use of org.apache.carbondata.core.metadata.blocklet.DataFileFooter in project carbondata by apache.
the class DataFileFooterConverterTest method testGetIndexInfo.
@Test
public void testGetIndexInfo() throws Exception {
DataFileFooterConverter dataFileFooterConverter = new DataFileFooterConverter();
final ThriftReader thriftReader = new ThriftReader("file");
List<Encoding> encoders = new ArrayList<>();
encoders.add(Encoding.INVERTED_INDEX);
encoders.add(Encoding.BIT_PACKED);
encoders.add(Encoding.DELTA);
encoders.add(Encoding.DICTIONARY);
encoders.add(Encoding.DIRECT_DICTIONARY);
encoders.add(Encoding.RLE);
ColumnSchema columnSchema = new ColumnSchema(DataType.INT, "column", "3", true, encoders, true);
ColumnSchema columnSchema1 = new ColumnSchema(DataType.ARRAY, "column", "3", true, encoders, true);
ColumnSchema columnSchema2 = new ColumnSchema(DataType.DECIMAL, "column", "3", true, encoders, true);
ColumnSchema columnSchema3 = new ColumnSchema(DataType.DOUBLE, "column", "3", true, encoders, true);
ColumnSchema columnSchema4 = new ColumnSchema(DataType.LONG, "column", "3", true, encoders, true);
ColumnSchema columnSchema5 = new ColumnSchema(DataType.SHORT, "column", "3", true, encoders, true);
ColumnSchema columnSchema6 = new ColumnSchema(DataType.STRUCT, "column", "3", true, encoders, true);
ColumnSchema columnSchema7 = new ColumnSchema(DataType.STRING, "column", "3", true, encoders, true);
final List<ColumnSchema> columnSchemas = new ArrayList<>();
columnSchemas.add(columnSchema);
columnSchemas.add(columnSchema1);
columnSchemas.add(columnSchema2);
columnSchemas.add(columnSchema3);
columnSchemas.add(columnSchema4);
columnSchemas.add(columnSchema5);
columnSchemas.add(columnSchema6);
columnSchemas.add(columnSchema7);
final BlockIndex blockIndex = new BlockIndex();
blockIndex.setBlock_index(new org.apache.carbondata.format.BlockletIndex());
org.apache.carbondata.format.BlockletIndex blockletIndex1 = new org.apache.carbondata.format.BlockletIndex();
BlockletBTreeIndex blockletBTreeIndex = new BlockletBTreeIndex();
blockletBTreeIndex.setStart_key("1".getBytes());
blockletBTreeIndex.setEnd_key("3".getBytes());
blockletIndex1.setB_tree_index(blockletBTreeIndex);
BlockletMinMaxIndex blockletMinMaxIndex = new BlockletMinMaxIndex();
blockletMinMaxIndex.setMax_values(Arrays.asList(ByteBuffer.allocate(1).put((byte) 2)));
blockletMinMaxIndex.setMin_values(Arrays.asList(ByteBuffer.allocate(1).put((byte) 1)));
blockletIndex1.setMin_max_index(blockletMinMaxIndex);
blockIndex.setBlock_index(blockletIndex1);
List<Integer> column_cardinalities = new ArrayList<>();
column_cardinalities.add(new Integer("1"));
final org.apache.carbondata.format.SegmentInfo segmentInfo1 = new org.apache.carbondata.format.SegmentInfo(3, column_cardinalities);
new MockUp<CarbonIndexFileReader>() {
boolean mockedHasNextStatus = true;
@SuppressWarnings("unused")
@Mock
public boolean hasNext() throws IOException {
boolean temp = mockedHasNextStatus;
mockedHasNextStatus = false;
return temp;
}
@SuppressWarnings("unused")
@Mock
public void openThriftReader(String filePath) throws IOException {
thriftReader.open();
}
@SuppressWarnings("unused")
@Mock
public IndexHeader readIndexHeader() throws IOException {
return new IndexHeader(1, columnSchemas, segmentInfo1);
}
@SuppressWarnings("unused")
@Mock
public BlockIndex readBlockIndexInfo() throws IOException {
return blockIndex;
}
@SuppressWarnings("unused")
@Mock
public void closeThriftReader() {
thriftReader.close();
}
};
new MockUp<IndexHeader>() {
@SuppressWarnings("unused")
@Mock
public List<ColumnSchema> getTable_columns() {
return columnSchemas;
}
};
ByteArrayInputStream byteArrayInputStream = new ByteArrayInputStream("1".getBytes());
final DataInputStream dataInputStream = new DataInputStream(byteArrayInputStream);
new MockUp<FileFactory>() {
@SuppressWarnings("unused")
@Mock
public DataInputStream getDataInputStream(String path, FileFactory.FileType fileType, int bufferSize) {
return dataInputStream;
}
};
String[] arr = { "a", "b", "c" };
String fileName = "/part-0-0_batchno0-0-1495074251740.carbondata";
TableBlockInfo tableBlockInfo = new TableBlockInfo(fileName, 3, "id", arr, 3, ColumnarFormatVersion.V1);
tableBlockInfo.getBlockletInfos().setNoOfBlockLets(3);
List<TableBlockInfo> tableBlockInfoList = new ArrayList<>();
tableBlockInfoList.add(tableBlockInfo);
String idxFileName = "0_batchno0-0-1495074251740.carbonindex";
List<DataFileFooter> dataFileFooterList = dataFileFooterConverter.getIndexInfo(idxFileName, tableBlockInfoList);
byte[] exp = dataFileFooterList.get(0).getBlockletIndex().getBtreeIndex().getStartKey();
byte[] res = "1".getBytes();
for (int i = 0; i < exp.length; i++) {
assertEquals(exp[i], res[i]);
}
}
use of org.apache.carbondata.core.metadata.blocklet.DataFileFooter in project carbondata by apache.
the class BTreeBlockFinderTest method getFileFooter.
private DataFileFooter getFileFooter(byte[] startKey, byte[] endKey, byte[] noDictionaryStartKey, byte[] noDictionaryEndKey) {
DataFileFooter footer = new DataFileFooter();
BlockletIndex index = new BlockletIndex();
BlockletBTreeIndex btreeIndex = new BlockletBTreeIndex();
ByteBuffer buffer = ByteBuffer.allocate(4 + startKey.length + 4 + noDictionaryStartKey.length);
buffer.putInt(startKey.length);
buffer.putInt(noDictionaryStartKey.length);
buffer.put(startKey);
buffer.put(noDictionaryStartKey);
buffer.rewind();
btreeIndex.setStartKey(buffer.array());
ByteBuffer buffer1 = ByteBuffer.allocate(4 + startKey.length + 4 + noDictionaryEndKey.length);
buffer1.putInt(endKey.length);
buffer1.putInt(noDictionaryEndKey.length);
buffer1.put(endKey);
buffer1.put(noDictionaryEndKey);
buffer1.rewind();
btreeIndex.setEndKey(buffer1.array());
BlockletMinMaxIndex minMax = new BlockletMinMaxIndex();
minMax.setMaxValues(new byte[][] { endKey, noDictionaryEndKey });
minMax.setMinValues(new byte[][] { startKey, noDictionaryStartKey });
index.setBtreeIndex(btreeIndex);
index.setMinMaxIndex(minMax);
footer.setBlockletIndex(index);
return footer;
}
use of org.apache.carbondata.core.metadata.blocklet.DataFileFooter in project carbondata by apache.
the class BTreeBlockFinderTest method getFileFooterListWithOnlyNoDictionaryKey.
private List<DataFileFooter> getFileFooterListWithOnlyNoDictionaryKey() {
List<DataFileFooter> list = new ArrayList<DataFileFooter>();
try {
int[] dimensionBitLength = CarbonUtil.getDimensionBitLength(new int[] { 10000, 10000 }, new int[] { 1, 1 });
KeyGenerator multiDimKeyVarLengthGenerator = new MultiDimKeyVarLengthGenerator(dimensionBitLength);
int i = 1;
while (i < 1001) {
byte[] startKey = multiDimKeyVarLengthGenerator.generateKey(new int[] { i, i });
byte[] endKey = multiDimKeyVarLengthGenerator.generateKey(new int[] { i + 10, i + 10 });
ByteBuffer buffer = ByteBuffer.allocate(2 + 4);
buffer.rewind();
buffer.putShort((short) 1);
buffer.putInt(i);
buffer.array();
byte[] noDictionaryStartKey = buffer.array();
ByteBuffer buffer1 = ByteBuffer.allocate(2 + 4);
buffer1.rewind();
buffer1.putShort((short) 2);
buffer1.putInt(i + 10);
buffer1.array();
byte[] noDictionaryEndKey = buffer.array();
DataFileFooter footer = getFileMatadataWithOnlyNoDictionaryKey(startKey, endKey, noDictionaryStartKey, noDictionaryEndKey);
list.add(footer);
i = i + 10;
}
} catch (Exception e) {
LOGGER.error(e);
}
return list;
}
use of org.apache.carbondata.core.metadata.blocklet.DataFileFooter in project carbondata by apache.
the class BTreeBlockFinderTest method testBtreeBuldingIsPorper.
@Test
public void testBtreeBuldingIsPorper() {
BtreeBuilder builder = new BlockBTreeBuilder();
List<DataFileFooter> footerList = getDataFileFooterList();
BTreeBuilderInfo infos = new BTreeBuilderInfo(footerList, null);
builder.build(infos);
}
use of org.apache.carbondata.core.metadata.blocklet.DataFileFooter in project carbondata by apache.
the class CarbonCompactionUtil method checkIfAnyRestructuredBlockExists.
/**
* This method will check for any restructured block in the blocks selected for compaction
*
* @param segmentMapping
* @param dataFileMetadataSegMapping
* @param tableLastUpdatedTime
* @return
*/
public static boolean checkIfAnyRestructuredBlockExists(Map<String, TaskBlockInfo> segmentMapping, Map<String, List<DataFileFooter>> dataFileMetadataSegMapping, long tableLastUpdatedTime) {
boolean restructuredBlockExists = false;
for (Map.Entry<String, TaskBlockInfo> taskMap : segmentMapping.entrySet()) {
String segmentId = taskMap.getKey();
List<DataFileFooter> listMetadata = dataFileMetadataSegMapping.get(segmentId);
for (DataFileFooter dataFileFooter : listMetadata) {
// it indicates it is a restructured block
if (tableLastUpdatedTime > dataFileFooter.getSchemaUpdatedTimeStamp()) {
restructuredBlockExists = true;
break;
}
}
if (restructuredBlockExists) {
break;
}
}
return restructuredBlockExists;
}
Aggregations