Examples with ColumnIndex org.apache.apex.malhar.lib.streamquery.index.ColumnIndex used on opensource projects

Example 1 with ColumnIndex

use of org.apache.apex.malhar.lib.streamquery.index.ColumnIndex in project apex-malhar by apache.

the class GroupByOperatorTest method testSqlGroupBy.

@SuppressWarnings({ "rawtypes", "unchecked" })
@Test
public void testSqlGroupBy() {
    // create operator
    GroupByHavingOperator oper = new GroupByHavingOperator();
    oper.addColumnGroupByIndex(new ColumnIndex("b", null));
    try {
        oper.addAggregateIndex(new SumFunction("c", null));
    } catch (Exception e) {
        // TODO Auto-generated catch block
        e.printStackTrace();
        return;
    }
    EqualValueCondition condition = new EqualValueCondition();
    condition.addEqualValue("a", 1);
    oper.setCondition(condition);
    CollectorTestSink sink = new CollectorTestSink();
    oper.outport.setSink(sink);
    oper.setup(null);
    oper.beginWindow(1);
    HashMap<String, Object> tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 1);
    tuple.put("c", 2);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 1);
    tuple.put("c", 4);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 2);
    tuple.put("c", 6);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 2);
    tuple.put("c", 7);
    oper.inport.process(tuple);
    oper.endWindow();
    oper.teardown();
    LOG.debug("{}", sink.collectedTuples);
}

Example 2 with ColumnIndex

use of org.apache.apex.malhar.lib.streamquery.index.ColumnIndex in project apex-malhar by apache.

the class InnerJoinOperatorTest method testSqlSelect.

@SuppressWarnings({ "rawtypes", "unchecked" })
@Test
public void testSqlSelect() {
    // create operator
    InnerJoinOperator oper = new InnerJoinOperator();
    CollectorTestSink sink = new CollectorTestSink();
    oper.outport.setSink(sink);
    // set column join condition
    Condition cond = new JoinColumnEqualCondition("a", "a");
    oper.setJoinCondition(cond);
    // add columns
    oper.selectTable1Column(new ColumnIndex("b", null));
    oper.selectTable2Column(new ColumnIndex("c", null));
    oper.setup(null);
    oper.beginWindow(1);
    HashMap<String, Object> tuple = new HashMap<String, Object>();
    tuple.put("a", 0);
    tuple.put("b", 1);
    tuple.put("c", 2);
    oper.inport1.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 3);
    tuple.put("c", 4);
    oper.inport1.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 0);
    tuple.put("b", 7);
    tuple.put("c", 8);
    oper.inport2.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 5);
    tuple.put("c", 6);
    oper.inport2.process(tuple);
    oper.endWindow();
    oper.teardown();
    LOG.debug("{}", sink.collectedTuples);
}

Example 3 with ColumnIndex

use of org.apache.apex.malhar.lib.streamquery.index.ColumnIndex in project apex-malhar by apache.

the class GroupByHavingOperator method endWindow.

/**
 * Create aggregate at end window.
 */
@Override
public void endWindow() {
    // group names
    if (columnGroupIndexes.size() == 0) {
        rows = new ArrayList<Map<String, Object>>();
        return;
    }
    // group rows
    HashMap<MultiKeyCompare, ArrayList<Map<String, Object>>> groups = new HashMap<MultiKeyCompare, ArrayList<Map<String, Object>>>();
    for (Map<String, Object> row : rows) {
        MultiKeyCompare key = new MultiKeyCompare();
        for (ColumnIndex index : columnGroupIndexes) {
            key.addCompareKey(row.get(index.getColumn()));
        }
        ArrayList<Map<String, Object>> subRows;
        if (groups.containsKey(key)) {
            subRows = groups.get(key);
        } else {
            subRows = new ArrayList<Map<String, Object>>();
            groups.put(key, subRows);
        }
        subRows.add(row);
    }
    // Iterate over groups and emit aggregate values
    for (Map.Entry<MultiKeyCompare, ArrayList<Map<String, Object>>> entry : groups.entrySet()) {
        ArrayList<Map<String, Object>> subRows = entry.getValue();
        // get result
        Map<String, Object> result = new HashMap<String, Object>();
        for (ColumnIndex index : columnGroupIndexes) {
            index.filter(subRows.get(0), result);
        }
        // append aggregate values
        for (FunctionIndex aggregate : aggregates) {
            try {
                aggregate.filter(subRows, result);
            } catch (Exception e) {
                e.printStackTrace();
            }
        }
        // check valid having aggregate
        boolean isValidHaving = true;
        for (HavingCondition condition : havingConditions) {
            try {
                isValidHaving &= condition.isValidAggregate(subRows);
            } catch (Exception e) {
                e.printStackTrace();
                return;
            }
        }
        if (isValidHaving) {
            outport.emit(result);
        }
    }
    rows = new ArrayList<Map<String, Object>>();
}

Example 4 with ColumnIndex

use of org.apache.apex.malhar.lib.streamquery.index.ColumnIndex in project apex-malhar by apache.

the class BetweenConditionTest method testSqlSelect.

@SuppressWarnings({ "rawtypes", "unchecked" })
@Test
public void testSqlSelect() {
    // create operator
    SelectOperator oper = new SelectOperator();
    oper.addIndex(new ColumnIndex("b", null));
    oper.addIndex(new ColumnIndex("c", null));
    BetweenCondition cond = new BetweenCondition("a", 0, 2);
    oper.setCondition(cond);
    CollectorTestSink sink = new CollectorTestSink();
    oper.outport.setSink(sink);
    oper.setup(null);
    oper.beginWindow(1);
    HashMap<String, Object> tuple = new HashMap<String, Object>();
    tuple.put("a", 0);
    tuple.put("b", 1);
    tuple.put("c", 2);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 3);
    tuple.put("c", 4);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 2);
    tuple.put("b", 5);
    tuple.put("c", 6);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 3);
    tuple.put("b", 7);
    tuple.put("c", 8);
    oper.inport.process(tuple);
    oper.endWindow();
    oper.teardown();
    LOG.debug("{}", sink.collectedTuples);
}

Example 5 with ColumnIndex

use of org.apache.apex.malhar.lib.streamquery.index.ColumnIndex in project apex-malhar by apache.

the class InConditionTest method testSqlSelect.

@SuppressWarnings({ "rawtypes", "unchecked" })
@Test
public void testSqlSelect() {
    // create operator
    SelectOperator oper = new SelectOperator();
    oper.addIndex(new ColumnIndex("b", null));
    oper.addIndex(new ColumnIndex("c", null));
    InCondition cond = new InCondition("a");
    cond.addInValue(0);
    cond.addInValue(1);
    oper.setCondition(cond);
    CollectorTestSink sink = new CollectorTestSink();
    oper.outport.setSink(sink);
    oper.setup(null);
    oper.beginWindow(1);
    HashMap<String, Object> tuple = new HashMap<String, Object>();
    tuple.put("a", 0);
    tuple.put("b", 1);
    tuple.put("c", 2);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 1);
    tuple.put("b", 3);
    tuple.put("c", 4);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 2);
    tuple.put("b", 5);
    tuple.put("c", 6);
    oper.inport.process(tuple);
    tuple = new HashMap<String, Object>();
    tuple.put("a", 3);
    tuple.put("b", 7);
    tuple.put("c", 8);
    oper.inport.process(tuple);
    oper.endWindow();
    oper.teardown();
    LOG.debug("{}", sink.collectedTuples);
}