public TimestampWritable evaluate(TimestampWritable t) {
if (t == null) {
return null;
}
long newTimestamp = granularity.truncate(t.getTimestamp().getTime());
result.setTime(newTimestamp);
return result;
}
private void compareToUDFUnixTimeStampDate(long t, long y) {
TimestampWritable tsw = toTimestampWritable(t);
LongWritable res = getLongWritable(tsw);
if (res.get() != y) {
System.out.printf(
"%d vs %d for %d, %d\n", res.get(), y, t, tsw.getTimestamp().getTime() / 1000);
}
Assert.assertEquals(res.get(), y);
}
public int compareTo(TimestampWritable t) {
checkBytes();
long s1 = this.getSeconds();
long s2 = t.getSeconds();
if (s1 == s2) {
int n1 = this.getNanos();
int n2 = t.getNanos();
if (n1 == n2) {
return 0;
}
return n1 - n2;
} else {
return s1 < s2 ? -1 : 1;
}
}
public void readFields(DataInput in) throws IOException {
in.readFully(internalBytes, 0, 4);
if (TimestampWritable.hasDecimalOrSecondVInt(internalBytes[0])) {
in.readFully(internalBytes, 4, 1);
int len = (byte) WritableUtils.decodeVIntSize(internalBytes[4]);
if (len > 1) {
in.readFully(internalBytes, 5, len - 1);
}
long vlong = LazyBinaryUtils.readVLongFromByteArray(internalBytes, 4);
if (vlong < -1000000000 || vlong > 999999999) {
throw new IOException(
"Invalid first vint value (encoded nanoseconds) of a TimestampWritable: "
+ vlong
+ ", expected to be between -1000000000 and 999999999.");
// Note that -1000000000 is a valid value corresponding to a nanosecond timestamp
// of 999999999, because if the second VInt is present, we use the value
// (-reversedNanoseconds - 1) as the second VInt.
}
if (vlong < 0) {
// This indicates there is a second VInt containing the additional bits of the seconds
// field.
in.readFully(internalBytes, 4 + len, 1);
int secondVIntLen = (byte) WritableUtils.decodeVIntSize(internalBytes[4 + len]);
if (secondVIntLen > 1) {
in.readFully(internalBytes, 5 + len, secondVIntLen - 1);
}
}
}
currentBytes = internalBytes;
this.offset = 0;
}
/**
* Convert from Timestamp to an integer. This is called for CAST(... AS INT)
*
* @param i The Timestamp value to convert
* @return IntWritable
*/
public IntWritable evaluate(TimestampWritable i) {
if (i == null) {
return null;
} else {
intWritable.set(i.getSeconds());
return intWritable;
}
}
public static void setTimestamp(Timestamp t, byte[] bytes, int offset) {
boolean hasDecimalOrSecondVInt = hasDecimalOrSecondVInt(bytes[offset]);
long seconds = (long) TimestampWritable.getSeconds(bytes, offset);
int nanos = 0;
if (hasDecimalOrSecondVInt) {
nanos = TimestampWritable.getNanos(bytes, offset + 4);
if (hasSecondVInt(bytes[offset + 4])) {
seconds +=
LazyBinaryUtils.readVLongFromByteArray(
bytes, offset + 4 + WritableUtils.decodeVIntSize(bytes[offset + 4]));
}
}
t.setTime(seconds * 1000);
if (nanos != 0) {
t.setNanos(nanos);
}
}
public IntWritable evaluate(TimestampWritable t) {
if (t == null) {
return null;
}
calendar.setTime(t.getTimestamp());
result.set(1 + calendar.get(Calendar.MONTH));
return result;
}
private LongWritable getLongWritable(TimestampWritable i) {
LongWritable result = new LongWritable();
if (i == null) {
return null;
} else {
result.set(i.getSeconds());
return result;
}
}
/** @return nanoseconds in this TimestampWritable */
public int getNanos() {
if (!timestampEmpty) {
return timestamp.getNanos();
} else if (!bytesEmpty) {
return hasDecimalOrSecondVInt() ? TimestampWritable.getNanos(currentBytes, offset + 4) : 0;
} else {
throw new IllegalStateException("Both timestamp and bytes are empty");
}
}
/** @return seconds corresponding to this TimestampWritable */
public long getSeconds() {
if (!timestampEmpty) {
return millisToSeconds(timestamp.getTime());
} else if (!bytesEmpty) {
return TimestampWritable.getSeconds(currentBytes, offset);
} else {
throw new IllegalStateException("Both timestamp and bytes are empty");
}
}
@Test
public void testTimestampPlusIntervalDayTime() throws Exception {
GenericUDFOPPlus udf = new GenericUDFOPPlus();
TimestampWritable left = new TimestampWritable(Timestamp.valueOf("2001-01-01 00:00:00"));
HiveIntervalDayTimeWritable right =
new HiveIntervalDayTimeWritable(HiveIntervalDayTime.valueOf("1 2:3:4.567"));
ObjectInspector[] inputOIs = {
PrimitiveObjectInspectorFactory.writableTimestampObjectInspector,
PrimitiveObjectInspectorFactory.writableHiveIntervalDayTimeObjectInspector
};
DeferredObject[] args = {
new DeferredJavaObject(left), new DeferredJavaObject(right),
};
PrimitiveObjectInspector oi = (PrimitiveObjectInspector) udf.initialize(inputOIs);
Assert.assertEquals(TypeInfoFactory.timestampTypeInfo, oi.getTypeInfo());
TimestampWritable res = (TimestampWritable) udf.evaluate(args);
Assert.assertEquals(Timestamp.valueOf("2001-01-02 2:3:4.567"), res.getTimestamp());
}
public void set(TimestampWritable t) {
if (t.bytesEmpty) {
set(t.getTimestamp());
return;
}
if (t.currentBytes == t.externalBytes) {
set(t.currentBytes, t.offset);
} else {
set(t.currentBytes, 0);
}
}
@Test
public void testTimestampPlusIntervalYearMonth() throws Exception {
GenericUDFOPPlus udf = new GenericUDFOPPlus();
TimestampWritable left =
new TimestampWritable(Timestamp.valueOf("2001-11-15 01:02:03.123456789"));
HiveIntervalYearMonthWritable right =
new HiveIntervalYearMonthWritable(HiveIntervalYearMonth.valueOf("2-2"));
ObjectInspector[] inputOIs = {
PrimitiveObjectInspectorFactory.writableTimestampObjectInspector,
PrimitiveObjectInspectorFactory.writableHiveIntervalYearMonthObjectInspector
};
DeferredObject[] args = {
new DeferredJavaObject(left), new DeferredJavaObject(right),
};
PrimitiveObjectInspector oi = (PrimitiveObjectInspector) udf.initialize(inputOIs);
Assert.assertEquals(TypeInfoFactory.timestampTypeInfo, oi.getTypeInfo());
TimestampWritable res = (TimestampWritable) udf.evaluate(args);
Assert.assertEquals(Timestamp.valueOf("2004-01-15 01:02:03.123456789"), res.getTimestamp());
}
/**
* Check a particular field and set its size and offset in bytes based on the field type and the
* bytes arrays.
*
* <p>For void, boolean, byte, short, int, long, float and double, there is no offset and the size
* is fixed. For string, map, list, struct, the first four bytes are used to store the size. So
* the offset is 4 and the size is computed by concating the first four bytes together. The first
* four bytes are defined with respect to the offset in the bytes arrays. For timestamp, if the
* first bit is 0, the record length is 4, otherwise a VInt begins at the 5th byte and its length
* is added to 4.
*
* @param objectInspector object inspector of the field
* @param bytes bytes arrays store the table row
* @param offset offset of this field
* @param recordInfo modify this byteinfo object and return it
*/
public static void checkObjectByteInfo(
ObjectInspector objectInspector, byte[] bytes, int offset, RecordInfo recordInfo, VInt vInt) {
Category category = objectInspector.getCategory();
switch (category) {
case PRIMITIVE:
PrimitiveCategory primitiveCategory =
((PrimitiveObjectInspector) objectInspector).getPrimitiveCategory();
switch (primitiveCategory) {
case VOID:
recordInfo.elementOffset = 0;
recordInfo.elementSize = 0;
break;
case BOOLEAN:
case BYTE:
recordInfo.elementOffset = 0;
recordInfo.elementSize = 1;
break;
case SHORT:
recordInfo.elementOffset = 0;
recordInfo.elementSize = 2;
break;
case FLOAT:
recordInfo.elementOffset = 0;
recordInfo.elementSize = 4;
break;
case DOUBLE:
recordInfo.elementOffset = 0;
recordInfo.elementSize = 8;
break;
case INT:
recordInfo.elementOffset = 0;
recordInfo.elementSize = WritableUtils.decodeVIntSize(bytes[offset]);
break;
case LONG:
recordInfo.elementOffset = 0;
recordInfo.elementSize = WritableUtils.decodeVIntSize(bytes[offset]);
break;
case STRING:
// using vint instead of 4 bytes
LazyBinaryUtils.readVInt(bytes, offset, vInt);
recordInfo.elementOffset = vInt.length;
recordInfo.elementSize = vInt.value;
break;
case CHAR:
case VARCHAR:
LazyBinaryUtils.readVInt(bytes, offset, vInt);
recordInfo.elementOffset = vInt.length;
recordInfo.elementSize = vInt.value;
break;
case BINARY:
// using vint instead of 4 bytes
LazyBinaryUtils.readVInt(bytes, offset, vInt);
recordInfo.elementOffset = vInt.length;
recordInfo.elementSize = vInt.value;
break;
case DATE:
recordInfo.elementOffset = 0;
recordInfo.elementSize = WritableUtils.decodeVIntSize(bytes[offset]);
break;
case TIMESTAMP:
recordInfo.elementOffset = 0;
recordInfo.elementSize = TimestampWritable.getTotalLength(bytes, offset);
break;
case DECIMAL:
// using vint instead of 4 bytes
LazyBinaryUtils.readVInt(bytes, offset, vInt);
recordInfo.elementOffset = 0;
recordInfo.elementSize = vInt.length;
LazyBinaryUtils.readVInt(bytes, offset + vInt.length, vInt);
recordInfo.elementSize += vInt.length + vInt.value;
break;
default:
{
throw new RuntimeException("Unrecognized primitive type: " + primitiveCategory);
}
}
break;
case LIST:
case MAP:
case STRUCT:
case UNION:
recordInfo.elementOffset = 4;
recordInfo.elementSize = LazyBinaryUtils.byteArrayToInt(bytes, offset);
break;
default:
{
throw new RuntimeException("Unrecognized non-primitive type: " + category);
}
}
}
static void serialize(OutputByteBuffer buffer, Object o, ObjectInspector oi, boolean invert) {
// Is this field a null?
if (o == null) {
buffer.write((byte) 0, invert);
return;
}
// This field is not a null.
buffer.write((byte) 1, invert);
switch (oi.getCategory()) {
case PRIMITIVE:
{
PrimitiveObjectInspector poi = (PrimitiveObjectInspector) oi;
switch (poi.getPrimitiveCategory()) {
case VOID:
{
return;
}
case BOOLEAN:
{
boolean v = ((BooleanObjectInspector) poi).get(o);
buffer.write((byte) (v ? 2 : 1), invert);
return;
}
case BYTE:
{
ByteObjectInspector boi = (ByteObjectInspector) poi;
byte v = boi.get(o);
buffer.write((byte) (v ^ 0x80), invert);
return;
}
case SHORT:
{
ShortObjectInspector spoi = (ShortObjectInspector) poi;
short v = spoi.get(o);
buffer.write((byte) ((v >> 8) ^ 0x80), invert);
buffer.write((byte) v, invert);
return;
}
case INT:
{
IntObjectInspector ioi = (IntObjectInspector) poi;
int v = ioi.get(o);
buffer.write((byte) ((v >> 24) ^ 0x80), invert);
buffer.write((byte) (v >> 16), invert);
buffer.write((byte) (v >> 8), invert);
buffer.write((byte) v, invert);
return;
}
case LONG:
{
LongObjectInspector loi = (LongObjectInspector) poi;
long v = loi.get(o);
buffer.write((byte) ((v >> 56) ^ 0x80), invert);
buffer.write((byte) (v >> 48), invert);
buffer.write((byte) (v >> 40), invert);
buffer.write((byte) (v >> 32), invert);
buffer.write((byte) (v >> 24), invert);
buffer.write((byte) (v >> 16), invert);
buffer.write((byte) (v >> 8), invert);
buffer.write((byte) v, invert);
return;
}
case FLOAT:
{
FloatObjectInspector foi = (FloatObjectInspector) poi;
int v = Float.floatToIntBits(foi.get(o));
if ((v & (1 << 31)) != 0) {
// negative number, flip all bits
v = ~v;
} else {
// positive number, flip the first bit
v = v ^ (1 << 31);
}
buffer.write((byte) (v >> 24), invert);
buffer.write((byte) (v >> 16), invert);
buffer.write((byte) (v >> 8), invert);
buffer.write((byte) v, invert);
return;
}
case DOUBLE:
{
DoubleObjectInspector doi = (DoubleObjectInspector) poi;
long v = Double.doubleToLongBits(doi.get(o));
if ((v & (1L << 63)) != 0) {
// negative number, flip all bits
v = ~v;
} else {
// positive number, flip the first bit
v = v ^ (1L << 63);
}
buffer.write((byte) (v >> 56), invert);
buffer.write((byte) (v >> 48), invert);
buffer.write((byte) (v >> 40), invert);
buffer.write((byte) (v >> 32), invert);
buffer.write((byte) (v >> 24), invert);
buffer.write((byte) (v >> 16), invert);
buffer.write((byte) (v >> 8), invert);
buffer.write((byte) v, invert);
return;
}
case STRING:
{
StringObjectInspector soi = (StringObjectInspector) poi;
Text t = soi.getPrimitiveWritableObject(o);
serializeBytes(buffer, t.getBytes(), t.getLength(), invert);
return;
}
case BINARY:
{
BinaryObjectInspector baoi = (BinaryObjectInspector) poi;
BytesWritable ba = baoi.getPrimitiveWritableObject(o);
byte[] toSer = new byte[ba.getLength()];
System.arraycopy(ba.getBytes(), 0, toSer, 0, ba.getLength());
serializeBytes(buffer, toSer, ba.getLength(), invert);
return;
}
case DATE:
{
DateObjectInspector doi = (DateObjectInspector) poi;
long v = doi.getPrimitiveWritableObject(o).getTimeInSeconds();
buffer.write((byte) ((v >> 56) ^ 0x80), invert);
buffer.write((byte) (v >> 48), invert);
buffer.write((byte) (v >> 40), invert);
buffer.write((byte) (v >> 32), invert);
buffer.write((byte) (v >> 24), invert);
buffer.write((byte) (v >> 16), invert);
buffer.write((byte) (v >> 8), invert);
buffer.write((byte) v, invert);
return;
}
case TIMESTAMP:
{
TimestampObjectInspector toi = (TimestampObjectInspector) poi;
TimestampWritable t = toi.getPrimitiveWritableObject(o);
byte[] data = t.getBinarySortable();
for (int i = 0; i < data.length; i++) {
buffer.write(data[i], invert);
}
return;
}
default:
{
throw new RuntimeException("Unrecognized type: " + poi.getPrimitiveCategory());
}
}
}
case LIST:
{
ListObjectInspector loi = (ListObjectInspector) oi;
ObjectInspector eoi = loi.getListElementObjectInspector();
// \1 followed by each element
int size = loi.getListLength(o);
for (int eid = 0; eid < size; eid++) {
buffer.write((byte) 1, invert);
serialize(buffer, loi.getListElement(o, eid), eoi, invert);
}
// and \0 to terminate
buffer.write((byte) 0, invert);
return;
}
case MAP:
{
MapObjectInspector moi = (MapObjectInspector) oi;
ObjectInspector koi = moi.getMapKeyObjectInspector();
ObjectInspector voi = moi.getMapValueObjectInspector();
// \1 followed by each key and then each value
Map<?, ?> map = moi.getMap(o);
for (Map.Entry<?, ?> entry : map.entrySet()) {
buffer.write((byte) 1, invert);
serialize(buffer, entry.getKey(), koi, invert);
serialize(buffer, entry.getValue(), voi, invert);
}
// and \0 to terminate
buffer.write((byte) 0, invert);
return;
}
case STRUCT:
{
StructObjectInspector soi = (StructObjectInspector) oi;
List<? extends StructField> fields = soi.getAllStructFieldRefs();
for (int i = 0; i < fields.size(); i++) {
serialize(
buffer,
soi.getStructFieldData(o, fields.get(i)),
fields.get(i).getFieldObjectInspector(),
invert);
}
return;
}
case UNION:
{
UnionObjectInspector uoi = (UnionObjectInspector) oi;
byte tag = uoi.getTag(o);
buffer.write(tag, invert);
serialize(buffer, uoi.getField(o), uoi.getObjectInspectors().get(tag), invert);
return;
}
default:
{
throw new RuntimeException("Unrecognized type: " + oi.getCategory());
}
}
}
public static Timestamp createTimestamp(byte[] bytes, int offset) {
Timestamp t = new Timestamp(0);
TimestampWritable.setTimestamp(t, bytes, offset);
return t;
}
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 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:
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());
}
}
static Object deserialize(InputByteBuffer buffer, TypeInfo type, boolean invert, Object reuse)
throws IOException {
// Is this field a null?
byte isNull = buffer.read(invert);
if (isNull == 0) {
return null;
}
assert (isNull == 1);
switch (type.getCategory()) {
case PRIMITIVE:
{
PrimitiveTypeInfo ptype = (PrimitiveTypeInfo) type;
switch (ptype.getPrimitiveCategory()) {
case VOID:
{
return null;
}
case BOOLEAN:
{
BooleanWritable r = reuse == null ? new BooleanWritable() : (BooleanWritable) reuse;
byte b = buffer.read(invert);
assert (b == 1 || b == 2);
r.set(b == 2);
return r;
}
case BYTE:
{
ByteWritable r = reuse == null ? new ByteWritable() : (ByteWritable) reuse;
r.set((byte) (buffer.read(invert) ^ 0x80));
return r;
}
case SHORT:
{
ShortWritable r = reuse == null ? new ShortWritable() : (ShortWritable) reuse;
int v = buffer.read(invert) ^ 0x80;
v = (v << 8) + (buffer.read(invert) & 0xff);
r.set((short) v);
return r;
}
case INT:
{
IntWritable r = reuse == null ? new IntWritable() : (IntWritable) reuse;
int v = buffer.read(invert) ^ 0x80;
for (int i = 0; i < 3; i++) {
v = (v << 8) + (buffer.read(invert) & 0xff);
}
r.set(v);
return r;
}
case LONG:
{
LongWritable r = reuse == null ? new LongWritable() : (LongWritable) reuse;
long v = buffer.read(invert) ^ 0x80;
for (int i = 0; i < 7; i++) {
v = (v << 8) + (buffer.read(invert) & 0xff);
}
r.set(v);
return r;
}
case FLOAT:
{
FloatWritable r = reuse == null ? new FloatWritable() : (FloatWritable) reuse;
int v = 0;
for (int i = 0; i < 4; i++) {
v = (v << 8) + (buffer.read(invert) & 0xff);
}
if ((v & (1 << 31)) == 0) {
// negative number, flip all bits
v = ~v;
} else {
// positive number, flip the first bit
v = v ^ (1 << 31);
}
r.set(Float.intBitsToFloat(v));
return r;
}
case DOUBLE:
{
DoubleWritable r = reuse == null ? new DoubleWritable() : (DoubleWritable) reuse;
long v = 0;
for (int i = 0; i < 8; i++) {
v = (v << 8) + (buffer.read(invert) & 0xff);
}
if ((v & (1L << 63)) == 0) {
// negative number, flip all bits
v = ~v;
} else {
// positive number, flip the first bit
v = v ^ (1L << 63);
}
r.set(Double.longBitsToDouble(v));
return r;
}
case STRING:
{
Text r = reuse == null ? new Text() : (Text) reuse;
// Get the actual length first
int start = buffer.tell();
int length = 0;
do {
byte b = buffer.read(invert);
if (b == 0) {
// end of string
break;
}
if (b == 1) {
// the last char is an escape char. read the actual char
buffer.read(invert);
}
length++;
} while (true);
if (length == buffer.tell() - start) {
// No escaping happened, so we are already done.
r.set(buffer.getData(), start, length);
} else {
// Escaping happened, we need to copy byte-by-byte.
// 1. Set the length first.
r.set(buffer.getData(), start, length);
// 2. Reset the pointer.
buffer.seek(start);
// 3. Copy the data.
byte[] rdata = r.getBytes();
for (int i = 0; i < length; i++) {
byte b = buffer.read(invert);
if (b == 1) {
// The last char is an escape char, read the actual char.
// The serialization format escape \0 to \1, and \1 to \2,
// to make sure the string is null-terminated.
b = (byte) (buffer.read(invert) - 1);
}
rdata[i] = b;
}
// 4. Read the null terminator.
byte b = buffer.read(invert);
assert (b == 0);
}
return r;
}
case BINARY:
{
BytesWritable bw = new BytesWritable();
// Get the actual length first
int start = buffer.tell();
int length = 0;
do {
byte b = buffer.read(invert);
if (b == 0) {
// end of string
break;
}
if (b == 1) {
// the last char is an escape char. read the actual char
buffer.read(invert);
}
length++;
} while (true);
if (length == buffer.tell() - start) {
// No escaping happened, so we are already done.
bw.set(buffer.getData(), start, length);
} else {
// Escaping happened, we need to copy byte-by-byte.
// 1. Set the length first.
bw.set(buffer.getData(), start, length);
// 2. Reset the pointer.
buffer.seek(start);
// 3. Copy the data.
byte[] rdata = bw.getBytes();
for (int i = 0; i < length; i++) {
byte b = buffer.read(invert);
if (b == 1) {
// The last char is an escape char, read the actual char.
// The serialization format escape \0 to \1, and \1 to \2,
// to make sure the string is null-terminated.
b = (byte) (buffer.read(invert) - 1);
}
rdata[i] = b;
}
// 4. Read the null terminator.
byte b = buffer.read(invert);
assert (b == 0);
}
return bw;
}
case DATE:
{
DateWritable d = reuse == null ? new DateWritable() : (DateWritable) reuse;
long v = buffer.read(invert) ^ 0x80;
for (int i = 0; i < 7; i++) {
v = (v << 8) + (buffer.read(invert) & 0xff);
}
d.set(DateWritable.timeToDate(v));
return d;
}
case TIMESTAMP:
TimestampWritable t =
(reuse == null ? new TimestampWritable() : (TimestampWritable) reuse);
byte[] bytes = new byte[8];
for (int i = 0; i < bytes.length; i++) {
bytes[i] = buffer.read(invert);
}
t.setBinarySortable(bytes, 0);
return t;
default:
{
throw new RuntimeException("Unrecognized type: " + ptype.getPrimitiveCategory());
}
}
}
case LIST:
{
ListTypeInfo ltype = (ListTypeInfo) type;
TypeInfo etype = ltype.getListElementTypeInfo();
// Create the list if needed
ArrayList<Object> r = reuse == null ? new ArrayList<Object>() : (ArrayList<Object>) reuse;
// Read the list
int size = 0;
while (true) {
int more = buffer.read(invert);
if (more == 0) {
// \0 to terminate
break;
}
// \1 followed by each element
assert (more == 1);
if (size == r.size()) {
r.add(null);
}
r.set(size, deserialize(buffer, etype, invert, r.get(size)));
size++;
}
// Remove additional elements if the list is reused
while (r.size() > size) {
r.remove(r.size() - 1);
}
return r;
}
case MAP:
{
MapTypeInfo mtype = (MapTypeInfo) type;
TypeInfo ktype = mtype.getMapKeyTypeInfo();
TypeInfo vtype = mtype.getMapValueTypeInfo();
// Create the map if needed
Map<Object, Object> r;
if (reuse == null) {
r = new HashMap<Object, Object>();
} else {
r = (HashMap<Object, Object>) reuse;
r.clear();
}
while (true) {
int more = buffer.read(invert);
if (more == 0) {
// \0 to terminate
break;
}
// \1 followed by each key and then each value
assert (more == 1);
Object k = deserialize(buffer, ktype, invert, null);
Object v = deserialize(buffer, vtype, invert, null);
r.put(k, v);
}
return r;
}
case STRUCT:
{
StructTypeInfo stype = (StructTypeInfo) type;
List<TypeInfo> fieldTypes = stype.getAllStructFieldTypeInfos();
int size = fieldTypes.size();
// Create the struct if needed
ArrayList<Object> r =
reuse == null ? new ArrayList<Object>(size) : (ArrayList<Object>) reuse;
assert (r.size() <= size);
// Set the size of the struct
while (r.size() < size) {
r.add(null);
}
// Read one field by one field
for (int eid = 0; eid < size; eid++) {
r.set(eid, deserialize(buffer, fieldTypes.get(eid), invert, r.get(eid)));
}
return r;
}
case UNION:
{
UnionTypeInfo utype = (UnionTypeInfo) type;
StandardUnion r = reuse == null ? new StandardUnion() : (StandardUnion) reuse;
// Read the tag
byte tag = buffer.read(invert);
r.setTag(tag);
r.setObject(
deserialize(buffer, utype.getAllUnionObjectTypeInfos().get(tag), invert, null));
return r;
}
default:
{
throw new RuntimeException("Unrecognized type: " + type.getCategory());
}
}
}
public TimestampWritable(TimestampWritable t) {
this(t.getBytes(), 0);
}
/** Compare two objects with their respective ObjectInspectors. */
public static int compare(
Object o1,
ObjectInspector oi1,
Object o2,
ObjectInspector oi2,
MapEqualComparer mapEqualComparer) {
if (oi1.getCategory() != oi2.getCategory()) {
return oi1.getCategory().compareTo(oi2.getCategory());
}
if (o1 == null) {
return o2 == null ? 0 : -1;
} else if (o2 == null) {
return 1;
}
switch (oi1.getCategory()) {
case PRIMITIVE:
{
PrimitiveObjectInspector poi1 = ((PrimitiveObjectInspector) oi1);
PrimitiveObjectInspector poi2 = ((PrimitiveObjectInspector) oi2);
if (poi1.getPrimitiveCategory() != poi2.getPrimitiveCategory()) {
return poi1.getPrimitiveCategory().compareTo(poi2.getPrimitiveCategory());
}
switch (poi1.getPrimitiveCategory()) {
case VOID:
return 0;
case BOOLEAN:
{
int v1 = ((BooleanObjectInspector) poi1).get(o1) ? 1 : 0;
int v2 = ((BooleanObjectInspector) poi2).get(o2) ? 1 : 0;
return v1 - v2;
}
case BYTE:
{
int v1 = ((ByteObjectInspector) poi1).get(o1);
int v2 = ((ByteObjectInspector) poi2).get(o2);
return v1 - v2;
}
case SHORT:
{
int v1 = ((ShortObjectInspector) poi1).get(o1);
int v2 = ((ShortObjectInspector) poi2).get(o2);
return v1 - v2;
}
case INT:
{
int v1 = ((IntObjectInspector) poi1).get(o1);
int v2 = ((IntObjectInspector) poi2).get(o2);
return v1 > v2 ? 1 : (v1 < v2 ? -1 : 0);
}
case LONG:
{
long v1 = ((LongObjectInspector) poi1).get(o1);
long v2 = ((LongObjectInspector) poi2).get(o2);
return v1 > v2 ? 1 : (v1 < v2 ? -1 : 0);
}
case FLOAT:
{
float v1 = ((FloatObjectInspector) poi1).get(o1);
float v2 = ((FloatObjectInspector) poi2).get(o2);
return Float.compare(v1, v2);
}
case DOUBLE:
{
double v1 = ((DoubleObjectInspector) poi1).get(o1);
double v2 = ((DoubleObjectInspector) poi2).get(o2);
return Double.compare(v1, v2);
}
case STRING:
{
if (poi1.preferWritable() || poi2.preferWritable()) {
Text t1 = (Text) poi1.getPrimitiveWritableObject(o1);
Text t2 = (Text) poi2.getPrimitiveWritableObject(o2);
return t1 == null ? (t2 == null ? 0 : -1) : (t2 == null ? 1 : t1.compareTo(t2));
} else {
String s1 = (String) poi1.getPrimitiveJavaObject(o1);
String s2 = (String) poi2.getPrimitiveJavaObject(o2);
return s1 == null ? (s2 == null ? 0 : -1) : (s2 == null ? 1 : s1.compareTo(s2));
}
}
case CHAR:
{
HiveCharWritable t1 =
((HiveCharObjectInspector) poi1).getPrimitiveWritableObject(o1);
HiveCharWritable t2 =
((HiveCharObjectInspector) poi2).getPrimitiveWritableObject(o2);
return t1.compareTo(t2);
}
case VARCHAR:
{
HiveVarcharWritable t1 =
((HiveVarcharObjectInspector) poi1).getPrimitiveWritableObject(o1);
HiveVarcharWritable t2 =
((HiveVarcharObjectInspector) poi2).getPrimitiveWritableObject(o2);
return t1.compareTo(t2);
}
case BINARY:
{
BytesWritable bw1 = ((BinaryObjectInspector) poi1).getPrimitiveWritableObject(o1);
BytesWritable bw2 = ((BinaryObjectInspector) poi2).getPrimitiveWritableObject(o2);
return bw1.compareTo(bw2);
}
case DATE:
{
DateWritable d1 = ((DateObjectInspector) poi1).getPrimitiveWritableObject(o1);
DateWritable d2 = ((DateObjectInspector) poi2).getPrimitiveWritableObject(o2);
return d1.compareTo(d2);
}
case TIMESTAMP:
{
TimestampWritable t1 =
((TimestampObjectInspector) poi1).getPrimitiveWritableObject(o1);
TimestampWritable t2 =
((TimestampObjectInspector) poi2).getPrimitiveWritableObject(o2);
return t1.compareTo(t2);
}
case INTERVAL_YEAR_MONTH:
{
HiveIntervalYearMonthWritable i1 =
((HiveIntervalYearMonthObjectInspector) poi1).getPrimitiveWritableObject(o1);
HiveIntervalYearMonthWritable i2 =
((HiveIntervalYearMonthObjectInspector) poi2).getPrimitiveWritableObject(o2);
return i1.compareTo(i2);
}
case INTERVAL_DAY_TIME:
{
HiveIntervalDayTimeWritable i1 =
((HiveIntervalDayTimeObjectInspector) poi1).getPrimitiveWritableObject(o1);
HiveIntervalDayTimeWritable i2 =
((HiveIntervalDayTimeObjectInspector) poi2).getPrimitiveWritableObject(o2);
return i1.compareTo(i2);
}
case DECIMAL:
{
HiveDecimalWritable t1 =
((HiveDecimalObjectInspector) poi1).getPrimitiveWritableObject(o1);
HiveDecimalWritable t2 =
((HiveDecimalObjectInspector) poi2).getPrimitiveWritableObject(o2);
return t1.compareTo(t2);
}
default:
{
throw new RuntimeException("Unknown type: " + poi1.getPrimitiveCategory());
}
}
}
case STRUCT:
{
StructObjectInspector soi1 = (StructObjectInspector) oi1;
StructObjectInspector soi2 = (StructObjectInspector) oi2;
List<? extends StructField> fields1 = soi1.getAllStructFieldRefs();
List<? extends StructField> fields2 = soi2.getAllStructFieldRefs();
int minimum = Math.min(fields1.size(), fields2.size());
for (int i = 0; i < minimum; i++) {
int r =
compare(
soi1.getStructFieldData(o1, fields1.get(i)),
fields1.get(i).getFieldObjectInspector(),
soi2.getStructFieldData(o2, fields2.get(i)),
fields2.get(i).getFieldObjectInspector(),
mapEqualComparer);
if (r != 0) {
return r;
}
}
return fields1.size() - fields2.size();
}
case LIST:
{
ListObjectInspector loi1 = (ListObjectInspector) oi1;
ListObjectInspector loi2 = (ListObjectInspector) oi2;
int minimum = Math.min(loi1.getListLength(o1), loi2.getListLength(o2));
for (int i = 0; i < minimum; i++) {
int r =
compare(
loi1.getListElement(o1, i),
loi1.getListElementObjectInspector(),
loi2.getListElement(o2, i),
loi2.getListElementObjectInspector(),
mapEqualComparer);
if (r != 0) {
return r;
}
}
return loi1.getListLength(o1) - loi2.getListLength(o2);
}
case MAP:
{
if (mapEqualComparer == null) {
throw new RuntimeException("Compare on map type not supported!");
} else {
return mapEqualComparer.compare(
o1, (MapObjectInspector) oi1, o2, (MapObjectInspector) oi2);
}
}
case UNION:
{
UnionObjectInspector uoi1 = (UnionObjectInspector) oi1;
UnionObjectInspector uoi2 = (UnionObjectInspector) oi2;
byte tag1 = uoi1.getTag(o1);
byte tag2 = uoi2.getTag(o2);
if (tag1 != tag2) {
return tag1 - tag2;
}
return compare(
uoi1.getField(o1),
uoi1.getObjectInspectors().get(tag1),
uoi2.getField(o2),
uoi2.getObjectInspectors().get(tag2),
mapEqualComparer);
}
default:
throw new RuntimeException("Compare on unknown type: " + oi1.getCategory());
}
}