use of io.trino.orc.StreamId in project trino by trinodb.
the class Checkpoints method getStreamCheckpoints.
public static Map<StreamId, StreamCheckpoint> getStreamCheckpoints(Set<OrcColumnId> columns, ColumnMetadata<OrcType> columnTypes, boolean compressed, int rowGroupId, ColumnMetadata<ColumnEncoding> columnEncodings, Map<StreamId, Stream> streams, Map<StreamId, List<RowGroupIndex>> columnIndexes) throws InvalidCheckpointException {
ImmutableSetMultimap.Builder<OrcColumnId, StreamKind> streamKindsBuilder = ImmutableSetMultimap.builder();
for (Stream stream : streams.values()) {
streamKindsBuilder.put(stream.getColumnId(), stream.getStreamKind());
}
SetMultimap<OrcColumnId, StreamKind> streamKinds = streamKindsBuilder.build();
ImmutableMap.Builder<StreamId, StreamCheckpoint> checkpoints = ImmutableMap.builder();
for (Map.Entry<StreamId, List<RowGroupIndex>> entry : columnIndexes.entrySet()) {
OrcColumnId columnId = entry.getKey().getColumnId();
if (!columns.contains(columnId)) {
continue;
}
List<Integer> positionsList = entry.getValue().get(rowGroupId).getPositions();
ColumnEncodingKind columnEncoding = columnEncodings.get(columnId).getColumnEncodingKind();
OrcTypeKind columnType = columnTypes.get(columnId).getOrcTypeKind();
Set<StreamKind> availableStreams = streamKinds.get(columnId);
ColumnPositionsList columnPositionsList = new ColumnPositionsList(columnId, columnType, positionsList);
switch(columnType) {
case BOOLEAN:
checkpoints.putAll(getBooleanColumnCheckpoints(columnId, compressed, availableStreams, columnPositionsList));
break;
case BYTE:
checkpoints.putAll(getByteColumnCheckpoints(columnId, compressed, availableStreams, columnPositionsList));
break;
case SHORT:
case INT:
case LONG:
case DATE:
checkpoints.putAll(getLongColumnCheckpoints(columnId, columnEncoding, compressed, availableStreams, columnPositionsList));
break;
case FLOAT:
checkpoints.putAll(getFloatColumnCheckpoints(columnId, compressed, availableStreams, columnPositionsList));
break;
case DOUBLE:
checkpoints.putAll(getDoubleColumnCheckpoints(columnId, compressed, availableStreams, columnPositionsList));
break;
case TIMESTAMP:
case TIMESTAMP_INSTANT:
checkpoints.putAll(getTimestampColumnCheckpoints(columnId, columnEncoding, compressed, availableStreams, columnPositionsList));
break;
case BINARY:
case STRING:
case VARCHAR:
case CHAR:
checkpoints.putAll(getSliceColumnCheckpoints(columnId, columnEncoding, compressed, availableStreams, columnPositionsList));
break;
case LIST:
case MAP:
checkpoints.putAll(getListOrMapColumnCheckpoints(columnId, columnEncoding, compressed, availableStreams, columnPositionsList));
break;
case STRUCT:
checkpoints.putAll(getStructColumnCheckpoints(columnId, compressed, availableStreams, columnPositionsList));
break;
case DECIMAL:
checkpoints.putAll(getDecimalColumnCheckpoints(columnId, columnEncoding, compressed, availableStreams, columnPositionsList));
break;
default:
throw new IllegalArgumentException("Unsupported column type " + columnType);
}
}
return checkpoints.buildOrThrow();
}
use of io.trino.orc.StreamId in project trino by trinodb.
the class InputStreamSources method getInputStreamSource.
public <S extends ValueInputStream<?>> InputStreamSource<S> getInputStreamSource(OrcColumn column, StreamKind streamKind, Class<S> streamType) {
requireNonNull(column, "column is null");
requireNonNull(streamType, "streamType is null");
InputStreamSource<?> streamSource = streamSources.get(new StreamId(column.getColumnId(), streamKind));
if (streamSource == null) {
streamSource = missingStreamSource(streamType);
}
checkArgument(streamType.isAssignableFrom(streamSource.getStreamType()), "%s must be of type %s, not %s", column, streamType.getName(), streamSource.getStreamType().getName());
return (InputStreamSource<S>) streamSource;
}
Aggregations