(MEDLINE_TABLE, query, attributeNames, phrase);
// Perform Check
boolean contains = TestUtils.equals(expectedResultList, results);
Assert.assertTrue(contains);
}
use of edu.uci.ics.textdb.api.field.TextField in project textdb by TextDB.
the class FuzzyTokenMatcherTest method TestFuzzyTokenMatcherWithLargeTokens.
@Test
public void TestFuzzyTokenMatcherWithLargeTokens() throws Exception {
String query = "Twelve Angry Men Came Cafe Have Coffee Eat Chocolate Burger Fries SandWidch Cool Food Drinks American drama film elements film noir adapted teleplay same name Reginald Rose Written Rose directed Sidney Lumet trial film tells story jury made deliberate guilt acquittal defendant basis reasonable doubt United States verdict most criminal ";
double threshold = 0.02;
ArrayList<String> attributeNames = new ArrayList<>();
attributeNames.add(TestConstants.DESCRIPTION);
Schema schema = Utils.addAttributeToSchema(TestConstants.SCHEMA_PEOPLE, RESULTS_ATTR);
List<Span> spanList1 = Arrays.asList(new Span(TestConstants.DESCRIPTION, 5, 10, "angry", "Angry", 1));
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>(spanList1) };
List<Span> spanList2 = Arrays.asList(new Span(TestConstants.DESCRIPTION, 6, 11, "angry", "Angry", 1));
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<Span>(spanList2) };
List<Span> spanList3 = Arrays.asList(new Span(TestConstants.DESCRIPTION, 40, 45, "angry", "Angry", 8));
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<Span>(spanList3) };
List<Span> spanList4 = Arrays.asList(new Span(TestConstants.DESCRIPTION, 6, 11, "angry", "angry", 1));
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<Span>(spanList4) };
Tuple tuple1 = new Tuple(schema, fields1);
Tuple tuple2 = new Tuple(schema, fields2);
Tuple tuple3 = new Tuple(schema, fields3);
Tuple tuple4 = new Tuple(schema, fields4);
List<Tuple> expectedResultList = new ArrayList<>();
expectedResultList.add(tuple1);
expectedResultList.add(tuple2);
expectedResultList.add(tuple3);
expectedResultList.add(tuple4);
List<Tuple> results = FuzzyTokenMatcherTestHelper.getQueryResults(PEOPLE_TABLE, query, threshold, attributeNames);
boolean contains = TestUtils.equals(expectedResultList, results);
Assert.assertTrue(contains);
}
use of edu.uci.ics.textdb.api.field.TextField in project textdb by TextDB.
the class JoinDistanceTest method testIdsMatchFieldsMatchSpanWithinThreshold.
/*
* This case tests for the scenario when the difference of keyword spans
* is within the given span threshold.
* e.g.
* [<11, 18>]
* [<27, 33>]
* threshold = 20 (within threshold)
* result: [<11, 33>]
*
* Test result: The list contains a tuple with all the fields and a span
* list consisting of the joined span. The joined span is made up of the
* field name, start and stop index (computed as <min(span1 spanStartIndex,
* span2 spanStartIndex), max(span1 spanEndIndex, span2 spanEndIndex)>)
* key (combination of span1 key and span2 key) and value (combination of
* span1 value and span2 value).
*
*/
@Test
public void testIdsMatchFieldsMatchSpanWithinThreshold() throws Exception {
JoinTestHelper.insertToTable(BOOK_TABLE, JoinTestConstants.bookGroup1.get(0));
KeywordMatcherSourceOperator keywordSourceOuter = JoinTestHelper.getKeywordSource(BOOK_TABLE, "special", conjunction);
KeywordMatcherSourceOperator keywordSourceInner = JoinTestHelper.getKeywordSource(BOOK_TABLE, "writer", conjunction);
List<Tuple> resultList = JoinTestHelper.getJoinDistanceResults(keywordSourceInner, keywordSourceOuter, new JoinDistancePredicate(JoinTestConstants.REVIEW, 20), Integer.MAX_VALUE, 0);
Schema resultSchema = Utils.createSpanSchema(JoinTestConstants.BOOK_SCHEMA);
List<Span> spanList = new ArrayList<>();
Span span1 = new Span(JoinTestConstants.REVIEW, 11, 33, "special_writer", "special kind of " + "writer");
spanList.add(span1);
IField[] book1 = { new IntegerField(52), new StringField("Mary Roach"), new StringField("Grunt: The Curious Science of Humans at War"), new IntegerField(288), new TextField("It takes a special kind " + "of writer to make topics ranging from death to our " + "gastrointestinal tract interesting (sometimes " + "hilariously so), and pop science writer Mary Roach is " + "always up to the task."), new ListField<>(spanList) };
Tuple expectedTuple = new Tuple(resultSchema, book1);
List<Tuple> expectedResult = new ArrayList<>();
expectedResult.add(expectedTuple);
Assert.assertEquals(1, resultList.size());
Assert.assertTrue(TestUtils.equals(expectedResult, resultList));
}
use of edu.uci.ics.textdb.api.field.TextField in project textdb by TextDB.
the class JoinDistanceTest method testBothOperatorsMultipleTuplesSpanWithinThreshold.
// --------------------<END of single tuple test cases>--------------------
/*
* This case tests for the scenario when both the operators' have multiple
* tuples and spans are within threshold.
*
* Test result: Join should result in a list containing tuples with spans.
* The number of tuples is equal to the number of tuples with spans within threshold.
*/
@Test
public void testBothOperatorsMultipleTuplesSpanWithinThreshold() throws Exception {
List<Tuple> tuples = new ArrayList<>();
tuples.add(JoinTestConstants.bookGroup1.get(3));
tuples.add(JoinTestConstants.bookGroup2.get(2));
tuples.add(JoinTestConstants.bookGroup2.get(4));
JoinTestHelper.insertToTable(BOOK_TABLE, tuples);
KeywordMatcherSourceOperator keywordSourceOuter = JoinTestHelper.getKeywordSource(BOOK_TABLE, "review", conjunction);
KeywordMatcherSourceOperator keywordSourceInner = JoinTestHelper.getKeywordSource(BOOK_TABLE, "book", conjunction);
List<Tuple> resultList = JoinTestHelper.getJoinDistanceResults(keywordSourceInner, keywordSourceOuter, new JoinDistancePredicate(JoinTestConstants.REVIEW, 12), Integer.MAX_VALUE, 0);
Schema resultSchema = Utils.createSpanSchema(JoinTestConstants.BOOK_SCHEMA);
List<Span> spanList = new ArrayList<>();
Span span1 = new Span(JoinTestConstants.REVIEW, 0, 16, "review_book", "Review of a " + "Book");
spanList.add(span1);
Span span2 = new Span(JoinTestConstants.REVIEW, 62, 73, "review_book", "book review");
spanList.add(span2);
Span span3 = new Span(JoinTestConstants.REVIEW, 235, 246, "review_book", "book review");
spanList.add(span3);
IField[] book1 = { new IntegerField(54), new StringField("Andria Williams"), new StringField("The Longest Night: A Novel"), new IntegerField(400), new TextField("Review of a Book. This is a typical " + "review. This is a test. A book review " + "test. A test to test queries without " + "actually using actual review. From " + "here onwards, we can pretend this to " + "be actually a review even if it is not " + "your typical book review."), new ListField<>(spanList) };
IField[] book2 = { new IntegerField(65), new StringField("Sharon Guskin"), new StringField("The Forgetting Time: A Novel"), new IntegerField(368), new TextField("Review of a Book. This is a typical " + "review. This is a test. A book review " + "test. A test to test queries without " + "actually using actual review. From " + "here onwards, we can pretend this to " + "be actually a review even if it is not " + "your typical book review."), new ListField<>(spanList) };
IField[] book3 = { new IntegerField(63), new StringField("Paul Kalanithi"), new StringField("When Breath Becomes Air"), new IntegerField(256), new TextField("Review of a Book. This is a typical " + "review. This is a test. A book review " + "test. A test to test queries without " + "actually using actual review. From " + "here onwards, we can pretend this to " + "be actually a review even if it is not " + "your typical book review."), new ListField<>(spanList) };
Tuple expectedTuple1 = new Tuple(resultSchema, book1);
Tuple expectedTuple2 = new Tuple(resultSchema, book2);
Tuple expectedTuple3 = new Tuple(resultSchema, book3);
List<Tuple> expectedResult = new ArrayList<>();
expectedResult.add(expectedTuple1);
expectedResult.add(expectedTuple2);
expectedResult.add(expectedTuple3);
Assert.assertEquals(3, resultList.size());
Assert.assertTrue(TestUtils.equals(expectedResult, resultList));
}
Aggregations