Hi all,


I read the code of ¡®CFileIterator::Scan¡¯, and found that it would be
better to pass ¡®remaining_sel¡¯ to function ¡®CopyNextValues¡¯ to skip
coping the unnecessary data for the columns that are not in predicates. In
other words, the decoder will copy all the data of the columns that are not
in predicates.



for (const auto& col_pred : (ctx->DecoderEvalNotDisabled()) {

RETURN_NOT_OK(pb->dblk_->CopyNextAndEval(&this_batch, ctx, &remaining_sel,

} else {

RETURN_NOT_OK(pb->dblk_->CopyNextValues(&this_batch, &remaining_dst));
<-- Here




For example: select column_a, column_b from table where column_c=¡¯c¡¯;

In function ¡®MaterializingIterator::MaterializeBlock¡¯:

    Because the column_c is a predicate, so it is in
¡®col_idx_predicates_¡¯. And then the decoder will evaluate and copy the
right data or sets the ¡®SelectionVector¡¯ to false.

    Next, column_a and column_b are not predicates, so they are in
¡®non_predicate_column_indexes_¡¯. This time, the decoder will copy the data
directly even though some rows have been set to false.


I just want to make sure I am interpreting this correctly. Thanks in



2017-04-10 16:06:24
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB