Examples with JsonStringHashMap org.apache.drill.exec.util.JsonStringHashMap used on opensource projects

Example 1 with JsonStringHashMap

use of org.apache.drill.exec.util.JsonStringHashMap in project drill by apache.

the class TestBuilder method mapOf.

/**
   * Convenience method to create a {@link JsonStringHashMap<String, Object> map} instance with the given key value sequence.
   *
   * Key value sequence consists of key - value pairs such that a key precedes its value. For instance:
   *
   * mapOf("name", "Adam", "age", 41) corresponds to {"name": "Adam", "age": 41} in JSON.
   */
public static JsonStringHashMap<String, Object> mapOf(Object... keyValueSequence) {
    Preconditions.checkArgument(keyValueSequence.length % 2 == 0, "Length of key value sequence must be even");
    final JsonStringHashMap<String, Object> map = new JsonStringHashMap<>();
    for (int i = 0; i < keyValueSequence.length; i += 2) {
        Object value = keyValueSequence[i + 1];
        if (value instanceof CharSequence) {
            value = new Text(value.toString());
        }
        map.put(String.class.cast(keyValueSequence[i]), value);
    }
    return map;
}

Example 2 with JsonStringHashMap

use of org.apache.drill.exec.util.JsonStringHashMap in project drill by axbaretto.

the class TestOutputBatchSize method testFlattenListOfMaps.

@Test
public void testFlattenListOfMaps() throws Exception {
    PhysicalOperator flatten = new FlattenPOP(null, SchemaPath.getSimplePath("c"));
    mockOpContext(flatten, initReservation, maxAllocation);
    // create input rows like this.
    // "a" : 5, "b" : wideString,
    // "c" : [ [{"trans_id":"t1", amount:100, trans_time:7777777, type:sports}, {"trans_id":"t1", amount:100, trans_time:8888888, type:groceries}],
    // [{"trans_id":"t1", amount:100, trans_time:7777777, type:sports}, {"trans_id":"t1", amount:100, trans_time:8888888, type:groceries}],
    // [{"trans_id":"t1", amount:100, trans_time:7777777, type:sports}, {"trans_id":"t1", amount:100, trans_time:8888888, type:groceries}] ]
    List<String> inputJsonBatches = Lists.newArrayList();
    StringBuilder batchString = new StringBuilder();
    batchString.append("[");
    for (int i = 0; i < numRows; i++) {
        batchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : [" + "[ { \"trans_id\":\"t1\", \"amount\":100, \"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ], " + "[ { \"trans_id\":\"t1\", \"amount\":100, \"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ], " + "[ { \"trans_id\":\"t1\", \"amount\":100, " + "\"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ]");
        batchString.append("]},");
    }
    batchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : [" + "[ { \"trans_id\":\"t1\", \"amount\":100, \"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ], " + "[ { \"trans_id\":\"t1\", \"amount\":100, \"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ], " + "[ { \"trans_id\":\"t1\", \"amount\":100, " + "\"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ]");
    batchString.append("]}]");
    inputJsonBatches.add(batchString.toString());
    // Figure out what will be approximate total output size out of flatten for input above
    // We will use this sizing information to set output batch size so we can produce desired
    // number of batches that can be verified.
    // output rows will be like this.
    // "a" : 5, "b" : wideString, "c" : [{"trans_id":"t1", amount:100, trans_time:7777777, type:sports}, {"trans_id":"t1", amount:100, trans_time:8888888, type:groceries}]
    // "a" : 5, "b" : wideString, "c" : [{"trans_id":"t1", amount:100, trans_time:7777777, type:sports}, {"trans_id":"t1", amount:100, trans_time:8888888, type:groceries}]
    List<String> expectedJsonBatches = Lists.newArrayList();
    StringBuilder expectedBatchString = new StringBuilder();
    expectedBatchString.append("[");
    for (int i = 0; i < numRows; i++) {
        expectedBatchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : " + "[ { \"trans_id\":\"t1\", \"amount\":100, \"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ]},");
        expectedBatchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : " + "[ { \"trans_id\":\"t1\", \"amount\":100, " + "\"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"}]},");
        expectedBatchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : " + "[ { \"trans_id\":\"t1\", \"amount\":100, " + "\"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"}]},");
    }
    expectedBatchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : " + "[ { \"trans_id\":\"t1\", \"amount\":100, \"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"} ]},");
    expectedBatchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : " + "[ { \"trans_id\":\"t1\", \"amount\":100, " + "\"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"}]},");
    expectedBatchString.append("{\"a\": 5, " + "\"b\" : " + "\"" + wideString + "\"," + "\"c\" : " + "[ { \"trans_id\":\"t1\", \"amount\":100, " + "\"trans_time\":7777777, \"type\":\"sports\"}," + " { " + "\"trans_id\":\"t2\", \"amount\":1000, \"trans_time\":8888888, \"type\":\"groceries\"}]}");
    expectedBatchString.append("]");
    expectedJsonBatches.add(expectedBatchString.toString());
    long totalSize = getExpectedSize(expectedJsonBatches);
    // set the output batch size to 1/2 of total size expected.
    // We will get approximately get 2 batches and max of 4.
    fragContext.getOptions().setLocalOption("drill.exec.memory.operator.output_batch_size", totalSize / 2);
    OperatorTestBuilder opTestBuilder = opTestBuilder().physicalOperator(flatten).inputDataStreamJson(inputJsonBatches).baselineColumns("a", "b", "c").expectedNumBatches(// verify number of batches
    2).expectedBatchSize(// verify batch size.
    totalSize / 2);
    final JsonStringHashMap<String, Object> resultExpected1 = new JsonStringHashMap<>();
    resultExpected1.put("trans_id", new Text("t1"));
    resultExpected1.put("amount", new Long(100));
    resultExpected1.put("trans_time", new Long(7777777));
    resultExpected1.put("type", new Text("sports"));
    final JsonStringHashMap<String, Object> resultExpected2 = new JsonStringHashMap<>();
    resultExpected2.put("trans_id", new Text("t2"));
    resultExpected2.put("amount", new Long(1000));
    resultExpected2.put("trans_time", new Long(8888888));
    resultExpected2.put("type", new Text("groceries"));
    final JsonStringArrayList<JsonStringHashMap<String, Object>> results = new JsonStringArrayList<JsonStringHashMap<String, Object>>() {

        {
            add(resultExpected1);
            add(resultExpected2);
        }
    };
    for (int i = 0; i < numRows + 1; i++) {
        opTestBuilder.baselineValues(5l, wideString, results);
        opTestBuilder.baselineValues(5l, wideString, results);
        opTestBuilder.baselineValues(5l, wideString, results);
    }
    opTestBuilder.go();
}

Example 3 with JsonStringHashMap

use of org.apache.drill.exec.util.JsonStringHashMap in project drill by axbaretto.

the class TestJsonReader method testUntypedPathWithUnion.

// DRILL-6020
@Test
public void testUntypedPathWithUnion() throws Exception {
    String fileName = "table.json";
    try (BufferedWriter writer = new BufferedWriter(new FileWriter(new File(dirTestWatcher.getRootDir(), fileName)))) {
        writer.write("{\"rk\": {\"a\": {\"b\": \"1\"}}}");
        writer.write("{\"rk\": {\"a\": \"2\"}}");
    }
    JsonStringHashMap<String, Text> map = new JsonStringHashMap<>();
    map.put("b", new Text("1"));
    try {
        testBuilder().sqlQuery("select t.rk.a as a from dfs.`%s` t", fileName).ordered().optionSettingQueriesForTestQuery("alter session set `exec.enable_union_type`=true").baselineColumns("a").baselineValues(map).baselineValues("2").go();
    } finally {
        testNoResult("alter session reset `exec.enable_union_type`");
    }
}

Example 4 with JsonStringHashMap

use of org.apache.drill.exec.util.JsonStringHashMap in project drill by axbaretto.

the class TestBuilder method mapOf.

/**
 * Convenience method to create a {@link JsonStringHashMap<String, Object> map} instance with the given key value sequence.
 *
 * Key value sequence consists of key - value pairs such that a key precedes its value. For instance:
 *
 * mapOf("name", "Adam", "age", 41) corresponds to {"name": "Adam", "age": 41} in JSON.
 */
public static JsonStringHashMap<String, Object> mapOf(Object... keyValueSequence) {
    Preconditions.checkArgument(keyValueSequence.length % 2 == 0, "Length of key value sequence must be even");
    final JsonStringHashMap<String, Object> map = new JsonStringHashMap<>();
    for (int i = 0; i < keyValueSequence.length; i += 2) {
        Object value = keyValueSequence[i + 1];
        if (value instanceof CharSequence) {
            value = new Text(value.toString());
        }
        map.put(String.class.cast(keyValueSequence[i]), value);
    }
    return map;
}

Example 5 with JsonStringHashMap

use of org.apache.drill.exec.util.JsonStringHashMap in project drill by apache.

the class TestFlatten method testFlatten_Drill2162_complex.

@Test
@Category(UnlikelyTest.class)
public void testFlatten_Drill2162_complex() throws Exception {
    String jsonRecords = BaseTestQuery.getFile("flatten/complex_transaction_example_data.json");
    int numCopies = 700;
    new TestConstantFolding.SmallFileCreator(pathDir).setRecord(jsonRecords).createFiles(1, numCopies, "json");
    List<JsonStringHashMap<String, Object>> data = Lists.newArrayList(mapOf("uid", 1l, "lst_lst_0", listOf(1l, 2l, 3l, 4l, 5l), "lst_lst_1", listOf(2l, 3l, 4l, 5l, 6l), "lst_lst", listOf(listOf(1l, 2l, 3l, 4l, 5l), listOf(2l, 3l, 4l, 5l, 6l))), mapOf("uid", 2l, "lst_lst_0", listOf(1l, 2l, 3l, 4l, 5l), "lst_lst_1", listOf(2l, 3l, 4l, 5l, 6l), "lst_lst", listOf(listOf(1l, 2l, 3l, 4l, 5l), listOf(2l, 3l, 4l, 5l, 6l))));
    List<JsonStringHashMap<String, Object>> result = flatten(flatten(flatten(data, "lst_lst_1"), "lst_lst_0"), "lst_lst");
    TestBuilder builder = testBuilder().sqlQuery("select uid, flatten(d.lst_lst[1]) lst1, flatten(d.lst_lst[0]) lst0, flatten(d.lst_lst) lst from " + "dfs.`%s/bigfile/bigfile.json` d", TEST_DIR).unOrdered().baselineColumns("uid", "lst1", "lst0", "lst");
    for (int i = 0; i < numCopies; i++) {
        for (JsonStringHashMap<String, Object> record : result) {
            builder.baselineValues(record.get("uid"), record.get("lst_lst_1"), record.get("lst_lst_0"), record.get("lst_lst"));
        }
    }
    builder.go();
}