You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@calcite.apache.org by "Danny Chen (Jira)" <ji...@apache.org> on 2019/11/16 06:09:00 UTC
[jira] [Commented] (CALCITE-3507) Mismatched type AssertionError or
incorrect logical plan on correlated query
[ https://issues.apache.org/jira/browse/CALCITE-3507?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16975601#comment-16975601 ]
Danny Chen commented on CALCITE-3507:
-------------------------------------
Thanks for reporting this [~tc260], supply a test case in RelOptRulesTest would help a lot to clarify this problem ~
> Mismatched type AssertionError or incorrect logical plan on correlated query
> ----------------------------------------------------------------------------
>
> Key: CALCITE-3507
> URL: https://issues.apache.org/jira/browse/CALCITE-3507
> Project: Calcite
> Issue Type: Bug
> Affects Versions: 1.21.0
> Reporter: Tiangang Chen
> Priority: Major
>
> Calcite fails to convert correlated SQL query to logical plan because of AssertionError (type mismatch) during decorrelateQuery. Forcing types to match results in incorrect logical plan.
> Stack trace:
>
> {code:java}
> Exception in thread "main" java.lang.AssertionError: mismatched type $1 VARCHAR
> at org.apache.calcite.rex.RexUtil$FixNullabilityShuttle.visitInputRef(RexUtil.java:2540)
> at org.apache.calcite.rex.RexUtil$FixNullabilityShuttle.visitInputRef(RexUtil.java:2518)
> at org.apache.calcite.rex.RexInputRef.accept(RexInputRef.java:112)
> at org.apache.calcite.rex.RexShuttle.visitList(RexShuttle.java:149)
> at org.apache.calcite.rex.RexShuttle.visitCall(RexShuttle.java:101)
> at org.apache.calcite.rex.RexShuttle.visitCall(RexShuttle.java:34)
> at org.apache.calcite.rex.RexCall.accept(RexCall.java:191)
> at org.apache.calcite.rex.RexShuttle.apply(RexShuttle.java:277)
> at org.apache.calcite.rex.RexShuttle.mutate(RexShuttle.java:239)
> at org.apache.calcite.rex.RexShuttle.apply(RexShuttle.java:257)
> at org.apache.calcite.rex.RexUtil.fixUp(RexUtil.java:1635)
> at org.apache.calcite.rel.rules.FilterJoinRule.perform(FilterJoinRule.java:249)
> at org.apache.calcite.rel.rules.FilterJoinRule$FilterIntoJoinRule.onMatch(FilterJoinRule.java:383)
> at org.apache.calcite.plan.AbstractRelOptPlanner.fireRule(AbstractRelOptPlanner.java:319)
> at org.apache.calcite.plan.hep.HepPlanner.applyRule(HepPlanner.java:560)
> at org.apache.calcite.plan.hep.HepPlanner.applyRules(HepPlanner.java:419)
> at org.apache.calcite.plan.hep.HepPlanner.executeInstruction(HepPlanner.java:256)
> at org.apache.calcite.plan.hep.HepInstruction$RuleInstance.execute(HepInstruction.java:127)
> at org.apache.calcite.plan.hep.HepPlanner.executeProgram(HepPlanner.java:215)
> at org.apache.calcite.plan.hep.HepPlanner.findBestExp(HepPlanner.java:202)
> at org.apache.calcite.sql2rel.RelDecorrelator.decorrelate(RelDecorrelator.java:250)
> at org.apache.calcite.sql2rel.RelDecorrelator.decorrelateQuery(RelDecorrelator.java:215)
> at org.apache.calcite.prepare.PlannerImpl.rel(PlannerImpl.java:259)
> at RunWithPlanner.main(RunWithPlanner.java:17)
> {code}
>
>
> Code to reproduce:
> {code:java}
> SchemaPlus rootSchema = Frameworks.createRootSchema(true);
> rootSchema.add("LIKES", new AbstractTable() {
> public RelDataType getRowType(final RelDataTypeFactory typeFactory) {
> RelDataTypeFactory.Builder builder = typeFactory.builder();
> builder.add("DRINKER", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> builder.add("BEER", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> return builder.build();
> }
> });
> rootSchema.add("SERVES", new AbstractTable() {
> public RelDataType getRowType(final RelDataTypeFactory typeFactory) {
> RelDataTypeFactory.Builder builder = typeFactory.builder();
> builder.add("BAR", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> builder.add("BEER", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> builder.add("PRICE", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.DECIMAL));
> return builder.build();
> }
> });
> rootSchema.add("DRINKER", new AbstractTable() {
> public RelDataType getRowType(final RelDataTypeFactory typeFactory) {
> RelDataTypeFactory.Builder builder = typeFactory.builder();
> builder.add("NAME", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> builder.add("ADDRESS", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> return builder.build();
> }
> });
> rootSchema.add("FREQUENTS", new AbstractTable() {
> public RelDataType getRowType(final RelDataTypeFactory typeFactory) {
> RelDataTypeFactory.Builder builder = typeFactory.builder();
> builder.add("DRINKER", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> builder.add("BAR", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.VARCHAR));
> builder.add("TIMES_A_WEEK", new BasicSqlType(new RelDataTypeSystemImpl() {
> }, SqlTypeName.SMALLINT));
> return builder.build();
> }
> });
> String SAMPLE_QUERY = "SELECT Frequents.drinker, Frequents.bar\n" +
> "FROM Frequents\n" +
> "WHERE NOT EXISTS(SELECT * FROM Likes, Serves\n" +
> "WHERE Likes.drinker = Frequents.drinker\n" +
> "AND Serves.bar = Frequents.bar\n" +
> "AND Likes.beer = Serves.beer)";
> Planner planner = Frameworks.getPlanner(Frameworks.newConfigBuilder().defaultSchema(rootSchema).build());
> SqlNode parsed = planner.parse(SAMPLE_QUERY);
> SqlNode validated = planner.validate(parsed);
> RelRoot relRoot = planner.rel(validated);
> System.out.println(RelOptUtil.toString(relRoot.rel));{code}
> Using character types (VARCHAR, CHAR) for above schema will cause the exception. Forcing all the types to be non-character types (say INTEGER, BOOLEAN, etc.) will not cause exception but the resulting logical plan is incorrect because of one projection on the wrong columns (see the marked line below):
> {code:java}
> LogicalProject(DRINKER=[$0], BAR=[$1])
> LogicalFilter(condition=[IS NULL($5)])
> LogicalJoin(condition=[AND(=($0, $3), =($1, $4))], joinType=[left])
> EnumerableTableScan(table=[[FREQUENTS]])
> LogicalAggregate(group=[{0, 1}], agg#0=[MIN($2)])
> LogicalProject(DRINKER1=[$7], BAR00=[$8], $f0=[true]) [this is incorrect]
> LogicalJoin(condition=[=($1, $5)], joinType=[inner])
> LogicalJoin(condition=[=($0, $2)], joinType=[inner])
> EnumerableTableScan(table=[[LIKES]])
> LogicalAggregate(group=[{0, 1}])
> LogicalProject(DRINKER=[$0], BAR=[$1])
> EnumerableTableScan(table=[[FREQUENTS]])
> LogicalJoin(condition=[=($0, $4)], joinType=[inner])
> EnumerableTableScan(table=[[SERVES]])
> LogicalAggregate(group=[{0, 1}])
> LogicalProject(DRINKER=[$0], BAR=[$1])
> EnumerableTableScan(table=[[FREQUENTS]])
> {code}
> This projection:
> LogicalProject(DRINKER1=[$7], BAR00=[$8], $f0=[true])
> is on the wrong columns and therefore will produce incorrect query result.
> Correct columns could be DRINKER1=[$0], BAR00=[$4]
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)