use of org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe in project hive by apache.
the class TestLazySimpleSerDe method testSerDeParameters.
/**
* Tests the deprecated usage of SerDeParameters.
*/
@Test
@SuppressWarnings("deprecation")
public void testSerDeParameters() throws SerDeException, IOException {
// Setup
LazySimpleSerDe serDe = new LazySimpleSerDe();
Configuration conf = new Configuration();
MyTestClass row = new MyTestClass();
ExtraTypeInfo extraTypeInfo = new ExtraTypeInfo();
row.randomFill(new Random(1234), extraTypeInfo);
StructObjectInspector rowOI = (StructObjectInspector) ObjectInspectorFactory.getReflectionObjectInspector(MyTestClass.class, ObjectInspectorOptions.JAVA);
String fieldNames = ObjectInspectorUtils.getFieldNames(rowOI);
String fieldTypes = ObjectInspectorUtils.getFieldTypes(rowOI);
Properties schema = new Properties();
schema.setProperty(serdeConstants.LIST_COLUMNS, fieldNames);
schema.setProperty(serdeConstants.LIST_COLUMN_TYPES, fieldTypes);
SerDeUtils.initializeSerDe(serDe, conf, schema, null);
SerDeParameters serdeParams = LazySimpleSerDe.initSerdeParams(conf, schema, "testSerdeName");
// Test
LazyStruct data = (LazyStruct) serializeAndDeserialize(row, rowOI, serDe, serdeParams);
assertEquals((boolean) row.myBool, ((LazyBoolean) data.getField(0)).getWritableObject().get());
assertEquals((int) row.myInt, ((LazyInteger) data.getField(3)).getWritableObject().get());
}
use of org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe in project hive by apache.
the class TestLazySimpleSerDe method testLazySimpleSerDeExtraColumns.
/**
* Test the LazySimpleSerDe class with extra columns.
*/
public void testLazySimpleSerDeExtraColumns() throws Throwable {
try {
// Create the SerDe
LazySimpleSerDe serDe = new LazySimpleSerDe();
Configuration conf = new Configuration();
Properties tbl = createProperties();
SerDeUtils.initializeSerDe(serDe, conf, tbl, null);
// Data
Text t = new Text("123\t456\t789\t1000\t5.3\thive and hadoop\t1.\ta\tb\t");
String s = "123\t456\t789\t1000\t5.3\thive and hadoop\t1\ta";
Object[] expectedFieldsData = { new ByteWritable((byte) 123), new ShortWritable((short) 456), new IntWritable(789), new LongWritable(1000), new DoubleWritable(5.3), new Text("hive and hadoop"), new IntWritable(1), new Text("a") };
// Test
deserializeAndSerialize(serDe, t, s, expectedFieldsData);
} catch (Throwable e) {
e.printStackTrace();
throw e;
}
}
use of org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe in project hive by apache.
the class TestCrossMapEqualComparer method serializeAndDeserialize.
Object serializeAndDeserialize(TextStringMapHolder o1, StructObjectInspector oi1, LazySimpleSerDe serde, LazySerDeParameters serdeParams) throws IOException, SerDeException {
ByteStream.Output serializeStream = new ByteStream.Output();
LazySimpleSerDe.serialize(serializeStream, o1, oi1, serdeParams.getSeparators(), 0, serdeParams.getNullSequence(), serdeParams.isEscaped(), serdeParams.getEscapeChar(), serdeParams.getNeedsEscape());
Text t = new Text(serializeStream.toByteArray());
return serde.deserialize(t);
}
use of org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe in project hive by apache.
the class TestSimpleMapEqualComparer method serializeAndDeserialize.
Object serializeAndDeserialize(TextStringMapHolder o1, StructObjectInspector oi1, LazySimpleSerDe serde, LazySerDeParameters serdeParams) throws IOException, SerDeException {
ByteStream.Output serializeStream = new ByteStream.Output();
LazySimpleSerDe.serialize(serializeStream, o1, oi1, serdeParams.getSeparators(), 0, serdeParams.getNullSequence(), serdeParams.isEscaped(), serdeParams.getEscapeChar(), serdeParams.getNeedsEscape());
Text t = new Text(serializeStream.toByteArray());
return serde.deserialize(t);
}
use of org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe in project hive by apache.
the class HCatBaseStorer method getJavaObj.
/**
* Convert from Pig value object to Hive value object
* This method assumes that {@link #validateSchema(org.apache.pig.impl.logicalLayer.schema.Schema.FieldSchema, org.apache.hive.hcatalog.data.schema.HCatFieldSchema, org.apache.pig.impl.logicalLayer.schema.Schema, org.apache.hive.hcatalog.data.schema.HCatSchema, int)}
* which checks the types in Pig schema are compatible with target Hive table, has been called.
*/
private Object getJavaObj(Object pigObj, HCatFieldSchema hcatFS) throws HCatException, BackendException {
try {
if (pigObj == null)
return null;
// The real work-horse. Spend time and energy in this method if there is
// need to keep HCatStorer lean and go fast.
Type type = hcatFS.getType();
switch(type) {
case BINARY:
return ((DataByteArray) pigObj).get();
case STRUCT:
HCatSchema structSubSchema = hcatFS.getStructSubSchema();
// Unwrap the tuple.
List<Object> all = ((Tuple) pigObj).getAll();
ArrayList<Object> converted = new ArrayList<Object>(all.size());
for (int i = 0; i < all.size(); i++) {
converted.add(getJavaObj(all.get(i), structSubSchema.get(i)));
}
return converted;
case ARRAY:
// Unwrap the bag.
DataBag pigBag = (DataBag) pigObj;
HCatFieldSchema tupFS = hcatFS.getArrayElementSchema().get(0);
boolean needTuple = tupFS.getType() == Type.STRUCT;
List<Object> bagContents = new ArrayList<Object>((int) pigBag.size());
Iterator<Tuple> bagItr = pigBag.iterator();
while (bagItr.hasNext()) {
// If there is only one element in tuple contained in bag, we throw away the tuple.
bagContents.add(getJavaObj(needTuple ? bagItr.next() : bagItr.next().get(0), tupFS));
}
return bagContents;
case MAP:
Map<?, ?> pigMap = (Map<?, ?>) pigObj;
Map<Object, Object> typeMap = new HashMap<Object, Object>();
for (Entry<?, ?> entry : pigMap.entrySet()) {
// the value has a schema and not a FieldSchema
typeMap.put(// Schema validation enforces that the Key is a String
(String) entry.getKey(), getJavaObj(entry.getValue(), hcatFS.getMapValueSchema().get(0)));
}
return typeMap;
case STRING:
case INT:
case BIGINT:
case FLOAT:
case DOUBLE:
return pigObj;
case SMALLINT:
if ((Integer) pigObj < Short.MIN_VALUE || (Integer) pigObj > Short.MAX_VALUE) {
handleOutOfRangeValue(pigObj, hcatFS);
return null;
}
return ((Integer) pigObj).shortValue();
case TINYINT:
if ((Integer) pigObj < Byte.MIN_VALUE || (Integer) pigObj > Byte.MAX_VALUE) {
handleOutOfRangeValue(pigObj, hcatFS);
return null;
}
return ((Integer) pigObj).byteValue();
case BOOLEAN:
if (pigObj instanceof String) {
if (((String) pigObj).trim().compareTo("0") == 0) {
return Boolean.FALSE;
}
if (((String) pigObj).trim().compareTo("1") == 0) {
return Boolean.TRUE;
}
throw new BackendException("Unexpected type " + type + " for value " + pigObj + " of class " + pigObj.getClass().getName(), PigHCatUtil.PIG_EXCEPTION_CODE);
}
return Boolean.parseBoolean(pigObj.toString());
case DECIMAL:
BigDecimal bd = (BigDecimal) pigObj;
DecimalTypeInfo dti = (DecimalTypeInfo) hcatFS.getTypeInfo();
if (bd.precision() > dti.precision() || bd.scale() > dti.scale()) {
handleOutOfRangeValue(pigObj, hcatFS);
return null;
}
return HiveDecimal.create(bd);
case CHAR:
String charVal = (String) pigObj;
CharTypeInfo cti = (CharTypeInfo) hcatFS.getTypeInfo();
if (charVal.length() > cti.getLength()) {
handleOutOfRangeValue(pigObj, hcatFS);
return null;
}
return new HiveChar(charVal, cti.getLength());
case VARCHAR:
String varcharVal = (String) pigObj;
VarcharTypeInfo vti = (VarcharTypeInfo) hcatFS.getTypeInfo();
if (varcharVal.length() > vti.getLength()) {
handleOutOfRangeValue(pigObj, hcatFS);
return null;
}
return new HiveVarchar(varcharVal, vti.getLength());
case TIMESTAMP:
DateTime dt = (DateTime) pigObj;
// getMillis() returns UTC time regardless of TZ
return new Timestamp(dt.getMillis());
case DATE:
/**
* We ignore any TZ setting on Pig value since java.sql.Date doesn't have it (in any
* meaningful way). So the assumption is that if Pig value has 0 time component (midnight)
* we assume it reasonably 'fits' into a Hive DATE. If time part is not 0, it's considered
* out of range for target type.
*/
DateTime dateTime = ((DateTime) pigObj);
if (dateTime.getMillisOfDay() != 0) {
handleOutOfRangeValue(pigObj, hcatFS, "Time component must be 0 (midnight) in local timezone; Local TZ val='" + pigObj + "'");
return null;
}
/*java.sql.Date is a poorly defined API. Some (all?) SerDes call toString() on it
[e.g. LazySimpleSerDe, uses LazyUtils.writePrimitiveUTF8()], which automatically adjusts
for local timezone. Date.valueOf() also uses local timezone (as does Date(int,int,int).
Also see PigHCatUtil#extractPigObject() for corresponding read op. This way a DATETIME from Pig,
when stored into Hive and read back comes back with the same value.*/
return new Date(dateTime.getYear() - 1900, dateTime.getMonthOfYear() - 1, dateTime.getDayOfMonth());
default:
throw new BackendException("Unexpected HCat type " + type + " for value " + pigObj + " of class " + pigObj.getClass().getName(), PigHCatUtil.PIG_EXCEPTION_CODE);
}
} catch (BackendException e) {
// provide the path to the field in the error message
throw new BackendException((hcatFS.getName() == null ? " " : hcatFS.getName() + ".") + e.getMessage(), e);
}
}
Aggregations