Aman Sinha has posted comments on this change. ( http://gerrit.cloudera.org:8080/16346 )
Change subject: IMPALA-10064: Support constant propagation for eligible range predicates ...................................................................... Patch Set 11: (4 comments) > Patch Set 10: > > > Patch Set 10: > > > > Change looks good to me. I should wait for the e2e test you mentioned right? > > Thanks for the review. Yes, I plan to add at least 1 e2e test with the > modified dataset. I hope to do it later today after some other ongoing work. Made these changes. http://gerrit.cloudera.org:8080/#/c/16346/10/fe/src/main/java/org/apache/impala/analysis/BinaryPredicate.java File fe/src/main/java/org/apache/impala/analysis/BinaryPredicate.java: http://gerrit.cloudera.org:8080/#/c/16346/10/fe/src/main/java/org/apache/impala/analysis/BinaryPredicate.java@95 PS10, Line 95: public static final com.google.common.base.Predicate<BinaryPredicate> > line too long (92 > 90) Done http://gerrit.cloudera.org:8080/#/c/16346/10/fe/src/main/java/org/apache/impala/analysis/Expr.java File fe/src/main/java/org/apache/impala/analysis/Expr.java: http://gerrit.cloudera.org:8080/#/c/16346/10/fe/src/main/java/org/apache/impala/analysis/Expr.java@1234 PS10, Line 1234: BitSet changed = propagateConstants(tmpConjuncts, candidates, keepConjuncts, > line too long (95 > 90) Done http://gerrit.cloudera.org:8080/#/c/16346/9/testdata/workloads/functional-planner/queries/PlannerTest/constant-propagation.test File testdata/workloads/functional-planner/queries/PlannerTest/constant-propagation.test: http://gerrit.cloudera.org:8080/#/c/16346/9/testdata/workloads/functional-planner/queries/PlannerTest/constant-propagation.test@419 PS9, Line 419: predicates: timestamp_col <= TIMESTAMP '2009-02-01 00:00:00', timestamp_col >= TIMESTAMP '2009-01-01 00:00:00', date_col = CAST(timestamp_col AS DATE) > Made the code change to preserve the original conjunct date_col = cast(tim I changed the alltypes_date_partition table to have rows with odd values of 'id' to have matching timestamp_col and date_col values and even values of 'id' to have non-matching (I added an INTERVAL to the timestamp). The same e2e test from previous patchset is used. http://gerrit.cloudera.org:8080/#/c/16346/9/testdata/workloads/functional-query/queries/QueryTest/range-constant-propagation.test File testdata/workloads/functional-query/queries/QueryTest/range-constant-propagation.test: http://gerrit.cloudera.org:8080/#/c/16346/9/testdata/workloads/functional-query/queries/QueryTest/range-constant-propagation.test@4 PS9, Line 4: alltypes_date_part > I would be ok with running with other data sets but I had some struggles in Removed the functional_parquet prefix. Also, I since these tests or the planner tests don't need parquet data, I modified the data generation script to only generate text version of alltypes_date_partition. Also reduced the size of the table so it does not run into the default maximum partitions limit. -- To view, visit http://gerrit.cloudera.org:8080/16346 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I811a1f8d605c27c7704d7fc759a91510c6db3c2b Gerrit-Change-Number: 16346 Gerrit-PatchSet: 11 Gerrit-Owner: Aman Sinha <[email protected]> Gerrit-Reviewer: Aman Sinha <[email protected]> Gerrit-Reviewer: Impala Public Jenkins <[email protected]> Gerrit-Reviewer: Qifan Chen <[email protected]> Gerrit-Reviewer: Shant Hovsepian <[email protected]> Gerrit-Reviewer: Tim Armstrong <[email protected]> Gerrit-Comment-Date: Wed, 02 Sep 2020 07:09:12 +0000 Gerrit-HasComments: Yes
