Tim Armstrong has posted comments on this change. ( http://gerrit.cloudera.org:8080/16052 )
Change subject: IMPALA-9784: Non correlated subqueries in HAVING. ...................................................................... Patch Set 2: (8 comments) I think this makes sense, doing the nested query thing and converting it to a where clause does simplify it a lot. http://gerrit.cloudera.org:8080/#/c/16052/2/fe/src/main/java/org/apache/impala/analysis/StmtRewriter.java File fe/src/main/java/org/apache/impala/analysis/StmtRewriter.java: http://gerrit.cloudera.org:8080/#/c/16052/2/fe/src/main/java/org/apache/impala/analysis/StmtRewriter.java@465 PS2, Line 465: // TODO: Remove this when independent subquery evaluation is implemented. Can we add a simple planner test for this rewrite? With the cardinality check node added. http://gerrit.cloudera.org:8080/#/c/16052/2/fe/src/main/java/org/apache/impala/analysis/StmtRewriter.java@466 PS2, Line 466: // TODO: IMPALA-5100 to cover all cases, we do let through runtime scalars with > I relaxed some of these rules to let through subqueries such as (select cou It might be helpful to file a separate JIRA with an example of this query, just so we have some record that we can run some new queries. Some simple things like: select * from functional.alltypes where id < (select count(bool_col) from functional.alltypes where int_col=1 group by int_col); http://gerrit.cloudera.org:8080/#/c/16052/2/fe/src/main/java/org/apache/impala/analysis/StmtRewriter.java@469 PS2, Line 469: presensce nit: presence http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/functional-query/queries/QueryTest/subquery.test File testdata/workloads/functional-query/queries/QueryTest/subquery.test: http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/functional-query/queries/QueryTest/subquery.test@1218 PS2, Line 1218: from functional.alltypestiny group by id nit: remove the functional database name here and below - the test framework should set the current db based on the file format in the test matrix.. A lot of other queries here don't follow the best practice. http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q23-1.test File testdata/workloads/tpcds/queries/tpcds-q23-1.test: http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q23-1.test@52 PS2, Line 52: ---- RESULTS Compared to https://github.com/gregrahn/tpcds-kit/blob/master/answer_sets/23_NULLS_LAST.ans http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q23-2.test File testdata/workloads/tpcds/queries/tpcds-q23-2.test: http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q23-2.test@59 PS2, Line 59: 'Brown','Monika',6031.52 Compared to https://github.com/gregrahn/tpcds-kit/blob/master/answer_sets/23_NULLS_LAST.ans http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q24-2.test File testdata/workloads/tpcds/queries/tpcds-q24-2.test: http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q24-2.test@53 PS2, Line 53: 'Hamlin','Heather','able',149.65 Compared to https://github.com/gregrahn/tpcds-kit/blob/master/answer_sets/24.ans http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q44.test File testdata/workloads/tpcds/queries/tpcds-q44.test: http://gerrit.cloudera.org:8080/#/c/16052/2/testdata/workloads/tpcds/queries/tpcds-q44.test@51 PS2, Line 51: 1,'oughtantiprin st','callyeingbarcallyought' Compared this to https://github.com/gregrahn/tpcds-kit/blob/master/answer_sets/44.ans -- To view, visit http://gerrit.cloudera.org:8080/16052 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-MessageType: comment Gerrit-Change-Id: I124a58a09a1a47e1222a22d84b54fe7d07844461 Gerrit-Change-Number: 16052 Gerrit-PatchSet: 2 Gerrit-Owner: Shant Hovsepian <sh...@cloudera.com> Gerrit-Reviewer: Aman Sinha <amsi...@cloudera.com> Gerrit-Reviewer: Impala Public Jenkins <impala-public-jenk...@cloudera.com> Gerrit-Reviewer: Shant Hovsepian <sh...@cloudera.com> Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-Comment-Date: Fri, 26 Jun 2020 19:35:58 +0000 Gerrit-HasComments: Yes
