impala-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Marcel Kornacker (Code Review)" <>
Subject [Impala-ASF-CR] IMPALA-1788: Fold constant expressions.
Date Tue, 22 Nov 2016 22:18:47 GMT
Marcel Kornacker has posted comments on this change.

Change subject: IMPALA-1788: Fold constant expressions.

Patch Set 2:

File be/src/exprs/

Line 249
that looks weird.

Line 472:     case TYPE_TIMESTAMP:
do we have end-to-end tests with timestamp literals?
File be/src/service/

Line 105:   query_ctx.request.query_options.max_errors = 10;
File fe/src/main/java/org/apache/impala/analysis/

Line 423:   private void getResultTypesAndLabels(StatementBase stmt, List<Type> resultTypes,
it looks like a better approach is to add a StatementBase.getResultExprs and .getColLabels
and override appropriately in the subclasses.

Line 447:   private void setResultTypesAndLabels(StatementBase stmt, List<Type> resultTypes,
move this to StatementBase? (maybe as castResultExprs() and setColLabels()?)
File fe/src/main/java/org/apache/impala/analysis/

Line 177:       System.out.println(e.getMessage());
File fe/src/main/java/org/apache/impala/rewrite/

Line 48:     for (Expr child: expr.getChildren()) {
single line
File fe/src/test/java/org/apache/impala/testutil/

Line 260:     // Disable rewrites because some analyzer tests have non-executable constant
does this mean they're disabled in the planner tests?
File testdata/workloads/functional-planner/queries/PlannerTest/constant-folding.test:

Line 121: |  output: sum(2 + id), count:merge(*)
> Agree. It's not super easy to fix because the agg info is computed during a
get rid of which part? i agree we don't need the 1+1 here.
File testdata/workloads/functional-planner/queries/PlannerTest/insert.test:

Line 551: WRITE TO HDFS [functional.alltypes, OVERWRITE=false, PARTITION-KEYS=(2009,5)]
oh nice
File testdata/workloads/functional-planner/queries/PlannerTest/kudu-selectivity.test:

Line 9:      tuple-ids=0 row-size=68B cardinality=unavailable
what happened here, did something get screwed up when loading kudu data?
File testdata/workloads/functional-planner/queries/PlannerTest/subquery-rewrite.test:

Line 1270:    predicates: g.bigint_col = 1, g.bigint_col < 1000
do you know why these get reordered? it seems like =1 should have been the first one to start

To view, visit
To unsubscribe, visit

Gerrit-MessageType: comment
Gerrit-Change-Id: If672b703db1ba0bfc26e5b9130161798b40a69e9
Gerrit-PatchSet: 2
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Alex Behm <>
Gerrit-Reviewer: Alex Behm <>
Gerrit-Reviewer: Marcel Kornacker <>
Gerrit-HasComments: Yes

View raw message