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

Example 61 with ListObjectInspector

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

the class LazyBinaryArray method parse.

/**
 * Parse the bytes and fill elementStart, elementLength, elementInited and
 * elementIsNull.
 */
private void parse() {
    byte[] bytes = this.bytes.getData();
    // get the vlong that represents the map size
    LazyBinaryUtils.readVInt(bytes, start, vInt);
    arraySize = vInt.value;
    if (0 == arraySize) {
        parsed = true;
        return;
    }
    // adjust arrays
    adjustArraySize(arraySize);
    // find out the null-bytes
    int arryByteStart = start + vInt.length;
    int nullByteCur = arryByteStart;
    int nullByteEnd = arryByteStart + (arraySize + 7) / 8;
    // the begin the real elements
    int lastElementByteEnd = nullByteEnd;
    // the list element object inspector
    ObjectInspector listEleObjectInspector = ((ListObjectInspector) oi).getListElementObjectInspector();
    // parsing elements one by one
    for (int i = 0; i < arraySize; i++) {
        elementIsNull[i] = true;
        if ((bytes[nullByteCur] & (1 << (i % 8))) != 0) {
            elementIsNull[i] = false;
            LazyBinaryUtils.checkObjectByteInfo(listEleObjectInspector, bytes, lastElementByteEnd, recordInfo, vInt);
            elementStart[i] = lastElementByteEnd + recordInfo.elementOffset;
            elementLength[i] = recordInfo.elementSize;
            lastElementByteEnd = elementStart[i] + elementLength[i];
        }
        // move onto the next null byte
        if (7 == (i % 8)) {
            nullByteCur++;
        }
    }
    Arrays.fill(elementInited, 0, arraySize, false);
    parsed = true;
}

Example 62 with ListObjectInspector

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

the class LazyBinarySerDe method serialize.

/**
 * A recursive function that serialize an object to a byte buffer based on its
 * object inspector.
 *
 * @param byteStream
 *          the byte stream storing the serialization data
 * @param obj
 *          the object to serialize
 * @param objInspector
 *          the object inspector
 * @param skipLengthPrefix a boolean indicating whether length prefix is
 *          needed for list/map/struct
 * @param warnedOnceNullMapKey a boolean indicating whether a warning
 *          has been issued once already when encountering null map keys
 */
public static void serialize(RandomAccessOutput byteStream, Object obj, ObjectInspector objInspector, boolean skipLengthPrefix, BooleanRef warnedOnceNullMapKey) throws SerDeException {
    // do nothing for null object
    if (null == obj) {
        return;
    }
    switch(objInspector.getCategory()) {
        case PRIMITIVE:
            {
                PrimitiveObjectInspector poi = (PrimitiveObjectInspector) objInspector;
                switch(poi.getPrimitiveCategory()) {
                    case VOID:
                        {
                            return;
                        }
                    case BOOLEAN:
                        {
                            boolean v = ((BooleanObjectInspector) poi).get(obj);
                            byteStream.write((byte) (v ? 1 : 0));
                            return;
                        }
                    case BYTE:
                        {
                            ByteObjectInspector boi = (ByteObjectInspector) poi;
                            byte v = boi.get(obj);
                            byteStream.write(v);
                            return;
                        }
                    case SHORT:
                        {
                            ShortObjectInspector spoi = (ShortObjectInspector) poi;
                            short v = spoi.get(obj);
                            byteStream.write((byte) (v >> 8));
                            byteStream.write((byte) (v));
                            return;
                        }
                    case INT:
                        {
                            IntObjectInspector ioi = (IntObjectInspector) poi;
                            int v = ioi.get(obj);
                            LazyBinaryUtils.writeVInt(byteStream, v);
                            return;
                        }
                    case LONG:
                        {
                            LongObjectInspector loi = (LongObjectInspector) poi;
                            long v = loi.get(obj);
                            LazyBinaryUtils.writeVLong(byteStream, v);
                            return;
                        }
                    case FLOAT:
                        {
                            FloatObjectInspector foi = (FloatObjectInspector) poi;
                            int v = Float.floatToIntBits(foi.get(obj));
                            byteStream.write((byte) (v >> 24));
                            byteStream.write((byte) (v >> 16));
                            byteStream.write((byte) (v >> 8));
                            byteStream.write((byte) (v));
                            return;
                        }
                    case DOUBLE:
                        {
                            DoubleObjectInspector doi = (DoubleObjectInspector) poi;
                            LazyBinaryUtils.writeDouble(byteStream, doi.get(obj));
                            return;
                        }
                    case STRING:
                        {
                            StringObjectInspector soi = (StringObjectInspector) poi;
                            Text t = soi.getPrimitiveWritableObject(obj);
                            serializeText(byteStream, t, skipLengthPrefix);
                            return;
                        }
                    case CHAR:
                        {
                            HiveCharObjectInspector hcoi = (HiveCharObjectInspector) poi;
                            Text t = hcoi.getPrimitiveWritableObject(obj).getTextValue();
                            serializeText(byteStream, t, skipLengthPrefix);
                            return;
                        }
                    case VARCHAR:
                        {
                            HiveVarcharObjectInspector hcoi = (HiveVarcharObjectInspector) poi;
                            Text t = hcoi.getPrimitiveWritableObject(obj).getTextValue();
                            serializeText(byteStream, t, skipLengthPrefix);
                            return;
                        }
                    case BINARY:
                        {
                            BinaryObjectInspector baoi = (BinaryObjectInspector) poi;
                            BytesWritable bw = baoi.getPrimitiveWritableObject(obj);
                            int length = bw.getLength();
                            if (!skipLengthPrefix) {
                                LazyBinaryUtils.writeVInt(byteStream, length);
                            } else {
                                if (length == 0) {
                                    throw new RuntimeException("LazyBinaryColumnarSerde cannot serialize a non-null zero " + "length binary field. Consider using either LazyBinarySerde or ColumnarSerde.");
                                }
                            }
                            byteStream.write(bw.getBytes(), 0, length);
                            return;
                        }
                    case DATE:
                        {
                            DateWritable d = ((DateObjectInspector) poi).getPrimitiveWritableObject(obj);
                            writeDateToByteStream(byteStream, d);
                            return;
                        }
                    case TIMESTAMP:
                        {
                            TimestampObjectInspector toi = (TimestampObjectInspector) poi;
                            TimestampWritable t = toi.getPrimitiveWritableObject(obj);
                            t.writeToByteStream(byteStream);
                            return;
                        }
                    case TIMESTAMPLOCALTZ:
                        {
                            TimestampLocalTZWritable t = ((TimestampLocalTZObjectInspector) poi).getPrimitiveWritableObject(obj);
                            t.writeToByteStream(byteStream);
                            return;
                        }
                    case INTERVAL_YEAR_MONTH:
                        {
                            HiveIntervalYearMonthWritable intervalYearMonth = ((HiveIntervalYearMonthObjectInspector) poi).getPrimitiveWritableObject(obj);
                            intervalYearMonth.writeToByteStream(byteStream);
                            return;
                        }
                    case INTERVAL_DAY_TIME:
                        {
                            HiveIntervalDayTimeWritable intervalDayTime = ((HiveIntervalDayTimeObjectInspector) poi).getPrimitiveWritableObject(obj);
                            intervalDayTime.writeToByteStream(byteStream);
                            return;
                        }
                    case DECIMAL:
                        {
                            HiveDecimalObjectInspector bdoi = (HiveDecimalObjectInspector) poi;
                            HiveDecimalWritable t = bdoi.getPrimitiveWritableObject(obj);
                            if (t == null) {
                                return;
                            }
                            writeToByteStream(byteStream, t);
                            return;
                        }
                    default:
                        {
                            throw new RuntimeException("Unrecognized type: " + poi.getPrimitiveCategory());
                        }
                }
            }
        case LIST:
            {
                ListObjectInspector loi = (ListObjectInspector) objInspector;
                ObjectInspector eoi = loi.getListElementObjectInspector();
                int byteSizeStart = 0;
                int listStart = 0;
                if (!skipLengthPrefix) {
                    // 1/ reserve spaces for the byte size of the list
                    // which is a integer and takes four bytes
                    byteSizeStart = byteStream.getLength();
                    byteStream.reserve(4);
                    listStart = byteStream.getLength();
                }
                // 2/ write the size of the list as a VInt
                int size = loi.getListLength(obj);
                LazyBinaryUtils.writeVInt(byteStream, size);
                // 3/ write the null bytes
                byte nullByte = 0;
                for (int eid = 0; eid < size; eid++) {
                    // set the bit to 1 if an element is not null
                    if (null != loi.getListElement(obj, eid)) {
                        nullByte |= 1 << (eid % 8);
                    }
                    // if this is the last element
                    if (7 == eid % 8 || eid == size - 1) {
                        byteStream.write(nullByte);
                        nullByte = 0;
                    }
                }
                // 4/ write element by element from the list
                for (int eid = 0; eid < size; eid++) {
                    serialize(byteStream, loi.getListElement(obj, eid), eoi, false, warnedOnceNullMapKey);
                }
                if (!skipLengthPrefix) {
                    // 5/ update the list byte size
                    int listEnd = byteStream.getLength();
                    int listSize = listEnd - listStart;
                    writeSizeAtOffset(byteStream, byteSizeStart, listSize);
                }
                return;
            }
        case MAP:
            {
                MapObjectInspector moi = (MapObjectInspector) objInspector;
                ObjectInspector koi = moi.getMapKeyObjectInspector();
                ObjectInspector voi = moi.getMapValueObjectInspector();
                Map<?, ?> map = moi.getMap(obj);
                int byteSizeStart = 0;
                int mapStart = 0;
                if (!skipLengthPrefix) {
                    // 1/ reserve spaces for the byte size of the map
                    // which is a integer and takes four bytes
                    byteSizeStart = byteStream.getLength();
                    byteStream.reserve(4);
                    mapStart = byteStream.getLength();
                }
                // 2/ write the size of the map which is a VInt
                int size = map.size();
                LazyBinaryUtils.writeVInt(byteStream, size);
                // 3/ write the null bytes
                int b = 0;
                byte nullByte = 0;
                for (Map.Entry<?, ?> entry : map.entrySet()) {
                    // set the bit to 1 if a key is not null
                    if (null != entry.getKey()) {
                        nullByte |= 1 << (b % 8);
                    } else if (warnedOnceNullMapKey != null) {
                        if (!warnedOnceNullMapKey.value) {
                            LOG.warn("Null map key encountered! Ignoring similar problems.");
                        }
                        warnedOnceNullMapKey.value = true;
                    }
                    b++;
                    // set the bit to 1 if a value is not null
                    if (null != entry.getValue()) {
                        nullByte |= 1 << (b % 8);
                    }
                    b++;
                    // or if this is the last key-value pair
                    if (0 == b % 8 || b == size * 2) {
                        byteStream.write(nullByte);
                        nullByte = 0;
                    }
                }
                // 4/ write key-value pairs one by one
                for (Map.Entry<?, ?> entry : map.entrySet()) {
                    serialize(byteStream, entry.getKey(), koi, false, warnedOnceNullMapKey);
                    serialize(byteStream, entry.getValue(), voi, false, warnedOnceNullMapKey);
                }
                if (!skipLengthPrefix) {
                    // 5/ update the byte size of the map
                    int mapEnd = byteStream.getLength();
                    int mapSize = mapEnd - mapStart;
                    writeSizeAtOffset(byteStream, byteSizeStart, mapSize);
                }
                return;
            }
        case STRUCT:
        case UNION:
            {
                int byteSizeStart = 0;
                int typeStart = 0;
                if (!skipLengthPrefix) {
                    // 1/ reserve spaces for the byte size of the struct
                    // which is a integer and takes four bytes
                    byteSizeStart = byteStream.getLength();
                    byteStream.reserve(4);
                    typeStart = byteStream.getLength();
                }
                if (ObjectInspector.Category.STRUCT.equals(objInspector.getCategory())) {
                    // 2/ serialize the struct
                    serializeStruct(byteStream, obj, (StructObjectInspector) objInspector, warnedOnceNullMapKey);
                } else {
                    // 2/ serialize the union
                    serializeUnion(byteStream, obj, (UnionObjectInspector) objInspector, warnedOnceNullMapKey);
                }
                if (!skipLengthPrefix) {
                    // 3/ update the byte size of the struct
                    int typeEnd = byteStream.getLength();
                    int typeSize = typeEnd - typeStart;
                    writeSizeAtOffset(byteStream, byteSizeStart, typeSize);
                }
                return;
            }
        default:
            {
                throw new RuntimeException("Unrecognized type: " + objInspector.getCategory());
            }
    }
}

Example 63 with ListObjectInspector

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

the class ObjectInspectorUtils method hashCode.

public static int hashCode(Object o, ObjectInspector objIns) {
    if (o == null) {
        return 0;
    }
    switch(objIns.getCategory()) {
        case PRIMITIVE:
            {
                PrimitiveObjectInspector poi = ((PrimitiveObjectInspector) objIns);
                switch(poi.getPrimitiveCategory()) {
                    case VOID:
                        return 0;
                    case BOOLEAN:
                        return ((BooleanObjectInspector) poi).get(o) ? 1 : 0;
                    case BYTE:
                        return ((ByteObjectInspector) poi).get(o);
                    case SHORT:
                        return ((ShortObjectInspector) poi).get(o);
                    case INT:
                        return ((IntObjectInspector) poi).get(o);
                    case LONG:
                        {
                            long a = ((LongObjectInspector) poi).get(o);
                            return (int) ((a >>> 32) ^ a);
                        }
                    case FLOAT:
                        return Float.floatToIntBits(((FloatObjectInspector) poi).get(o));
                    case DOUBLE:
                        {
                            // This hash function returns the same result as Double.hashCode()
                            // while DoubleWritable.hashCode returns a different result.
                            long a = Double.doubleToLongBits(((DoubleObjectInspector) poi).get(o));
                            return (int) ((a >>> 32) ^ a);
                        }
                    case STRING:
                        {
                            // This hash function returns the same result as String.hashCode() when
                            // all characters are ASCII, while Text.hashCode() always returns a
                            // different result.
                            Text t = ((StringObjectInspector) poi).getPrimitiveWritableObject(o);
                            int r = 0;
                            for (int i = 0; i < t.getLength(); i++) {
                                r = r * 31 + t.getBytes()[i];
                            }
                            return r;
                        }
                    case CHAR:
                        return ((HiveCharObjectInspector) poi).getPrimitiveWritableObject(o).hashCode();
                    case VARCHAR:
                        return ((HiveVarcharObjectInspector) poi).getPrimitiveWritableObject(o).hashCode();
                    case BINARY:
                        return ((BinaryObjectInspector) poi).getPrimitiveWritableObject(o).hashCode();
                    case DATE:
                        return ((DateObjectInspector) poi).getPrimitiveWritableObject(o).hashCode();
                    case TIMESTAMP:
                        TimestampWritable t = ((TimestampObjectInspector) poi).getPrimitiveWritableObject(o);
                        return t.hashCode();
                    case TIMESTAMPLOCALTZ:
                        TimestampLocalTZWritable tstz = ((TimestampLocalTZObjectInspector) poi).getPrimitiveWritableObject(o);
                        return tstz.hashCode();
                    case INTERVAL_YEAR_MONTH:
                        HiveIntervalYearMonthWritable intervalYearMonth = ((HiveIntervalYearMonthObjectInspector) poi).getPrimitiveWritableObject(o);
                        return intervalYearMonth.hashCode();
                    case INTERVAL_DAY_TIME:
                        HiveIntervalDayTimeWritable intervalDayTime = ((HiveIntervalDayTimeObjectInspector) poi).getPrimitiveWritableObject(o);
                        return intervalDayTime.hashCode();
                    case DECIMAL:
                        // compatible hash code.
                        return ((HiveDecimalObjectInspector) poi).getPrimitiveWritableObject(o).hashCode();
                    default:
                        {
                            throw new RuntimeException("Unknown type: " + poi.getPrimitiveCategory());
                        }
                }
            }
        case LIST:
            {
                int r = 0;
                ListObjectInspector listOI = (ListObjectInspector) objIns;
                ObjectInspector elemOI = listOI.getListElementObjectInspector();
                for (int ii = 0; ii < listOI.getListLength(o); ++ii) {
                    r = 31 * r + hashCode(listOI.getListElement(o, ii), elemOI);
                }
                return r;
            }
        case MAP:
            {
                int r = 0;
                MapObjectInspector mapOI = (MapObjectInspector) objIns;
                ObjectInspector keyOI = mapOI.getMapKeyObjectInspector();
                ObjectInspector valueOI = mapOI.getMapValueObjectInspector();
                Map<?, ?> map = mapOI.getMap(o);
                for (Map.Entry<?, ?> entry : map.entrySet()) {
                    r += hashCode(entry.getKey(), keyOI) ^ hashCode(entry.getValue(), valueOI);
                }
                return r;
            }
        case STRUCT:
            int r = 0;
            StructObjectInspector structOI = (StructObjectInspector) objIns;
            List<? extends StructField> fields = structOI.getAllStructFieldRefs();
            for (StructField field : fields) {
                r = 31 * r + hashCode(structOI.getStructFieldData(o, field), field.getFieldObjectInspector());
            }
            return r;
        case UNION:
            UnionObjectInspector uOI = (UnionObjectInspector) objIns;
            byte tag = uOI.getTag(o);
            return hashCode(uOI.getField(o), uOI.getObjectInspectors().get(tag));
        default:
            throw new RuntimeException("Unknown type: " + objIns.getTypeName());
    }
}

Example 64 with ListObjectInspector

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

the class ObjectInspectorUtils method compareSupported.

/**
 * Whether comparison is supported for this type.
 * Currently all types that references any map are not comparable.
 */
public static boolean compareSupported(ObjectInspector oi) {
    switch(oi.getCategory()) {
        case PRIMITIVE:
            return true;
        case LIST:
            ListObjectInspector loi = (ListObjectInspector) oi;
            return compareSupported(loi.getListElementObjectInspector());
        case STRUCT:
            StructObjectInspector soi = (StructObjectInspector) oi;
            List<? extends StructField> fields = soi.getAllStructFieldRefs();
            for (int f = 0; f < fields.size(); f++) {
                if (!compareSupported(fields.get(f).getFieldObjectInspector())) {
                    return false;
                }
            }
            return true;
        case MAP:
            return false;
        case UNION:
            UnionObjectInspector uoi = (UnionObjectInspector) oi;
            for (ObjectInspector eoi : uoi.getObjectInspectors()) {
                if (!compareSupported(eoi)) {
                    return false;
                }
            }
            return true;
        default:
            return false;
    }
}

Example 65 with ListObjectInspector

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

the class LazySimpleSerDe method serialize.

/**
 * Serialize the row into the StringBuilder.
 *
 * @param out
 *          The StringBuilder to store the serialized data.
 * @param obj
 *          The object for the current field.
 * @param objInspector
 *          The ObjectInspector for the current Object.
 * @param separators
 *          The separators array.
 * @param level
 *          The current level of separator.
 * @param nullSequence
 *          The byte sequence representing the NULL value.
 * @param escaped
 *          Whether we need to escape the data when writing out
 * @param escapeChar
 *          Which char to use as the escape char, e.g. '\\'
 * @param needsEscape
 *          Which byte needs to be escaped for 256 bytes.
 * @throws IOException
 * @throws SerDeException
 */
public static void serialize(ByteStream.Output out, Object obj, ObjectInspector objInspector, byte[] separators, int level, Text nullSequence, boolean escaped, byte escapeChar, boolean[] needsEscape) throws IOException, SerDeException {
    if (obj == null) {
        out.write(nullSequence.getBytes(), 0, nullSequence.getLength());
        return;
    }
    char separator;
    List<?> list;
    switch(objInspector.getCategory()) {
        case PRIMITIVE:
            LazyUtils.writePrimitiveUTF8(out, obj, (PrimitiveObjectInspector) objInspector, escaped, escapeChar, needsEscape);
            return;
        case LIST:
            separator = (char) LazyUtils.getSeparator(separators, level);
            ListObjectInspector loi = (ListObjectInspector) objInspector;
            list = loi.getList(obj);
            ObjectInspector eoi = loi.getListElementObjectInspector();
            if (list == null) {
                out.write(nullSequence.getBytes(), 0, nullSequence.getLength());
            } else {
                for (int i = 0; i < list.size(); i++) {
                    if (i > 0) {
                        out.write(separator);
                    }
                    serialize(out, list.get(i), eoi, separators, level + 1, nullSequence, escaped, escapeChar, needsEscape);
                }
            }
            return;
        case MAP:
            separator = (char) LazyUtils.getSeparator(separators, level);
            char keyValueSeparator = (char) LazyUtils.getSeparator(separators, level + 1);
            MapObjectInspector moi = (MapObjectInspector) objInspector;
            ObjectInspector koi = moi.getMapKeyObjectInspector();
            ObjectInspector voi = moi.getMapValueObjectInspector();
            Map<?, ?> map = moi.getMap(obj);
            if (map == null) {
                out.write(nullSequence.getBytes(), 0, nullSequence.getLength());
            } else {
                boolean first = true;
                for (Map.Entry<?, ?> entry : map.entrySet()) {
                    if (first) {
                        first = false;
                    } else {
                        out.write(separator);
                    }
                    serialize(out, entry.getKey(), koi, separators, level + 2, nullSequence, escaped, escapeChar, needsEscape);
                    out.write(keyValueSeparator);
                    serialize(out, entry.getValue(), voi, separators, level + 2, nullSequence, escaped, escapeChar, needsEscape);
                }
            }
            return;
        case STRUCT:
            separator = (char) LazyUtils.getSeparator(separators, level);
            StructObjectInspector soi = (StructObjectInspector) objInspector;
            List<? extends StructField> fields = soi.getAllStructFieldRefs();
            list = soi.getStructFieldsDataAsList(obj);
            if (list == null) {
                out.write(nullSequence.getBytes(), 0, nullSequence.getLength());
            } else {
                for (int i = 0; i < list.size(); i++) {
                    if (i > 0) {
                        out.write(separator);
                    }
                    serialize(out, list.get(i), fields.get(i).getFieldObjectInspector(), separators, level + 1, nullSequence, escaped, escapeChar, needsEscape);
                }
            }
            return;
        case UNION:
            separator = (char) LazyUtils.getSeparator(separators, level);
            UnionObjectInspector uoi = (UnionObjectInspector) objInspector;
            List<? extends ObjectInspector> ois = uoi.getObjectInspectors();
            if (ois == null) {
                out.write(nullSequence.getBytes(), 0, nullSequence.getLength());
            } else {
                LazyUtils.writePrimitiveUTF8(out, new Byte(uoi.getTag(obj)), PrimitiveObjectInspectorFactory.javaByteObjectInspector, escaped, escapeChar, needsEscape);
                out.write(separator);
                serialize(out, uoi.getField(obj), ois.get(uoi.getTag(obj)), separators, level + 1, nullSequence, escaped, escapeChar, needsEscape);
            }
            return;
        default:
            break;
    }
    throw new RuntimeException("Unknown category type: " + objInspector.getCategory());
}