Skye Wanderman-Milne has posted comments on this change.

Change subject: IMPALA-2736: Basic column-wise slot materialization in Parquet 
scanner.
......................................................................


Patch Set 6:

(3 comments)

http://gerrit.cloudera.org:8080/#/c/2779/6/be/src/exec/hdfs-parquet-scanner.cc
File be/src/exec/hdfs-parquet-scanner.cc:

Line 1164:   pool->AcquireData(decompressed_data_pool_.get(), false);
Personally, I think it's more clear if you change this to 
scratch_batch_->mem_pool() instead of plumbing pool around. When I initially 
read this, my immediate question was, "Which pool is this?" and then I had to 
trace this all the way back up to AssembleRows(). Going the other way, it's not 
obvious from the ReadValueBatch() signature what the is used for since it 
happens so far down the callstack (although you could say it in the comment I 
suppose). Your call.


Line 1754:     int num_null_tuples = min(batch_->capacity() - 
batch_->num_rows(),
nit: I would just call this "num_tuples", they're technically NULL but that's 
kind of a bug.


Line 1897:     if (column_readers[0]->RowGroupAtEnd()) break;
nit: I think this should go in the while loop because it describes the logic 
more clearly. If anything continue_execution seems like the exceptional case 
that could go in a break statement, e.g.:

while not at end of row group:
  read some rows
  if (UNLIKELY(!continue_execution)) break;


-- 
To view, visit http://gerrit.cloudera.org:8080/2779
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I72a613fa805c542e39df20588fb25c57b5f139aa
Gerrit-PatchSet: 6
Gerrit-Project: Impala
Gerrit-Branch: cdh5-trunk
Gerrit-Owner: Alex Behm <[email protected]>
Gerrit-Reviewer: Alex Behm <[email protected]>
Gerrit-Reviewer: Mostafa Mokhtar <[email protected]>
Gerrit-Reviewer: Skye Wanderman-Milne <[email protected]>
Gerrit-Reviewer: Tim Armstrong <[email protected]>
Gerrit-HasComments: Yes

Reply via email to