public static class LimitStatsRule extends DefaultStatsRule implements NodeProcessor {
@Override
public Object process(Node nd, Stack<Node> stack, NodeProcessorCtx procCtx,
Object... nodeOutputs) throws SemanticException {
LimitOperator lop = (LimitOperator) nd;
Operator<? extends OperatorDesc> parent = lop.getParentOperators().get(0);
Statistics parentStats = parent.getStatistics();
AnnotateStatsProcCtx aspCtx = (AnnotateStatsProcCtx) procCtx;
HiveConf conf = aspCtx.getConf();
try {
long limit = -1;
limit = lop.getConf().getLimit();
if (satisfyPrecondition(parentStats)) {
Statistics stats = parentStats.clone();
// if limit is greater than available rows then do not update
// statistics
if (limit <= parentStats.getNumRows()) {
updateStats(stats, limit, true);
}
lop.setStatistics(stats);
if (LOG.isDebugEnabled()) {
LOG.debug("[0] STATS-" + lop.toString() + ": " + stats.extendedToString());
}
} else {
if (parentStats != null) {
// in the absence of column statistics, compute data size based on
// based on average row size
Statistics wcStats = parentStats.clone();
if (limit <= parentStats.getNumRows()) {
long numRows = limit;
long avgRowSize = parentStats.getAvgRowSize();
long dataSize = avgRowSize * limit;
wcStats.setNumRows(numRows);
wcStats.setDataSize(dataSize);
}
lop.setStatistics(wcStats);
if (LOG.isDebugEnabled()) {
LOG.debug("[1] STATS-" + lop.toString() + ": " + wcStats.extendedToString());
}
}
}
} catch (CloneNotSupportedException e) {
throw new SemanticException(ErrorMsg.STATISTICS_CLONING_FAILED.getMsg());