Examples with ObjectInspectorFactory.getStandardStructObjectInspector org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory.getStandardStructObjectInspector used on opensource projects

Example 96 with ObjectInspectorFactory.getStandardStructObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory.getStandardStructObjectInspector in project hive by apache.

the class TestStandardObjectInspectors method testStandardUnionObjectInspector.

@SuppressWarnings("unchecked")
@Test
public void testStandardUnionObjectInspector() throws Throwable {
    try {
        ArrayList<ObjectInspector> objectInspectors = new ArrayList<ObjectInspector>();
        // add primitive types
        objectInspectors.add(PrimitiveObjectInspectorFactory.javaIntObjectInspector);
        objectInspectors.add(PrimitiveObjectInspectorFactory.javaStringObjectInspector);
        objectInspectors.add(PrimitiveObjectInspectorFactory.javaBooleanObjectInspector);
        // add a list
        objectInspectors.add(ObjectInspectorFactory.getStandardListObjectInspector(PrimitiveObjectInspectorFactory.javaIntObjectInspector));
        // add a map
        objectInspectors.add(ObjectInspectorFactory.getStandardMapObjectInspector(PrimitiveObjectInspectorFactory.javaIntObjectInspector, PrimitiveObjectInspectorFactory.javaStringObjectInspector));
        // add a struct
        List<String> fieldNames = new ArrayList<String>();
        fieldNames.add("myDouble");
        fieldNames.add("myLong");
        ArrayList<ObjectInspector> fieldObjectInspectors = new ArrayList<ObjectInspector>();
        fieldObjectInspectors.add(PrimitiveObjectInspectorFactory.javaDoubleObjectInspector);
        fieldObjectInspectors.add(PrimitiveObjectInspectorFactory.javaLongObjectInspector);
        objectInspectors.add(ObjectInspectorFactory.getStandardStructObjectInspector(fieldNames, fieldObjectInspectors));
        StandardUnionObjectInspector uoi1 = ObjectInspectorFactory.getStandardUnionObjectInspector(objectInspectors);
        StandardUnionObjectInspector uoi2 = ObjectInspectorFactory.getStandardUnionObjectInspector((ArrayList<ObjectInspector>) objectInspectors.clone());
        assertEquals(uoi1, uoi2);
        assertEquals(ObjectInspectorUtils.getObjectInspectorName(uoi1), ObjectInspectorUtils.getObjectInspectorName(uoi2));
        assertTrue(ObjectInspectorUtils.compareTypes(uoi1, uoi2));
        // compareSupported returns false because Union can contain
        // an object of Map
        assertFalse(ObjectInspectorUtils.compareSupported(uoi1));
        // construct unionObjectInspector without Map field.
        ArrayList<ObjectInspector> ois = (ArrayList<ObjectInspector>) objectInspectors.clone();
        ois.set(4, PrimitiveObjectInspectorFactory.javaIntObjectInspector);
        assertTrue(ObjectInspectorUtils.compareSupported(ObjectInspectorFactory.getStandardUnionObjectInspector(ois)));
        // metadata
        assertEquals(Category.UNION, uoi1.getCategory());
        List<? extends ObjectInspector> uois = uoi1.getObjectInspectors();
        assertEquals(6, uois.size());
        for (int i = 0; i < 6; i++) {
            assertEquals(objectInspectors.get(i), uois.get(i));
        }
        StringBuilder unionTypeName = new StringBuilder();
        unionTypeName.append("uniontype<");
        for (int i = 0; i < uois.size(); i++) {
            if (i > 0) {
                unionTypeName.append(",");
            }
            unionTypeName.append(uois.get(i).getTypeName());
        }
        unionTypeName.append(">");
        assertEquals(unionTypeName.toString(), uoi1.getTypeName());
        // TypeInfo
        TypeInfo typeInfo1 = TypeInfoUtils.getTypeInfoFromObjectInspector(uoi1);
        assertEquals(Category.UNION, typeInfo1.getCategory());
        assertEquals(UnionTypeInfo.class.getName(), typeInfo1.getClass().getName());
        assertEquals(typeInfo1.getTypeName(), uoi1.getTypeName());
        assertEquals(typeInfo1, TypeInfoUtils.getTypeInfoFromTypeString(uoi1.getTypeName()));
        TypeInfo typeInfo2 = TypeInfoUtils.getTypeInfoFromObjectInspector(uoi2);
        assertEquals(typeInfo1, typeInfo2);
        assertEquals(TypeInfoUtils.getStandardJavaObjectInspectorFromTypeInfo(typeInfo1), TypeInfoUtils.getStandardJavaObjectInspectorFromTypeInfo(typeInfo2));
        assertEquals(TypeInfoUtils.getStandardWritableObjectInspectorFromTypeInfo(typeInfo1), TypeInfoUtils.getStandardWritableObjectInspectorFromTypeInfo(typeInfo2));
        // null
        assertNull(uoi1.getField(null));
        assertEquals(-1, uoi1.getTag(null));
        // Union
        UnionObject union = new StandardUnion((byte) 0, 1);
        assertEquals(0, uoi1.getTag(union));
        assertEquals(1, uoi1.getField(union));
        assertEquals("{0:1}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 0, 1), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(union));
        union = new StandardUnion((byte) 1, "two");
        assertEquals(1, uoi1.getTag(union));
        assertEquals("two", uoi1.getField(union));
        assertEquals("{1:\"two\"}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 1, "two"), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(union));
        union = new StandardUnion((byte) 2, true);
        assertEquals(2, uoi1.getTag(union));
        assertEquals(true, uoi1.getField(union));
        assertEquals("{2:true}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 2, true), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(union));
        ArrayList<Integer> iList = new ArrayList<Integer>();
        iList.add(4);
        iList.add(5);
        union = new StandardUnion((byte) 3, iList);
        assertEquals(3, uoi1.getTag(union));
        assertEquals(iList, uoi1.getField(union));
        assertEquals("{3:[4,5]}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 3, iList.clone()), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(union));
        HashMap<Integer, String> map = new LinkedHashMap<Integer, String>();
        map.put(6, "six");
        map.put(7, "seven");
        map.put(8, "eight");
        union = new StandardUnion((byte) 4, map);
        assertEquals(4, uoi1.getTag(union));
        assertEquals(map, uoi1.getField(union));
        assertEquals("{4:{6:\"six\",7:\"seven\",8:\"eight\"}}", SerDeUtils.getJSONString(union, uoi1));
        Throwable th = null;
        try {
            ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 4, map.clone()), uoi2, null);
        } catch (Throwable t) {
            th = t;
        }
        assertNotNull(th);
        assertEquals("Compare on map type not supported!", th.getMessage());
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(union));
        ArrayList<Object> struct = new ArrayList<Object>(2);
        struct.add(9.0);
        struct.add(10L);
        union = new StandardUnion((byte) 5, struct);
        assertEquals(5, uoi1.getTag(union));
        assertEquals(struct, uoi1.getField(union));
        assertEquals("{5:{\"mydouble\":9.0,\"mylong\":10}}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 5, struct.clone()), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(union));
    } catch (Throwable e) {
        e.printStackTrace();
        throw e;
    }
}

Example 97 with ObjectInspectorFactory.getStandardStructObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory.getStandardStructObjectInspector in project hive by apache.

the class SerdeRandomRowSource method chooseSchema.

private void chooseSchema(SupportedTypes supportedTypes, int maxComplexDepth) {
    Set<Integer> hashSet = null;
    final boolean allTypes;
    final boolean onlyOne = (r.nextInt(100) == 7);
    if (onlyOne) {
        columnCount = 1;
        allTypes = false;
    } else {
        allTypes = r.nextBoolean();
        if (allTypes) {
            switch(supportedTypes) {
                case ALL:
                    columnCount = possibleHivePrimitiveTypeNames.length + possibleHiveComplexTypeNames.length;
                    break;
                case ALL_EXCEPT_MAP:
                    columnCount = possibleHivePrimitiveTypeNames.length + possibleHiveComplexTypeNames.length - 1;
                    break;
                case PRIMITIVE:
                    columnCount = possibleHivePrimitiveTypeNames.length;
                    break;
            }
            hashSet = new HashSet<Integer>();
        } else {
            columnCount = 1 + r.nextInt(20);
        }
    }
    typeNames = new ArrayList<String>(columnCount);
    categories = new Category[columnCount];
    typeInfos = new TypeInfo[columnCount];
    objectInspectorList = new ArrayList<ObjectInspector>(columnCount);
    primitiveCategories = new PrimitiveCategory[columnCount];
    primitiveTypeInfos = new PrimitiveTypeInfo[columnCount];
    primitiveObjectInspectorList = new ArrayList<ObjectInspector>(columnCount);
    final List<String> columnNames = new ArrayList<String>(columnCount);
    for (int c = 0; c < columnCount; c++) {
        columnNames.add(String.format("col%d", c));
        String typeName;
        if (onlyOne) {
            typeName = getRandomTypeName(supportedTypes);
        } else {
            int typeNum;
            if (allTypes) {
                int maxTypeNum = 0;
                switch(supportedTypes) {
                    case PRIMITIVE:
                        maxTypeNum = possibleHivePrimitiveTypeNames.length;
                        break;
                    case ALL_EXCEPT_MAP:
                        maxTypeNum = possibleHivePrimitiveTypeNames.length + possibleHiveComplexTypeNames.length - 1;
                        break;
                    case ALL:
                        maxTypeNum = possibleHivePrimitiveTypeNames.length + possibleHiveComplexTypeNames.length;
                        break;
                }
                while (true) {
                    typeNum = r.nextInt(maxTypeNum);
                    final Integer typeNumInteger = Integer.valueOf(typeNum);
                    if (!hashSet.contains(typeNumInteger)) {
                        hashSet.add(typeNumInteger);
                        break;
                    }
                }
            } else {
                if (supportedTypes == SupportedTypes.PRIMITIVE || r.nextInt(10) != 0) {
                    typeNum = r.nextInt(possibleHivePrimitiveTypeNames.length);
                } else {
                    typeNum = possibleHivePrimitiveTypeNames.length + r.nextInt(possibleHiveComplexTypeNames.length);
                    if (supportedTypes == SupportedTypes.ALL_EXCEPT_MAP) {
                        typeNum--;
                    }
                }
            }
            if (typeNum < possibleHivePrimitiveTypeNames.length) {
                typeName = possibleHivePrimitiveTypeNames[typeNum];
            } else {
                typeName = possibleHiveComplexTypeNames[typeNum - possibleHivePrimitiveTypeNames.length];
            }
        }
        final String decoratedTypeName = getDecoratedTypeName(typeName, supportedTypes, 0, maxComplexDepth);
        final TypeInfo typeInfo;
        try {
            typeInfo = TypeInfoUtils.getTypeInfoFromTypeString(decoratedTypeName);
        } catch (Exception e) {
            throw new RuntimeException("Cannot convert type name " + decoratedTypeName + " to a type " + e);
        }
        typeInfos[c] = typeInfo;
        final Category category = typeInfo.getCategory();
        categories[c] = category;
        ObjectInspector objectInspector = getObjectInspector(typeInfo);
        switch(category) {
            case PRIMITIVE:
                {
                    final PrimitiveTypeInfo primitiveTypeInfo = (PrimitiveTypeInfo) typeInfo;
                    final PrimitiveCategory primitiveCategory = primitiveTypeInfo.getPrimitiveCategory();
                    objectInspector = PrimitiveObjectInspectorFactory.getPrimitiveWritableObjectInspector(primitiveTypeInfo);
                    primitiveTypeInfos[c] = primitiveTypeInfo;
                    primitiveCategories[c] = primitiveCategory;
                    primitiveObjectInspectorList.add(objectInspector);
                }
                break;
            case LIST:
            case MAP:
            case STRUCT:
            case UNION:
                primitiveObjectInspectorList.add(null);
                break;
            default:
                throw new RuntimeException("Unexpected category " + category);
        }
        objectInspectorList.add(objectInspector);
        if (category == Category.PRIMITIVE) {
        }
        typeNames.add(decoratedTypeName);
    }
    rowStructObjectInspector = ObjectInspectorFactory.getStandardStructObjectInspector(columnNames, objectInspectorList);
    alphabets = new String[columnCount];
}

Example 98 with ObjectInspectorFactory.getStandardStructObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory.getStandardStructObjectInspector in project DataX by alibaba.

the class HdfsHelper method orcFileStartWrite.

/**
 * 写orcfile类型文件
 * @param lineReceiver
 * @param config
 * @param fileName
 * @param taskPluginCollector
 */
public void orcFileStartWrite(RecordReceiver lineReceiver, Configuration config, String fileName, TaskPluginCollector taskPluginCollector) {
    List<Configuration> columns = config.getListConfiguration(Key.COLUMN);
    String compress = config.getString(Key.COMPRESS, null);
    List<String> columnNames = getColumnNames(columns);
    List<ObjectInspector> columnTypeInspectors = getColumnTypeInspectors(columns);
    StructObjectInspector inspector = (StructObjectInspector) ObjectInspectorFactory.getStandardStructObjectInspector(columnNames, columnTypeInspectors);
    OrcSerde orcSerde = new OrcSerde();
    FileOutputFormat outFormat = new OrcOutputFormat();
    if (!"NONE".equalsIgnoreCase(compress) && null != compress) {
        Class<? extends CompressionCodec> codecClass = getCompressCodec(compress);
        if (null != codecClass) {
            outFormat.setOutputCompressorClass(conf, codecClass);
        }
    }
    try {
        RecordWriter writer = outFormat.getRecordWriter(fileSystem, conf, fileName, Reporter.NULL);
        Record record = null;
        while ((record = lineReceiver.getFromReader()) != null) {
            MutablePair<List<Object>, Boolean> transportResult = transportOneRecord(record, columns, taskPluginCollector);
            if (!transportResult.getRight()) {
                writer.write(NullWritable.get(), orcSerde.serialize(transportResult.getLeft(), inspector));
            }
        }
        writer.close(Reporter.NULL);
    } catch (Exception e) {
        String message = String.format("写文件文件[%s]时发生IO异常,请检查您的网络是否正常！", fileName);
        LOG.error(message);
        Path path = new Path(fileName);
        deleteDir(path.getParent());
        throw DataXException.asDataXException(HdfsWriterErrorCode.Write_FILE_IO_ERROR, e);
    }
}

Example 99 with ObjectInspectorFactory.getStandardStructObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory.getStandardStructObjectInspector in project hadoop-pcap by RIPE-NCC.

the class PcapDeserializer method initialize.

@Override
public void initialize(Configuration cfg, Properties props) throws SerDeException {
    String columnNameProperty = props.getProperty(Constants.LIST_COLUMNS);
    columnNames = Arrays.asList(columnNameProperty.split(","));
    numColumns = columnNames.size();
    String columnTypeProperty = props.getProperty(Constants.LIST_COLUMN_TYPES);
    List<TypeInfo> columnTypes = TypeInfoUtils.getTypeInfosFromTypeString(columnTypeProperty);
    // Ensure we have the same number of column names and types
    assert numColumns == columnTypes.size();
    List<ObjectInspector> inspectors = new ArrayList<ObjectInspector>(numColumns);
    row = new ArrayList<Object>(numColumns);
    for (int c = 0; c < numColumns; c++) {
        ObjectInspector oi = TypeInfoUtils.getStandardJavaObjectInspectorFromTypeInfo(columnTypes.get(c));
        inspectors.add(oi);
        row.add(null);
    }
    inspector = ObjectInspectorFactory.getStandardStructObjectInspector(columnNames, inspectors);
}

Example 100 with ObjectInspectorFactory.getStandardStructObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory.getStandardStructObjectInspector in project flink by apache.

the class HiveInspectors method getObjectInspector.

public static ObjectInspector getObjectInspector(TypeInfo type) {
    switch(type.getCategory()) {
        case PRIMITIVE:
            PrimitiveTypeInfo primitiveType = (PrimitiveTypeInfo) type;
            return PrimitiveObjectInspectorFactory.getPrimitiveJavaObjectInspector(primitiveType);
        case LIST:
            ListTypeInfo listType = (ListTypeInfo) type;
            return ObjectInspectorFactory.getStandardListObjectInspector(getObjectInspector(listType.getListElementTypeInfo()));
        case MAP:
            MapTypeInfo mapType = (MapTypeInfo) type;
            return ObjectInspectorFactory.getStandardMapObjectInspector(getObjectInspector(mapType.getMapKeyTypeInfo()), getObjectInspector(mapType.getMapValueTypeInfo()));
        case STRUCT:
            StructTypeInfo structType = (StructTypeInfo) type;
            List<TypeInfo> fieldTypes = structType.getAllStructFieldTypeInfos();
            List<ObjectInspector> fieldInspectors = new ArrayList<ObjectInspector>();
            for (TypeInfo fieldType : fieldTypes) {
                fieldInspectors.add(getObjectInspector(fieldType));
            }
            return ObjectInspectorFactory.getStandardStructObjectInspector(structType.getAllStructFieldNames(), fieldInspectors);
        default:
            throw new CatalogException("Unsupported Hive type category " + type.getCategory());
    }
}