use of org.apache.carbondata.core.metadata.schema.table.column.CarbonDimension in project carbondata by apache.
the class CarbonParserFactory method createParser.
/**
* This method may be called recursively if the carbon column is complex type.
*
* @param carbonColumn
* @param complexDelimiters, these delimiters which are used to separate the complex data types.
* @param depth It is like depth of tree, if column has children then depth is 1,
* And depth becomes 2 if children has children.
* This depth is used select the complex
* delimiters
* @return GenericParser
*/
private static GenericParser createParser(CarbonColumn carbonColumn, String[] complexDelimiters, String nullFormat, int depth) {
switch(carbonColumn.getDataType()) {
case ARRAY:
List<CarbonDimension> listOfChildDimensions = ((CarbonDimension) carbonColumn).getListOfChildDimensions();
// Create array parser with complex delimiter
ArrayParserImpl arrayParser = new ArrayParserImpl(complexDelimiters[depth], nullFormat);
for (CarbonDimension dimension : listOfChildDimensions) {
arrayParser.addChildren(createParser(dimension, complexDelimiters, nullFormat, depth + 1));
}
return arrayParser;
case STRUCT:
List<CarbonDimension> dimensions = ((CarbonDimension) carbonColumn).getListOfChildDimensions();
// Create struct parser with complex delimiter
StructParserImpl parser = new StructParserImpl(complexDelimiters[depth], nullFormat);
for (CarbonDimension dimension : dimensions) {
parser.addChildren(createParser(dimension, complexDelimiters, nullFormat, depth + 1));
}
return parser;
case MAP:
throw new UnsupportedOperationException("Complex type Map is not supported yet");
default:
return new PrimitiveParserImpl();
}
}
use of org.apache.carbondata.core.metadata.schema.table.column.CarbonDimension in project carbondata by apache.
the class CarbonCompactionExecutor method prepareQueryModel.
/**
* Preparing of the query model.
*
* @param blockList
* @return
*/
private QueryModel prepareQueryModel(List<TableBlockInfo> blockList) {
QueryModel model = new QueryModel();
model.setTableBlockInfos(blockList);
model.setForcedDetailRawQuery(true);
model.setFilterExpressionResolverTree(null);
List<QueryDimension> dims = new ArrayList<>(CarbonCommonConstants.DEFAULT_COLLECTION_SIZE);
List<CarbonDimension> dimensions = carbonTable.getDimensionByTableName(carbonTable.getFactTableName());
for (CarbonDimension dim : dimensions) {
// check if dimension is deleted
QueryDimension queryDimension = new QueryDimension(dim.getColName());
queryDimension.setDimension(dim);
dims.add(queryDimension);
}
model.setQueryDimension(dims);
List<QueryMeasure> msrs = new ArrayList<>(CarbonCommonConstants.DEFAULT_COLLECTION_SIZE);
List<CarbonMeasure> measures = carbonTable.getMeasureByTableName(carbonTable.getFactTableName());
for (CarbonMeasure carbonMeasure : measures) {
// check if measure is deleted
QueryMeasure queryMeasure = new QueryMeasure(carbonMeasure.getColName());
queryMeasure.setMeasure(carbonMeasure);
msrs.add(queryMeasure);
}
model.setQueryMeasures(msrs);
model.setQueryId(System.nanoTime() + "");
model.setAbsoluteTableIdentifier(carbonTable.getAbsoluteTableIdentifier());
model.setTable(carbonTable);
return model;
}
use of org.apache.carbondata.core.metadata.schema.table.column.CarbonDimension in project carbondata by apache.
the class CompactionResultSortProcessor method initSortDataRows.
/**
* create an instance of sort data rows
*/
private void initSortDataRows() throws Exception {
measureCount = carbonTable.getMeasureByTableName(tableName).size();
List<CarbonDimension> dimensions = carbonTable.getDimensionByTableName(tableName);
noDictionaryColMapping = new boolean[dimensions.size()];
int i = 0;
for (CarbonDimension dimension : dimensions) {
if (CarbonUtil.hasEncoding(dimension.getEncoder(), Encoding.DICTIONARY)) {
i++;
continue;
}
noDictionaryColMapping[i++] = true;
noDictionaryCount++;
}
dimensionColumnCount = dimensions.size();
SortParameters parameters = createSortParameters();
intermediateFileMerger = new SortIntermediateFileMerger(parameters);
// TODO: Now it is only supported onheap merge, but we can have unsafe merge
// as well by using UnsafeSortDataRows.
this.sortDataRows = new SortDataRows(parameters, intermediateFileMerger);
try {
this.sortDataRows.initialize();
} catch (CarbonSortKeyAndGroupByException e) {
LOGGER.error(e);
throw new Exception("Error initializing sort data rows object during compaction: " + e.getMessage());
}
}
use of org.apache.carbondata.core.metadata.schema.table.column.CarbonDimension in project carbondata by apache.
the class StoreCreator method writeDictionary.
private static void writeDictionary(String factFilePath, CarbonTable table) throws Exception {
BufferedReader reader = new BufferedReader(new FileReader(factFilePath));
String header = reader.readLine();
String[] split = header.split(",");
List<CarbonColumn> allCols = new ArrayList<CarbonColumn>();
List<CarbonDimension> dims = table.getDimensionByTableName(table.getFactTableName());
allCols.addAll(dims);
List<CarbonMeasure> msrs = table.getMeasureByTableName(table.getFactTableName());
allCols.addAll(msrs);
Set<String>[] set = new HashSet[dims.size()];
for (int i = 0; i < set.length; i++) {
set[i] = new HashSet<String>();
}
String line = reader.readLine();
while (line != null) {
String[] data = line.split(",");
for (int i = 0; i < set.length; i++) {
set[i].add(data[i]);
}
line = reader.readLine();
}
Cache dictCache = CacheProvider.getInstance().createCache(CacheType.REVERSE_DICTIONARY, absoluteTableIdentifier.getStorePath());
for (int i = 0; i < set.length; i++) {
ColumnIdentifier columnIdentifier = new ColumnIdentifier(dims.get(i).getColumnId(), null, null);
CarbonDictionaryWriter writer = new CarbonDictionaryWriterImpl(absoluteTableIdentifier.getStorePath(), absoluteTableIdentifier.getCarbonTableIdentifier(), columnIdentifier);
for (String value : set[i]) {
writer.write(value);
}
writer.close();
writer.commit();
Dictionary dict = (Dictionary) dictCache.get(new DictionaryColumnUniqueIdentifier(absoluteTableIdentifier.getCarbonTableIdentifier(), columnIdentifier, dims.get(i).getDataType()));
CarbonDictionarySortInfoPreparator preparator = new CarbonDictionarySortInfoPreparator();
List<String> newDistinctValues = new ArrayList<String>();
CarbonDictionarySortInfo dictionarySortInfo = preparator.getDictionarySortInfo(newDistinctValues, dict, dims.get(i).getDataType());
CarbonDictionarySortIndexWriter carbonDictionaryWriter = new CarbonDictionarySortIndexWriterImpl(absoluteTableIdentifier.getCarbonTableIdentifier(), columnIdentifier, absoluteTableIdentifier.getStorePath());
try {
carbonDictionaryWriter.writeSortIndex(dictionarySortInfo.getSortIndex());
carbonDictionaryWriter.writeInvertedSortIndex(dictionarySortInfo.getSortIndexInverted());
} finally {
carbonDictionaryWriter.close();
}
}
reader.close();
}
use of org.apache.carbondata.core.metadata.schema.table.column.CarbonDimension in project carbondata by apache.
the class CarbonInputFormatUtil method processFilterExpression.
public static void processFilterExpression(Expression filterExpression, CarbonTable carbonTable) {
List<CarbonDimension> dimensions = carbonTable.getDimensionByTableName(carbonTable.getFactTableName());
List<CarbonMeasure> measures = carbonTable.getMeasureByTableName(carbonTable.getFactTableName());
QueryModel.processFilterExpression(filterExpression, dimensions, measures);
if (null != filterExpression) {
// Optimize Filter Expression and fit RANGE filters is conditions apply.
FilterOptimizer rangeFilterOptimizer = new RangeFilterOptmizer(new FilterOptimizerBasic(), filterExpression);
rangeFilterOptimizer.optimizeFilter();
}
}
Aggregations