org.apache.calcite.sql2rel.SqlToRelConverter Maven / Gradle / Ivy
Show all versions of flink-table-planner_2.12 Show documentation
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to you under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.calcite.sql2rel;
import org.apache.flink.table.api.TableConfig;
import org.apache.flink.table.data.TimestampData;
import org.apache.flink.table.planner.calcite.FlinkSqlCallBinding;
import org.apache.flink.table.planner.calcite.TimestampSchemaVersion;
import org.apache.flink.table.planner.hint.ClearQueryHintsWithInvalidPropagationShuttle;
import org.apache.flink.table.planner.hint.FlinkHints;
import org.apache.flink.table.planner.plan.FlinkCalciteCatalogSnapshotReader;
import org.apache.flink.table.planner.plan.utils.FlinkRelOptUtil;
import org.apache.flink.table.planner.utils.ShortcutUtils;
import com.google.common.base.Preconditions;
import com.google.common.collect.ImmutableList;
import com.google.common.collect.ImmutableMap;
import com.google.common.collect.ImmutableSet;
import com.google.common.collect.Iterables;
import org.apache.calcite.avatica.util.Spaces;
import org.apache.calcite.jdbc.CalciteSchema;
import org.apache.calcite.linq4j.Ord;
import org.apache.calcite.linq4j.tree.TableExpressionFactory;
import org.apache.calcite.plan.RelOptCluster;
import org.apache.calcite.plan.RelOptPlanner;
import org.apache.calcite.plan.RelOptSamplingParameters;
import org.apache.calcite.plan.RelOptTable;
import org.apache.calcite.plan.RelOptUtil;
import org.apache.calcite.plan.RelTraitSet;
import org.apache.calcite.plan.ViewExpanders;
import org.apache.calcite.prepare.Prepare;
import org.apache.calcite.prepare.RelOptTableImpl;
import org.apache.calcite.rel.RelCollation;
import org.apache.calcite.rel.RelCollationTraitDef;
import org.apache.calcite.rel.RelCollations;
import org.apache.calcite.rel.RelFieldCollation;
import org.apache.calcite.rel.RelNode;
import org.apache.calcite.rel.RelRoot;
import org.apache.calcite.rel.RelShuttleImpl;
import org.apache.calcite.rel.SingleRel;
import org.apache.calcite.rel.core.Aggregate;
import org.apache.calcite.rel.core.AggregateCall;
import org.apache.calcite.rel.core.Collect;
import org.apache.calcite.rel.core.CorrelationId;
import org.apache.calcite.rel.core.Filter;
import org.apache.calcite.rel.core.Join;
import org.apache.calcite.rel.core.JoinInfo;
import org.apache.calcite.rel.core.JoinRelType;
import org.apache.calcite.rel.core.Project;
import org.apache.calcite.rel.core.RelFactories;
import org.apache.calcite.rel.core.Sample;
import org.apache.calcite.rel.core.Sort;
import org.apache.calcite.rel.hint.HintStrategyTable;
import org.apache.calcite.rel.hint.Hintable;
import org.apache.calcite.rel.hint.RelHint;
import org.apache.calcite.rel.logical.LogicalAggregate;
import org.apache.calcite.rel.logical.LogicalCorrelate;
import org.apache.calcite.rel.logical.LogicalFilter;
import org.apache.calcite.rel.logical.LogicalIntersect;
import org.apache.calcite.rel.logical.LogicalJoin;
import org.apache.calcite.rel.logical.LogicalMatch;
import org.apache.calcite.rel.logical.LogicalMinus;
import org.apache.calcite.rel.logical.LogicalProject;
import org.apache.calcite.rel.logical.LogicalSort;
import org.apache.calcite.rel.logical.LogicalTableFunctionScan;
import org.apache.calcite.rel.logical.LogicalTableModify;
import org.apache.calcite.rel.logical.LogicalTableScan;
import org.apache.calcite.rel.logical.LogicalUnion;
import org.apache.calcite.rel.logical.LogicalValues;
import org.apache.calcite.rel.metadata.RelColumnMapping;
import org.apache.calcite.rel.metadata.RelMetadataQuery;
import org.apache.calcite.rel.stream.Delta;
import org.apache.calcite.rel.stream.LogicalDelta;
import org.apache.calcite.rel.type.RelDataType;
import org.apache.calcite.rel.type.RelDataTypeFactory;
import org.apache.calcite.rel.type.RelDataTypeField;
import org.apache.calcite.rex.RexBuilder;
import org.apache.calcite.rex.RexCall;
import org.apache.calcite.rex.RexCorrelVariable;
import org.apache.calcite.rex.RexDynamicParam;
import org.apache.calcite.rex.RexFieldAccess;
import org.apache.calcite.rex.RexFieldCollation;
import org.apache.calcite.rex.RexInputRef;
import org.apache.calcite.rex.RexLiteral;
import org.apache.calcite.rex.RexNode;
import org.apache.calcite.rex.RexPatternFieldRef;
import org.apache.calcite.rex.RexRangeRef;
import org.apache.calcite.rex.RexShuttle;
import org.apache.calcite.rex.RexSubQuery;
import org.apache.calcite.rex.RexUtil;
import org.apache.calcite.rex.RexWindowBound;
import org.apache.calcite.rex.RexWindowBounds;
import org.apache.calcite.schema.ColumnStrategy;
import org.apache.calcite.schema.ModifiableTable;
import org.apache.calcite.schema.ModifiableView;
import org.apache.calcite.schema.SchemaVersion;
import org.apache.calcite.schema.Schemas;
import org.apache.calcite.schema.Table;
import org.apache.calcite.schema.TranslatableTable;
import org.apache.calcite.schema.Wrapper;
import org.apache.calcite.sql.JoinConditionType;
import org.apache.calcite.sql.JoinType;
import org.apache.calcite.sql.SqlAggFunction;
import org.apache.calcite.sql.SqlBasicCall;
import org.apache.calcite.sql.SqlCall;
import org.apache.calcite.sql.SqlCallBinding;
import org.apache.calcite.sql.SqlDataTypeSpec;
import org.apache.calcite.sql.SqlDelete;
import org.apache.calcite.sql.SqlDynamicParam;
import org.apache.calcite.sql.SqlExplainFormat;
import org.apache.calcite.sql.SqlExplainLevel;
import org.apache.calcite.sql.SqlFunction;
import org.apache.calcite.sql.SqlIdentifier;
import org.apache.calcite.sql.SqlInsert;
import org.apache.calcite.sql.SqlIntervalQualifier;
import org.apache.calcite.sql.SqlJoin;
import org.apache.calcite.sql.SqlKind;
import org.apache.calcite.sql.SqlLiteral;
import org.apache.calcite.sql.SqlMatchRecognize;
import org.apache.calcite.sql.SqlMerge;
import org.apache.calcite.sql.SqlNode;
import org.apache.calcite.sql.SqlNodeList;
import org.apache.calcite.sql.SqlNumericLiteral;
import org.apache.calcite.sql.SqlOperator;
import org.apache.calcite.sql.SqlOperatorTable;
import org.apache.calcite.sql.SqlOrderBy;
import org.apache.calcite.sql.SqlPivot;
import org.apache.calcite.sql.SqlSampleSpec;
import org.apache.calcite.sql.SqlSelect;
import org.apache.calcite.sql.SqlSelectKeyword;
import org.apache.calcite.sql.SqlSetOperator;
import org.apache.calcite.sql.SqlSnapshot;
import org.apache.calcite.sql.SqlTableRef;
import org.apache.calcite.sql.SqlUnnestOperator;
import org.apache.calcite.sql.SqlUnpivot;
import org.apache.calcite.sql.SqlUpdate;
import org.apache.calcite.sql.SqlUtil;
import org.apache.calcite.sql.SqlValuesOperator;
import org.apache.calcite.sql.SqlWindow;
import org.apache.calcite.sql.SqlWith;
import org.apache.calcite.sql.SqlWithItem;
import org.apache.calcite.sql.fun.SqlCase;
import org.apache.calcite.sql.fun.SqlInOperator;
import org.apache.calcite.sql.fun.SqlQuantifyOperator;
import org.apache.calcite.sql.fun.SqlRowOperator;
import org.apache.calcite.sql.fun.SqlStdOperatorTable;
import org.apache.calcite.sql.parser.SqlParserPos;
import org.apache.calcite.sql.type.SqlReturnTypeInference;
import org.apache.calcite.sql.type.SqlTypeName;
import org.apache.calcite.sql.type.SqlTypeUtil;
import org.apache.calcite.sql.type.TableFunctionReturnTypeInference;
import org.apache.calcite.sql.util.SqlBasicVisitor;
import org.apache.calcite.sql.util.SqlVisitor;
import org.apache.calcite.sql.validate.AggregatingSelectScope;
import org.apache.calcite.sql.validate.CollectNamespace;
import org.apache.calcite.sql.validate.DelegatingScope;
import org.apache.calcite.sql.validate.ListScope;
import org.apache.calcite.sql.validate.MatchRecognizeScope;
import org.apache.calcite.sql.validate.ParameterScope;
import org.apache.calcite.sql.validate.SelectScope;
import org.apache.calcite.sql.validate.SqlMonotonicity;
import org.apache.calcite.sql.validate.SqlNameMatcher;
import org.apache.calcite.sql.validate.SqlQualified;
import org.apache.calcite.sql.validate.SqlUserDefinedTableFunction;
import org.apache.calcite.sql.validate.SqlUserDefinedTableMacro;
import org.apache.calcite.sql.validate.SqlValidator;
import org.apache.calcite.sql.validate.SqlValidatorImpl;
import org.apache.calcite.sql.validate.SqlValidatorNamespace;
import org.apache.calcite.sql.validate.SqlValidatorScope;
import org.apache.calcite.sql.validate.SqlValidatorTable;
import org.apache.calcite.sql.validate.SqlValidatorUtil;
import org.apache.calcite.tools.RelBuilder;
import org.apache.calcite.tools.RelBuilderFactory;
import org.apache.calcite.util.ImmutableBitSet;
import org.apache.calcite.util.ImmutableIntList;
import org.apache.calcite.util.Litmus;
import org.apache.calcite.util.NlsString;
import org.apache.calcite.util.NumberUtil;
import org.apache.calcite.util.Pair;
import org.apache.calcite.util.TimestampString;
import org.apache.calcite.util.Util;
import org.apache.calcite.util.trace.CalciteTrace;
import org.checkerframework.checker.nullness.qual.Nullable;
import org.immutables.value.Value;
import org.slf4j.Logger;
import java.lang.reflect.Type;
import java.math.BigDecimal;
import java.time.ZoneId;
import java.util.AbstractList;
import java.util.ArrayDeque;
import java.util.ArrayList;
import java.util.BitSet;
import java.util.Collection;
import java.util.Collections;
import java.util.Deque;
import java.util.EnumSet;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Objects;
import java.util.Set;
import java.util.TreeSet;
import java.util.function.BiFunction;
import java.util.function.Supplier;
import java.util.function.UnaryOperator;
import java.util.stream.Collectors;
import static java.util.Objects.requireNonNull;
import static org.apache.calcite.linq4j.Nullness.castNonNull;
import static org.apache.calcite.sql.SqlUtil.stripAs;
import static org.apache.flink.util.Preconditions.checkNotNull;
/**
* Converts a SQL parse tree (consisting of {@link org.apache.calcite.sql.SqlNode} objects) into a
* relational algebra expression (consisting of {@link org.apache.calcite.rel.RelNode} objects).
*
* The public entry points are: {@link #convertQuery}, {@link #convertExpression(SqlNode)}.
*
*
FLINK modifications are at lines
*
*
* - Added in FLINK-29081, FLINK-28682, FLINK-33395: Lines 655 ~ 673
*
- Added in Flink-24024: Lines 1437 ~ 1447, Lines 1461 ~ 1503
*
- Added in FLINK-28682: Lines 2325 ~ 2342
*
- Added in FLINK-28682: Lines 2379 ~ 2407
*
- Added in FLINK-32474: Lines 2877 ~ 2889
*
- Added in FLINK-32474: Lines 2989 ~ 3023
*
- Added in FLINK-20873: Lines 5521 ~ 5530
*
- Added in FLINK-34312: Lines 5641 ~ 5644
*
- Added in FLINK-34057, FLINK-34058, FLINK-34312: Lines 6093 ~ 6111
*
*/
@SuppressWarnings("UnstableApiUsage")
@Value.Enclosing
public class SqlToRelConverter {
// ~ Static fields/initializers ---------------------------------------------
/** Default configuration. */
public static final Config CONFIG =
ImmutableSqlToRelConverter.Config.builder()
.withRelBuilderFactory(RelFactories.LOGICAL_BUILDER)
.withRelBuilderConfigTransform(c -> c.withPushJoinCondition(true))
.withHintStrategyTable(HintStrategyTable.EMPTY)
.build();
protected static final Logger SQL2REL_LOGGER = CalciteTrace.getSqlToRelTracer();
/** Size of the smallest IN list that will be converted to a semijoin to a static table. */
public static final int DEFAULT_IN_SUB_QUERY_THRESHOLD = 20;
@Deprecated // to be removed before 2.0
public static final int DEFAULT_IN_SUBQUERY_THRESHOLD = DEFAULT_IN_SUB_QUERY_THRESHOLD;
// ~ Instance fields --------------------------------------------------------
public final @Nullable SqlValidator validator;
protected final RexBuilder rexBuilder;
protected final Prepare.CatalogReader catalogReader;
protected final RelOptCluster cluster;
private SubQueryConverter subQueryConverter;
protected final Map leaves = new HashMap<>();
private final List<@Nullable SqlDynamicParam> dynamicParamSqlNodes = new ArrayList<>();
private final SqlOperatorTable opTab;
protected final RelDataTypeFactory typeFactory;
private final SqlNodeToRexConverter exprConverter;
private final HintStrategyTable hintStrategies;
private int explainParamCount;
public final SqlToRelConverter.Config config;
private final RelBuilder relBuilder;
/** Fields used in name resolution for correlated sub-queries. */
private final Map mapCorrelToDeferred = new HashMap<>();
/**
* Stack of names of datasets requested by the
* TABLE(SAMPLE(<datasetName>, <query>)) construct.
*/
private final Deque datasetStack = new ArrayDeque<>();
/**
* Mapping of non-correlated sub-queries that have been converted to their equivalent constants.
* Used to avoid re-evaluating the sub-query if it's already been evaluated.
*/
private final Map mapConvertedNonCorrSubqs = new HashMap<>();
public final RelOptTable.ViewExpander viewExpander;
// ~ Constructors -----------------------------------------------------------
/**
* Creates a converter.
*
* @param viewExpander Preparing statement
* @param validator Validator
* @param catalogReader Schema
* @param planner Planner
* @param rexBuilder Rex builder
* @param convertletTable Expression converter
*/
@Deprecated // to be removed before 2.0
public SqlToRelConverter(
RelOptTable.ViewExpander viewExpander,
SqlValidator validator,
Prepare.CatalogReader catalogReader,
RelOptPlanner planner,
RexBuilder rexBuilder,
SqlRexConvertletTable convertletTable) {
this(
viewExpander,
validator,
catalogReader,
RelOptCluster.create(planner, rexBuilder),
convertletTable,
SqlToRelConverter.config());
}
@Deprecated // to be removed before 2.0
public SqlToRelConverter(
RelOptTable.ViewExpander viewExpander,
SqlValidator validator,
Prepare.CatalogReader catalogReader,
RelOptCluster cluster,
SqlRexConvertletTable convertletTable) {
this(
viewExpander,
validator,
catalogReader,
cluster,
convertletTable,
SqlToRelConverter.config());
}
/* Creates a converter. */
public SqlToRelConverter(
RelOptTable.ViewExpander viewExpander,
@Nullable SqlValidator validator,
Prepare.CatalogReader catalogReader,
RelOptCluster cluster,
SqlRexConvertletTable convertletTable,
Config config) {
this.viewExpander = viewExpander;
this.opTab =
(validator == null) ? SqlStdOperatorTable.instance() : validator.getOperatorTable();
this.validator = validator;
this.catalogReader = catalogReader;
this.subQueryConverter = new NoOpSubQueryConverter();
this.rexBuilder = cluster.getRexBuilder();
this.typeFactory = rexBuilder.getTypeFactory();
this.exprConverter = new SqlNodeToRexConverterImpl(convertletTable);
this.explainParamCount = 0;
this.config = requireNonNull(config, "config");
this.relBuilder =
config.getRelBuilderFactory()
.create(cluster, null)
.transform(config.getRelBuilderConfigTransform());
this.hintStrategies = config.getHintStrategyTable();
cluster.setHintStrategies(this.hintStrategies);
this.cluster = requireNonNull(cluster, "cluster");
}
// ~ Methods ----------------------------------------------------------------
private SqlValidator validator() {
return requireNonNull(validator, "validator");
}
private T getNamespace(SqlNode node) {
//noinspection unchecked
return (T)
requireNonNull(
getNamespaceOrNull(node), () -> "Namespace is not found for " + node);
}
@SuppressWarnings("unchecked")
private @Nullable T getNamespaceOrNull(SqlNode node) {
return (@Nullable T) validator().getNamespace(node);
}
/** Returns the RelOptCluster in use. */
public RelOptCluster getCluster() {
return cluster;
}
/** Returns the row-expression builder. */
public RexBuilder getRexBuilder() {
return rexBuilder;
}
/**
* Returns the number of dynamic parameters encountered during translation; this must only be
* called after {@link #convertQuery}.
*
* @return number of dynamic parameters
*/
public int getDynamicParamCount() {
return dynamicParamSqlNodes.size();
}
/**
* Returns the type inferred for a dynamic parameter.
*
* @param index 0-based index of dynamic parameter
* @return inferred type, never null
*/
public RelDataType getDynamicParamType(int index) {
SqlNode sqlNode = dynamicParamSqlNodes.get(index);
if (sqlNode == null) {
throw Util.needToImplement("dynamic param type inference");
}
return validator().getValidatedNodeType(sqlNode);
}
/**
* Returns the current count of the number of dynamic parameters in an EXPLAIN PLAN statement.
*
* @param increment if true, increment the count
* @return the current count before the optional increment
*/
public int getDynamicParamCountInExplain(boolean increment) {
int retVal = explainParamCount;
if (increment) {
++explainParamCount;
}
return retVal;
}
/**
* Returns the mapping of non-correlated sub-queries that have been converted to the constants
* that they evaluate to.
*/
public Map getMapConvertedNonCorrSubqs() {
return mapConvertedNonCorrSubqs;
}
/**
* Adds to the current map of non-correlated converted sub-queries the elements from another map
* that contains non-correlated sub-queries that have been converted by another
* SqlToRelConverter.
*
* @param alreadyConvertedNonCorrSubqs the other map
*/
public void addConvertedNonCorrSubqs(Map alreadyConvertedNonCorrSubqs) {
mapConvertedNonCorrSubqs.putAll(alreadyConvertedNonCorrSubqs);
}
/**
* Sets a new SubQueryConverter. To have any effect, this must be called before any convert
* method.
*
* @param converter new SubQueryConverter
*/
public void setSubQueryConverter(SubQueryConverter converter) {
subQueryConverter = converter;
}
/**
* Sets the number of dynamic parameters in the current EXPLAIN PLAN statement.
*
* @param explainParamCount number of dynamic parameters in the statement
*/
public void setDynamicParamCountInExplain(int explainParamCount) {
assert config.isExplain();
this.explainParamCount = explainParamCount;
}
private void checkConvertedType(SqlNode query, RelNode result) {
if (query.isA(SqlKind.DML)) {
return;
}
// Verify that conversion from SQL to relational algebra did
// not perturb any type information. (We can't do this if the
// SQL statement is something like an INSERT which has no
// validator type information associated with its result,
// hence the namespace check above.)
final List validatedFields =
validator().getValidatedNodeType(query).getFieldList();
final RelDataType validatedRowType =
validator()
.getTypeFactory()
.createStructType(
Pair.right(validatedFields),
SqlValidatorUtil.uniquify(
Pair.left(validatedFields),
catalogReader.nameMatcher().isCaseSensitive()));
final List convertedFields =
result.getRowType().getFieldList().subList(0, validatedFields.size());
final RelDataType convertedRowType =
validator().getTypeFactory().createStructType(convertedFields);
if (!RelOptUtil.equal(
"validated row type",
validatedRowType,
"converted row type",
convertedRowType,
Litmus.IGNORE)) {
throw new AssertionError(
"Conversion to relational algebra failed to "
+ "preserve datatypes:\n"
+ "validated type:\n"
+ validatedRowType.getFullTypeString()
+ "\nconverted type:\n"
+ convertedRowType.getFullTypeString()
+ "\nrel:\n"
+ RelOptUtil.toString(result));
}
}
public RelNode flattenTypes(RelNode rootRel, boolean restructure) {
RelStructuredTypeFlattener typeFlattener =
new RelStructuredTypeFlattener(
relBuilder,
rexBuilder,
createToRelContext(ImmutableList.of()),
restructure);
return typeFlattener.rewrite(rootRel);
}
/**
* If sub-query is correlated and decorrelation is enabled, performs decorrelation.
*
* @param query Query
* @param rootRel Root relational expression
* @return New root relational expression after decorrelation
*/
public RelNode decorrelate(SqlNode query, RelNode rootRel) {
if (!config.isDecorrelationEnabled()) {
return rootRel;
}
final RelNode result = decorrelateQuery(rootRel);
if (result != rootRel) {
checkConvertedType(query, result);
}
return result;
}
/**
* Walks over a tree of relational expressions, replacing each {@link RelNode} with a 'slimmed
* down' relational expression that projects only the fields required by its consumer.
*
* This may make things easier for the optimizer, by removing crud that would expand the
* search space, but is difficult for the optimizer itself to do it, because optimizer rules
* must preserve the number and type of fields. Hence, this transform that operates on the
* entire tree, similar to the {@link RelStructuredTypeFlattener type-flattening transform}.
*
*
Currently this functionality is disabled in farrago/luciddb; the default implementation of
* this method does nothing.
*
* @param ordered Whether the relational expression must produce results in a particular order
* (typically because it has an ORDER BY at top level)
* @param rootRel Relational expression that is at the root of the tree
* @return Trimmed relational expression
*/
public RelNode trimUnusedFields(boolean ordered, RelNode rootRel) {
// Trim fields that are not used by their consumer.
if (config.isTrimUnusedFields()) {
final RelFieldTrimmer trimmer = newFieldTrimmer();
final List collations =
rootRel.getTraitSet().getTraits(RelCollationTraitDef.INSTANCE);
rootRel = trimmer.trim(rootRel);
if (!ordered
&& collations != null
&& !collations.isEmpty()
&& !collations.equals(ImmutableList.of(RelCollations.EMPTY))) {
final RelTraitSet traitSet =
rootRel.getTraitSet().replace(RelCollationTraitDef.INSTANCE, collations);
rootRel = rootRel.copy(traitSet, rootRel.getInputs());
}
if (SQL2REL_LOGGER.isDebugEnabled()) {
SQL2REL_LOGGER.debug(
RelOptUtil.dumpPlan(
"Plan after trimming unused fields",
rootRel,
SqlExplainFormat.TEXT,
SqlExplainLevel.EXPPLAN_ATTRIBUTES));
}
}
return rootRel;
}
/**
* Creates a RelFieldTrimmer.
*
* @return Field trimmer
*/
protected RelFieldTrimmer newFieldTrimmer() {
return new RelFieldTrimmer(validator, relBuilder);
}
/**
* Converts an unvalidated query's parse tree into a relational expression.
*
* @param query Query to convert
* @param needsValidation Whether to validate the query before converting; false if
* the query has already been validated.
* @param top Whether the query is top-level, say if its result will become a JDBC result set;
* false if the query will be part of a view.
*/
public RelRoot convertQuery(SqlNode query, final boolean needsValidation, final boolean top) {
if (needsValidation) {
query = validator().validate(query);
}
RelNode result = convertQueryRecursive(query, top, null).rel;
if (top) {
if (isStream(query)) {
result = new LogicalDelta(cluster, result.getTraitSet(), result);
}
}
RelCollation collation = RelCollations.EMPTY;
if (!query.isA(SqlKind.DML)) {
if (isOrdered(query)) {
collation = requiredCollation(result);
}
}
checkConvertedType(query, result);
if (SQL2REL_LOGGER.isDebugEnabled()) {
SQL2REL_LOGGER.debug(
RelOptUtil.dumpPlan(
"Plan after converting SqlNode to RelNode",
result,
SqlExplainFormat.TEXT,
SqlExplainLevel.EXPPLAN_ATTRIBUTES));
}
final RelDataType validatedRowType = validator().getValidatedNodeType(query);
List hints = new ArrayList<>();
if (query.getKind() == SqlKind.SELECT) {
final SqlSelect select = (SqlSelect) query;
if (select.hasHints()) {
hints = SqlUtil.getRelHint(hintStrategies, select.getHints());
}
}
if (config.isAddJsonTypeOperatorEnabled()) {
result = result.accept(new NestedJsonFunctionRelRewriter());
}
// ----- FLINK MODIFICATION BEGIN -----
// propagate the hints.
// The method FlinkRelOptUtil#propagateRelHints not only finds and propagates hints
// throughout the entire rel tree but also within subqueries.
result = FlinkRelOptUtil.propagateRelHints(result, false);
// replace all query hints with upper case
result = FlinkHints.capitalizeQueryHints(result);
// clear query hints which are propagated into wrong query block
// The hint QueryBlockAlias will be added when building a RelNode tree before. It is used to
// distinguish the query block in the SQL.
result = result.accept(new ClearQueryHintsWithInvalidPropagationShuttle());
// clear the hints on some nodes where these hints should not be attached
result = FlinkHints.clearQueryHintsOnUnmatchedNodes(result);
// ----- FLINK MODIFICATION END -----
return RelRoot.of(result, validatedRowType, query.getKind())
.withCollation(collation)
.withHints(hints);
}
private static boolean isStream(SqlNode query) {
return query instanceof SqlSelect
&& ((SqlSelect) query).isKeywordPresent(SqlSelectKeyword.STREAM);
}
public static boolean isOrdered(SqlNode query) {
switch (query.getKind()) {
case SELECT:
SqlNodeList orderList = ((SqlSelect) query).getOrderList();
return orderList != null && orderList.size() > 0;
case WITH:
return isOrdered(((SqlWith) query).body);
case ORDER_BY:
return ((SqlOrderBy) query).orderList.size() > 0;
default:
return false;
}
}
private static RelCollation requiredCollation(RelNode r) {
if (r instanceof Sort) {
return ((Sort) r).collation;
}
if (r instanceof Project) {
return requiredCollation(((Project) r).getInput());
}
if (r instanceof Delta) {
return requiredCollation(((Delta) r).getInput());
}
throw new AssertionError();
}
/** Converts a SELECT statement's parse tree into a relational expression. */
public RelNode convertSelect(SqlSelect select, boolean top) {
final SqlValidatorScope selectScope = validator().getWhereScope(select);
final Blackboard bb = createBlackboard(selectScope, null, top);
convertSelectImpl(bb, select);
return castNonNull(bb.root);
}
/** Factory method for creating translation workspace. */
protected Blackboard createBlackboard(
@Nullable SqlValidatorScope scope,
@Nullable Map nameToNodeMap,
boolean top) {
return new Blackboard(scope, nameToNodeMap, top);
}
/** Implementation of {@link #convertSelect(SqlSelect, boolean)}; derived class may override. */
protected void convertSelectImpl(final Blackboard bb, SqlSelect select) {
convertFrom(bb, select.getFrom());
// We would like to remove ORDER BY clause from an expanded view, except if
// it is top-level or affects semantics.
//
// Top-level example. Given the view definition
// CREATE VIEW v AS SELECT * FROM t ORDER BY x
// we would retain the view's ORDER BY in
// SELECT * FROM v
// or
// SELECT * FROM v WHERE y = 5
// but remove the view's ORDER BY in
// SELECT * FROM v ORDER BY z
// and
// SELECT deptno, COUNT(*) FROM v GROUP BY deptno
// because the ORDER BY and GROUP BY mean that the view is not 'top level' in
// the query.
//
// Semantics example. Given the view definition
// CREATE VIEW v2 AS SELECT * FROM t ORDER BY x LIMIT 10
// we would never remove the ORDER BY, because "ORDER BY ... LIMIT" is about
// semantics. It is not a 'pure order'.
if (RelOptUtil.isPureOrder(castNonNull(bb.root)) && config.isRemoveSortInSubQuery()) {
// Remove the Sort if the view is at the top level. Also remove the Sort
// if there are other nodes, which will cause the view to be in the
// sub-query.
if (!bb.top
|| validator().isAggregate(select)
|| select.isDistinct()
|| select.hasOrderBy()
|| select.getFetch() != null
|| select.getOffset() != null) {
bb.setRoot(castNonNull(bb.root).getInput(0), true);
}
}
convertWhere(bb, select.getWhere());
final List orderExprList = new ArrayList<>();
final List collationList = new ArrayList<>();
gatherOrderExprs(bb, select, select.getOrderList(), orderExprList, collationList);
final RelCollation collation = cluster.traitSet().canonize(RelCollations.of(collationList));
if (validator().isAggregate(select)) {
convertAgg(bb, select, orderExprList);
} else {
convertSelectList(bb, select, orderExprList);
}
if (select.isDistinct()) {
distinctify(bb, true);
}
convertOrder(select, bb, collation, orderExprList, select.getOffset(), select.getFetch());
if (select.hasHints()) {
final List hints = SqlUtil.getRelHint(hintStrategies, select.getHints());
// Attach the hints to the first Hintable node we found from the root node.
bb.setRoot(
bb.root()
.accept(
new RelShuttleImpl() {
boolean attached = false;
@Override
public RelNode visitChild(
RelNode parent, int i, RelNode child) {
if (parent instanceof Hintable && !attached) {
attached = true;
return ((Hintable) parent).attachHints(hints);
} else {
return super.visitChild(parent, i, child);
}
}
}),
true);
} else {
bb.setRoot(bb.root(), true);
}
}
/**
* Having translated 'SELECT ... FROM ... [GROUP BY ...] [HAVING ...]', adds a relational
* expression to make the results unique.
*
* If the SELECT clause contains duplicate expressions, adds {@link
* org.apache.calcite.rel.logical.LogicalProject}s so that we are grouping on the minimal set of
* keys. The performance gain isn't huge, but it is difficult to detect these duplicate
* expressions later.
*
* @param bb Blackboard
* @param checkForDupExprs Check for duplicate expressions
*/
private void distinctify(Blackboard bb, boolean checkForDupExprs) {
// Look for duplicate expressions in the project.
// Say we have 'select x, y, x, z'.
// Then dups will be {[2, 0]}
// and oldToNew will be {[0, 0], [1, 1], [2, 0], [3, 2]}
RelNode rel = bb.root;
if (checkForDupExprs && (rel instanceof LogicalProject)) {
LogicalProject project = (LogicalProject) rel;
final List projectExprs = project.getProjects();
final List origins = new ArrayList<>();
int dupCount = 0;
for (int i = 0; i < projectExprs.size(); i++) {
int x = projectExprs.indexOf(projectExprs.get(i));
if (x >= 0 && x < i) {
origins.add(x);
++dupCount;
} else {
origins.add(i);
}
}
if (dupCount == 0) {
distinctify(bb, false);
return;
}
final Map squished = new HashMap<>();
final List fields = rel.getRowType().getFieldList();
final List> newProjects = new ArrayList<>();
for (int i = 0; i < fields.size(); i++) {
if (origins.get(i) == i) {
squished.put(i, newProjects.size());
newProjects.add(RexInputRef.of2(i, fields));
}
}
rel =
LogicalProject.create(
rel,
ImmutableList.of(),
Pair.left(newProjects),
Pair.right(newProjects));
bb.root = rel;
distinctify(bb, false);
rel = bb.root();
// Create the expressions to reverse the mapping.
// Project($0, $1, $0, $2).
final List> undoProjects = new ArrayList<>();
for (int i = 0; i < fields.size(); i++) {
final int origin = origins.get(i);
RelDataTypeField field = fields.get(i);
undoProjects.add(
Pair.of(
new RexInputRef(castNonNull(squished.get(origin)), field.getType()),
field.getName()));
}
rel =
LogicalProject.create(
rel,
ImmutableList.of(),
Pair.left(undoProjects),
Pair.right(undoProjects));
bb.setRoot(rel, false);
return;
}
assert rel != null : "rel must not be null, root = " + bb.root;
// Usual case: all of the expressions in the SELECT clause are
// different.
final ImmutableBitSet groupSet = ImmutableBitSet.range(rel.getRowType().getFieldCount());
rel = createAggregate(bb, groupSet, ImmutableList.of(groupSet), ImmutableList.of());
bb.setRoot(rel, false);
}
/**
* Converts a query's ORDER BY clause, if any.
*
* Ignores the ORDER BY clause if the query is not top-level and FETCH or OFFSET are not
* present.
*
* @param select Query
* @param bb Blackboard
* @param collation Collation list
* @param orderExprList Method populates this list with orderBy expressions not present in
* selectList
* @param offset Expression for number of rows to discard before returning first row
* @param fetch Expression for number of rows to fetch
*/
protected void convertOrder(
SqlSelect select,
Blackboard bb,
RelCollation collation,
List orderExprList,
@Nullable SqlNode offset,
@Nullable SqlNode fetch) {
if (removeSortInSubQuery(bb.top)
|| select.getOrderList() == null
|| select.getOrderList().isEmpty()) {
assert removeSortInSubQuery(bb.top) || collation.getFieldCollations().isEmpty();
if ((offset == null
|| (offset instanceof SqlLiteral
&& Objects.equals(
((SqlLiteral) offset).bigDecimalValue(),
BigDecimal.ZERO)))
&& fetch == null) {
return;
}
}
// Create a sorter using the previously constructed collations.
bb.setRoot(
LogicalSort.create(
bb.root(),
collation,
offset == null ? null : convertExpression(offset),
fetch == null ? null : convertExpression(fetch)),
false);
// If extra expressions were added to the project list for sorting,
// add another project to remove them. But make the collation empty, because
// we can't represent the real collation.
//
// If it is the top node, use the real collation, but don't trim fields.
if (orderExprList.size() > 0 && !bb.top) {
final List exprs = new ArrayList<>();
final RelDataType rowType = bb.root().getRowType();
final int fieldCount = rowType.getFieldCount() - orderExprList.size();
for (int i = 0; i < fieldCount; i++) {
exprs.add(rexBuilder.makeInputRef(bb.root(), i));
}
bb.setRoot(
LogicalProject.create(
bb.root(),
ImmutableList.of(),
exprs,
rowType.getFieldNames().subList(0, fieldCount)),
false);
}
}
/**
* Returns whether we should remove the sort for the subsequent query conversion.
*
* @param top Whether the rel to convert is the root of the query
*/
private boolean removeSortInSubQuery(boolean top) {
return config.isRemoveSortInSubQuery() && !top;
}
/**
* Returns whether a given node contains a {@link SqlInOperator}.
*
* @param node a RexNode tree
*/
private static boolean containsInOperator(SqlNode node) {
try {
SqlVisitor visitor =
new SqlBasicVisitor() {
@Override
public Void visit(SqlCall call) {
if (call.getOperator() instanceof SqlInOperator) {
throw new Util.FoundOne(call);
}
return super.visit(call);
}
};
node.accept(visitor);
return false;
} catch (Util.FoundOne e) {
Util.swallow(e, null);
return true;
}
}
/**
* Push down all the NOT logical operators into any IN/NOT IN operators.
*
* @param scope Scope where {@code sqlNode} occurs
* @param sqlNode the root node from which to look for NOT operators
* @return the transformed SqlNode representation with NOT pushed down.
*/
private static SqlNode pushDownNotForIn(SqlValidatorScope scope, SqlNode sqlNode) {
if (!(sqlNode instanceof SqlCall) || !containsInOperator(sqlNode)) {
return sqlNode;
}
final SqlCall sqlCall = (SqlCall) sqlNode;
switch (sqlCall.getKind()) {
case AND:
case OR:
final List operands = new ArrayList<>();
for (SqlNode operand : sqlCall.getOperandList()) {
operands.add(pushDownNotForIn(scope, operand));
}
final SqlCall newCall =
sqlCall.getOperator().createCall(sqlCall.getParserPosition(), operands);
return reg(scope, newCall);
case NOT:
assert sqlCall.operand(0) instanceof SqlCall;
final SqlCall call = sqlCall.operand(0);
switch (sqlCall.operand(0).getKind()) {
case CASE:
final SqlCase caseNode = (SqlCase) call;
final SqlNodeList thenOperands = new SqlNodeList(SqlParserPos.ZERO);
for (SqlNode thenOperand : caseNode.getThenOperands()) {
final SqlCall not =
SqlStdOperatorTable.NOT.createCall(
SqlParserPos.ZERO, thenOperand);
thenOperands.add(pushDownNotForIn(scope, reg(scope, not)));
}
SqlNode elseOperand =
requireNonNull(
caseNode.getElseOperand(),
"getElseOperand for " + caseNode);
if (!SqlUtil.isNull(elseOperand)) {
// "not(unknown)" is "unknown", so no need to simplify
final SqlCall not =
SqlStdOperatorTable.NOT.createCall(
SqlParserPos.ZERO, elseOperand);
elseOperand = pushDownNotForIn(scope, reg(scope, not));
}
return reg(
scope,
SqlStdOperatorTable.CASE.createCall(
SqlParserPos.ZERO,
caseNode.getValueOperand(),
caseNode.getWhenOperands(),
thenOperands,
elseOperand));
case AND:
final List orOperands = new ArrayList<>();
for (SqlNode operand : call.getOperandList()) {
orOperands.add(
pushDownNotForIn(
scope,
reg(
scope,
SqlStdOperatorTable.NOT.createCall(
SqlParserPos.ZERO, operand))));
}
return reg(
scope,
SqlStdOperatorTable.OR.createCall(SqlParserPos.ZERO, orOperands));
case OR:
final List andOperands = new ArrayList<>();
for (SqlNode operand : call.getOperandList()) {
andOperands.add(
pushDownNotForIn(
scope,
reg(
scope,
SqlStdOperatorTable.NOT.createCall(
SqlParserPos.ZERO, operand))));
}
return reg(
scope,
SqlStdOperatorTable.AND.createCall(SqlParserPos.ZERO, andOperands));
case NOT:
assert call.operandCount() == 1;
return pushDownNotForIn(scope, call.operand(0));
case NOT_IN:
return reg(
scope,
SqlStdOperatorTable.IN.createCall(
SqlParserPos.ZERO, call.getOperandList()));
case IN:
return reg(
scope,
SqlStdOperatorTable.NOT_IN.createCall(
SqlParserPos.ZERO, call.getOperandList()));
default:
break;
}
break;
default:
break;
}
return sqlNode;
}
/**
* Registers with the validator a {@link SqlNode} that has been created during the Sql-to-Rel
* process.
*/
private static SqlNode reg(SqlValidatorScope scope, SqlNode e) {
scope.getValidator().deriveType(scope, e);
return e;
}
/**
* Converts a WHERE clause.
*
* @param bb Blackboard
* @param where WHERE clause, may be null
*/
private void convertWhere(final Blackboard bb, final @Nullable SqlNode where) {
if (where == null) {
return;
}
SqlNode newWhere = pushDownNotForIn(bb.scope(), where);
replaceSubQueries(bb, newWhere, RelOptUtil.Logic.UNKNOWN_AS_FALSE);
final RexNode convertedWhere = bb.convertExpression(newWhere);
final RexNode convertedWhere2 = RexUtil.removeNullabilityCast(typeFactory, convertedWhere);
// only allocate filter if the condition is not TRUE
if (convertedWhere2.isAlwaysTrue()) {
return;
}
final RelFactories.FilterFactory filterFactory = RelFactories.DEFAULT_FILTER_FACTORY;
final RelNode filter =
filterFactory.createFilter(bb.root(), convertedWhere2, ImmutableSet.of());
final RelNode r;
final CorrelationUse p = getCorrelationUse(bb, filter);
if (p != null) {
assert p.r instanceof Filter;
Filter f = (Filter) p.r;
r = LogicalFilter.create(f.getInput(), f.getCondition(), ImmutableSet.of(p.id));
} else {
r = filter;
}
bb.setRoot(r, false);
}
private void replaceSubQueries(
final Blackboard bb, final SqlNode expr, RelOptUtil.Logic logic) {
findSubQueries(bb, expr, logic, false);
for (SubQuery node : bb.subQueryList) {
substituteSubQuery(bb, node);
}
}
private void substituteSubQuery(Blackboard bb, SubQuery subQuery) {
final RexNode expr = subQuery.expr;
if (expr != null) {
// Already done.
return;
}
final SqlBasicCall call;
final RelNode rel;
final SqlNode query;
final RelOptUtil.Exists converted;
switch (subQuery.node.getKind()) {
case CURSOR:
convertCursor(bb, subQuery);
return;
case ARRAY_QUERY_CONSTRUCTOR:
case MAP_QUERY_CONSTRUCTOR:
case MULTISET_QUERY_CONSTRUCTOR:
if (!config.isExpand()) {
return;
}
// fall through
case MULTISET_VALUE_CONSTRUCTOR:
rel = convertMultisets(ImmutableList.of(subQuery.node), bb);
subQuery.expr = bb.register(rel, JoinRelType.INNER);
return;
case IN:
case NOT_IN:
case SOME:
case ALL:
call = (SqlBasicCall) subQuery.node;
query = call.operand(1);
if (!config.isExpand() && !(query instanceof SqlNodeList)) {
return;
}
final SqlNode leftKeyNode = call.operand(0);
final List leftKeys;
switch (leftKeyNode.getKind()) {
case ROW:
leftKeys = new ArrayList<>();
for (SqlNode sqlExpr : ((SqlBasicCall) leftKeyNode).getOperandList()) {
leftKeys.add(bb.convertExpression(sqlExpr));
}
break;
default:
leftKeys = ImmutableList.of(bb.convertExpression(leftKeyNode));
}
if (query instanceof SqlNodeList) {
SqlNodeList valueList = (SqlNodeList) query;
// When the list size under the threshold or the list references columns, we
// convert to OR.
if (valueList.size() < config.getInSubQueryThreshold()
|| valueList.accept(new SqlIdentifierFinder())) {
subQuery.expr =
convertInToOr(
bb,
leftKeys,
valueList,
(SqlInOperator) call.getOperator());
return;
}
// Otherwise, let convertExists translate
// values list into an inline table for the
// reference to Q below.
}
// Project out the search columns from the left side
// Q1:
// "select from emp where emp.deptno in (select col1 from T)"
//
// is converted to
//
// "select from
// emp inner join (select distinct col1 from T)) q
// on emp.deptno = q.col1
//
// Q2:
// "select from emp where emp.deptno not in (Q)"
//
// is converted to
//
// "select from
// emp left outer join (select distinct col1, TRUE from T) q
// on emp.deptno = q.col1
// where emp.deptno <> null
// and q.indicator <> TRUE"
//
// Note: Sub-query can be used as SqlUpdate#condition like below:
//
// UPDATE emp
// SET empno = 1 WHERE emp.empno IN (
// SELECT emp.empno FROM emp WHERE emp.empno = 2)
//
// In such case, when converting SqlUpdate#condition, bb.root is null
// and it makes no sense to do the sub-query substitution.
if (bb.root == null) {
return;
}
final RelDataType targetRowType =
SqlTypeUtil.promoteToRowType(
typeFactory, validator().getValidatedNodeType(leftKeyNode), null);
final boolean notIn = call.getOperator().kind == SqlKind.NOT_IN;
converted =
convertExists(
query,
RelOptUtil.SubQueryType.IN,
subQuery.logic,
notIn,
targetRowType);
if (converted.indicator) {
// Generate
// emp CROSS JOIN (SELECT COUNT(*) AS c,
// COUNT(deptno) AS ck FROM dept)
final RelDataType longType = typeFactory.createSqlType(SqlTypeName.BIGINT);
final RelNode seek = converted.r.getInput(0); // fragile
final int keyCount = leftKeys.size();
final List args = ImmutableIntList.range(0, keyCount);
LogicalAggregate aggregate =
LogicalAggregate.create(
seek,
ImmutableList.of(),
ImmutableBitSet.of(),
null,
ImmutableList.of(
AggregateCall.create(
SqlStdOperatorTable.COUNT,
false,
false,
false,
ImmutableList.of(),
-1,
null,
RelCollations.EMPTY,
longType,
null),
AggregateCall.create(
SqlStdOperatorTable.COUNT,
false,
false,
false,
args,
-1,
null,
RelCollations.EMPTY,
longType,
null)));
LogicalJoin join =
LogicalJoin.create(
bb.root(),
aggregate,
ImmutableList.of(),
rexBuilder.makeLiteral(true),
ImmutableSet.of(),
JoinRelType.INNER);
bb.setRoot(join, false);
}
final RexNode rex =
bb.register(
converted.r,
converted.outerJoin ? JoinRelType.LEFT : JoinRelType.INNER,
leftKeys);
RelOptUtil.Logic logic = subQuery.logic;
switch (logic) {
case TRUE_FALSE_UNKNOWN:
case UNKNOWN_AS_TRUE:
if (!converted.indicator) {
logic = RelOptUtil.Logic.TRUE_FALSE;
}
break;
default:
break;
}
subQuery.expr = translateIn(logic, bb.root, rex);
if (notIn) {
subQuery.expr = rexBuilder.makeCall(SqlStdOperatorTable.NOT, subQuery.expr);
}
return;
case EXISTS:
// "select from emp where exists (select a from T)"
//
// is converted to the following if the sub-query is correlated:
//
// "select from emp left outer join (select AGG_TRUE() as indicator
// from T group by corr_var) q where q.indicator is true"
//
// If there is no correlation, the expression is replaced with a
// boolean indicating whether the sub-query returned 0 or >= 1 row.
if (!config.isExpand()) {
return;
}
call = (SqlBasicCall) subQuery.node;
query = call.operand(0);
final SqlValidatorScope seekScope =
(query instanceof SqlSelect)
? validator().getSelectScope((SqlSelect) query)
: null;
final Blackboard seekBb = createBlackboard(seekScope, null, false);
final RelNode seekRel = convertQueryOrInList(seekBb, query, null);
requireNonNull(seekRel, () -> "seekRel is null for query " + query);
// An EXIST sub-query whose inner child has at least 1 tuple
// (e.g. an Aggregate with no grouping columns or non-empty Values
// node) should be simplified to a Boolean constant expression.
final RelMetadataQuery mq = seekRel.getCluster().getMetadataQuery();
final Double minRowCount = mq.getMinRowCount(seekRel);
if (minRowCount != null && minRowCount >= 1D) {
subQuery.expr = rexBuilder.makeLiteral(true);
return;
}
converted =
RelOptUtil.createExistsPlan(
seekRel,
RelOptUtil.SubQueryType.EXISTS,
subQuery.logic,
true,
relBuilder);
assert !converted.indicator;
if (convertNonCorrelatedSubQuery(subQuery, bb, converted.r, true)) {
return;
}
subQuery.expr = bb.register(converted.r, JoinRelType.LEFT);
return;
case UNIQUE:
return;
case SCALAR_QUERY:
// Convert the sub-query. If it's non-correlated, convert it
// to a constant expression.
if (!config.isExpand()) {
return;
}
call = (SqlBasicCall) subQuery.node;
query = call.operand(0);
converted =
convertExists(
query, RelOptUtil.SubQueryType.SCALAR, subQuery.logic, true, null);
assert !converted.indicator;
if (convertNonCorrelatedSubQuery(subQuery, bb, converted.r, false)) {
return;
}
rel = convertToSingleValueSubq(query, converted.r);
subQuery.expr = bb.register(rel, JoinRelType.LEFT);
return;
case SELECT:
// This is used when converting multiset queries:
//
// select * from unnest(select multiset[deptno] from emps);
//
converted =
convertExists(
subQuery.node,
RelOptUtil.SubQueryType.SCALAR,
subQuery.logic,
true,
null);
assert !converted.indicator;
subQuery.expr = bb.register(converted.r, JoinRelType.LEFT);
// This is used when converting window table functions:
//
// select * from table(tumble(table emps, descriptor(deptno), interval '3' DAY))
//
bb.cursors.add(converted.r);
return;
case SET_SEMANTICS_TABLE:
// ----- FLINK MODIFICATION BEGIN -----
// We always expand the SET SEMANTICS TABLE for two reasons:
// 1. Calcite has a bug when not expanding the SET SEMANTICS TABLE. For more
// information, see CALCITE-6204.
// 2. Currently, Flink’s built-in Session Window TVF is the only PTF with SET
// SEMANTICS. We will expand it by default, like other built-in window TVFs, to
// reuse some subsequent processing and optimization logic.
// if (!config.isExpand()) {
// return;
// }
// ----- FLINK MODIFICATION END -----
substituteSubQueryOfSetSemanticsInputTable(bb, subQuery);
return;
default:
throw new AssertionError("unexpected kind of sub-query: " + subQuery.node);
}
}
private void substituteSubQueryOfSetSemanticsInputTable(Blackboard bb, SubQuery subQuery) {
SqlBasicCall call;
SqlNode query;
call = (SqlBasicCall) subQuery.node;
query = call.operand(0);
// FLINK MODIFICATION BEGIN
// We modified it for two reasons:
// 1. In Flink, Exchange nodes should not appear in the logical stage, which will bring
// uncertainty to the implementation of plan optimization in the current logical stage.
// Instead, Flink will add exchanges based on traits during the physical phase.
// 2. Currently, Flink’s built-in Session Window TVF is the only SET SEMANTICS
// TABLE. We will convert it into the same plan tree as other Window TVFs. The partition key
// and order key will be recorded using a custom RexCall when subsequently converting the
// SqlCall of SET SEMANTICS TABLE. See more at
// FlinkConvertletTable#convertSetSemanticsWindowTableFunction
final RelNode inputOfSetSemanticsTable =
convertQueryRecursive(query, false, null).project();
relBuilder.push(inputOfSetSemanticsTable);
// final SqlValidatorScope innerTableScope =
// (query instanceof SqlSelect) ? validator().getSelectScope((SqlSelect) query) :
// null;
// final Blackboard setSemanticsTableBb = createBlackboard(innerTableScope, null, false);
// final RelNode inputOfSetSemanticsTable =
// convertQueryRecursive(query, false, null).project();
// relBuilder.push(inputOfSetSemanticsTable);
// requireNonNull(inputOfSetSemanticsTable, () -> "input RelNode is null for query " +
// query);
// SqlNodeList partitionList = call.operand(1);
// final ImmutableBitSet partitionKeys =
// buildPartitionKeys(setSemanticsTableBb, partitionList);
// For set semantics table, distribution is singleton if does not specify
// partition keys
// RelDistribution distribution =
// partitionKeys.isEmpty()
// ? RelDistributions.SINGLETON
// : RelDistributions.hash(partitionKeys.asList());
// ORDER BY
// final SqlNodeList orderList = call.operand(2);
// final RelCollation orders = buildCollation(setSemanticsTableBb, orderList);
// if (orderList.isEmpty()) {
// relBuilder.exchange(distribution);
// } else {
// relBuilder.sortExchange(distribution, orders);
// }
// FLINK MODIFICATION END
RelNode tableRel = relBuilder.build();
subQuery.expr = bb.register(tableRel, JoinRelType.LEFT);
// This is used when converting window table functions:
//
// select * from table(tumble(table emps, descriptor(deptno),
// interval '3' DAY))
//
bb.cursors.add(tableRel);
}
private ImmutableBitSet buildPartitionKeys(Blackboard bb, SqlNodeList partitionList) {
final ImmutableBitSet.Builder partitionKeys = ImmutableBitSet.builder();
for (SqlNode partition : partitionList) {
validator().deriveType(bb.scope(), partition);
RexNode e = bb.convertExpression(partition);
partitionKeys.set(parseFieldIdx(e));
}
return partitionKeys.build();
}
/**
* Note: The ORDER BY clause for input table parameter differs from the ORDER BY clause in some
* other contexts in that only columns may be sorted (not arbitrary expressions).
*
* @param bb Scope within which to resolve identifiers
* @param orderList Order by clause, may be null
* @return ordering of input table
*/
private RelCollation buildCollation(Blackboard bb, SqlNodeList orderList) {
final List orderKeys = new ArrayList<>();
for (SqlNode orderItem : orderList) {
orderKeys.add(
convertOrderItem(
bb,
orderItem,
RelFieldCollation.Direction.ASCENDING,
RelFieldCollation.NullDirection.UNSPECIFIED));
}
return cluster.traitSet().canonize(RelCollations.of(orderKeys));
}
private RelFieldCollation convertOrderItem(
Blackboard bb,
SqlNode orderItem,
RelFieldCollation.Direction direction,
RelFieldCollation.NullDirection nullDirection) {
switch (orderItem.getKind()) {
case DESCENDING:
return convertOrderItem(
bb,
((SqlCall) orderItem).operand(0),
RelFieldCollation.Direction.DESCENDING,
nullDirection);
case NULLS_FIRST:
return convertOrderItem(
bb,
((SqlCall) orderItem).operand(0),
direction,
RelFieldCollation.NullDirection.FIRST);
case NULLS_LAST:
return convertOrderItem(
bb,
((SqlCall) orderItem).operand(0),
direction,
RelFieldCollation.NullDirection.LAST);
default:
break;
}
switch (nullDirection) {
case UNSPECIFIED:
nullDirection =
validator().config().defaultNullCollation().last(desc(direction))
? RelFieldCollation.NullDirection.LAST
: RelFieldCollation.NullDirection.FIRST;
break;
default:
break;
}
RexNode e = bb.convertExpression(orderItem);
return new RelFieldCollation(parseFieldIdx(e), direction, nullDirection);
}
private static int parseFieldIdx(RexNode e) {
switch (e.getKind()) {
case FIELD_ACCESS:
final RexFieldAccess f = (RexFieldAccess) e;
return f.getField().getIndex();
case INPUT_REF:
final RexInputRef ref = (RexInputRef) e;
return ref.getIndex();
default:
throw new AssertionError();
}
}
private RexNode translateIn(RelOptUtil.Logic logic, @Nullable RelNode root, final RexNode rex) {
switch (logic) {
case TRUE:
return rexBuilder.makeLiteral(true);
case TRUE_FALSE:
case UNKNOWN_AS_FALSE:
assert rex instanceof RexRangeRef;
final int fieldCount = rex.getType().getFieldCount();
RexNode rexNode = rexBuilder.makeFieldAccess(rex, fieldCount - 1);
rexNode = rexBuilder.makeCall(SqlStdOperatorTable.IS_TRUE, rexNode);
// Then append the IS NOT NULL(leftKeysForIn).
//
// RexRangeRef contains the following fields:
// leftKeysForIn,
// rightKeysForIn (the original sub-query select list),
// nullIndicator
//
// The first two lists contain the same number of fields.
final int k = (fieldCount - 1) / 2;
ImmutableList.Builder rexNodeBuilder = ImmutableList.builder();
rexNodeBuilder.add(rexNode);
for (int i = 0; i < k; i++) {
rexNodeBuilder.add(
rexBuilder.makeCall(
SqlStdOperatorTable.IS_NOT_NULL,
rexBuilder.makeFieldAccess(rex, i)));
}
rexNode =
rexBuilder.makeCall(
rexNode.getType(),
SqlStdOperatorTable.AND,
RexUtil.flatten(rexNodeBuilder.build(), SqlStdOperatorTable.AND));
return rexNode;
case TRUE_FALSE_UNKNOWN:
case UNKNOWN_AS_TRUE:
// select e.deptno,
// case
// when ct.c = 0 then false
// when dt.i is not null then true
// when e.deptno is null then null
// when ct.ck < ct.c then null
// else false
// end
// from e
// cross join (select count(*) as c, count(deptno) as ck from v) as ct
// left join (select distinct deptno, true as i from v) as dt
// on e.deptno = dt.deptno
final Join join = (Join) requireNonNull(root, "root");
final Project left = (Project) join.getLeft();
final RelNode leftLeft = ((Join) left.getInput()).getLeft();
final int leftLeftCount = leftLeft.getRowType().getFieldCount();
final RelDataType longType = typeFactory.createSqlType(SqlTypeName.BIGINT);
final RexNode cRef = rexBuilder.makeInputRef(root, leftLeftCount);
final RexNode ckRef = rexBuilder.makeInputRef(root, leftLeftCount + 1);
final RexNode iRef =
rexBuilder.makeInputRef(root, root.getRowType().getFieldCount() - 1);
final RexLiteral zero = rexBuilder.makeExactLiteral(BigDecimal.ZERO, longType);
final RexLiteral trueLiteral = rexBuilder.makeLiteral(true);
final RexLiteral falseLiteral = rexBuilder.makeLiteral(false);
final RexNode unknownLiteral = rexBuilder.makeNullLiteral(trueLiteral.getType());
final ImmutableList.Builder args = ImmutableList.builder();
args.add(
rexBuilder.makeCall(SqlStdOperatorTable.EQUALS, cRef, zero),
falseLiteral,
rexBuilder.makeCall(SqlStdOperatorTable.IS_NOT_NULL, iRef),
trueLiteral);
final JoinInfo joinInfo = join.analyzeCondition();
for (int leftKey : joinInfo.leftKeys) {
final RexNode kRef = rexBuilder.makeInputRef(root, leftKey);
args.add(
rexBuilder.makeCall(SqlStdOperatorTable.IS_NULL, kRef), unknownLiteral);
}
args.add(
rexBuilder.makeCall(SqlStdOperatorTable.LESS_THAN, ckRef, cRef),
unknownLiteral,
falseLiteral);
return rexBuilder.makeCall(SqlStdOperatorTable.CASE, args.build());
default:
throw new AssertionError(logic);
}
}
/**
* Determines if a sub-query is non-correlated and if so, converts it to a constant.
*
* @param subQuery the call that references the sub-query
* @param bb blackboard used to convert the sub-query
* @param converted RelNode tree corresponding to the sub-query
* @param isExists true if the sub-query is part of an EXISTS expression
* @return Whether the sub-query can be converted to a constant
*/
private boolean convertNonCorrelatedSubQuery(
SubQuery subQuery, Blackboard bb, RelNode converted, boolean isExists) {
SqlCall call = (SqlBasicCall) subQuery.node;
if (subQueryConverter.canConvertSubQuery() && isSubQueryNonCorrelated(converted, bb)) {
// First check if the sub-query has already been converted
// because it's a nested sub-query. If so, don't re-evaluate
// it again.
RexNode constExpr = mapConvertedNonCorrSubqs.get(call);
if (constExpr == null) {
constExpr =
subQueryConverter.convertSubQuery(call, this, isExists, config.isExplain());
}
if (constExpr != null) {
subQuery.expr = constExpr;
mapConvertedNonCorrSubqs.put(call, constExpr);
return true;
}
}
return false;
}
/**
* Converts the RelNode tree for a select statement to a select that produces a single value.
*
* @param query the query
* @param plan the original RelNode tree corresponding to the statement
* @return the converted RelNode tree
*/
public RelNode convertToSingleValueSubq(SqlNode query, RelNode plan) {
// Check whether query is guaranteed to produce a single value.
if (query instanceof SqlSelect) {
SqlSelect select = (SqlSelect) query;
SqlNodeList selectList = select.getSelectList();
SqlNodeList groupList = select.getGroup();
if ((selectList.size() == 1) && ((groupList == null) || (groupList.size() == 0))) {
SqlNode selectExpr = selectList.get(0);
if (selectExpr instanceof SqlCall) {
SqlCall selectExprCall = (SqlCall) selectExpr;
if (Util.isSingleValue(selectExprCall)) {
return plan;
}
}
// If there is a limit with 0 or 1,
// it is ensured to produce a single value
SqlNode fetch = select.getFetch();
if (fetch instanceof SqlNumericLiteral) {
long value = ((SqlNumericLiteral) fetch).getValueAs(Long.class);
if (value < 2) {
return plan;
}
}
}
} else if (query instanceof SqlCall) {
// If the query is (values ...),
// it is necessary to look into the operands to determine
// whether SingleValueAgg is necessary
SqlCall exprCall = (SqlCall) query;
if (exprCall.getOperator() instanceof SqlValuesOperator
&& Util.isSingleValue(exprCall)) {
return plan;
}
}
// If not, project SingleValueAgg
return RelOptUtil.createSingleValueAggRel(cluster, plan);
}
/**
* Converts "x IN (1, 2, ...)" to "x=1 OR x=2 OR ...".
*
* @param leftKeys LHS
* @param valuesList RHS
* @param op The operator (IN, NOT IN, > SOME, ...)
* @return converted expression
*/
private @Nullable RexNode convertInToOr(
final Blackboard bb,
final List leftKeys,
SqlNodeList valuesList,
SqlInOperator op) {
final List comparisons = new ArrayList<>();
for (SqlNode rightVals : valuesList) {
RexNode rexComparison;
final SqlOperator comparisonOp;
if (op instanceof SqlQuantifyOperator) {
comparisonOp =
RelOptUtil.op(
((SqlQuantifyOperator) op).comparisonKind,
SqlStdOperatorTable.EQUALS);
} else {
comparisonOp = SqlStdOperatorTable.EQUALS;
}
if (leftKeys.size() == 1) {
rexComparison =
rexBuilder.makeCall(
comparisonOp,
leftKeys.get(0),
ensureSqlType(
leftKeys.get(0).getType(),
bb.convertExpression(rightVals)));
} else {
assert rightVals instanceof SqlCall;
final SqlBasicCall call = (SqlBasicCall) rightVals;
assert (call.getOperator() instanceof SqlRowOperator)
&& call.operandCount() == leftKeys.size();
rexComparison =
RexUtil.composeConjunction(
rexBuilder,
Util.transform(
Pair.zip(leftKeys, call.getOperandList()),
pair ->
rexBuilder.makeCall(
comparisonOp,
pair.left,
// TODO: remove requireNonNull when
// checkerframework issue resolved
ensureSqlType(
requireNonNull(
pair.left,
"pair.left")
.getType(),
bb.convertExpression(
pair.right)))));
}
comparisons.add(rexComparison);
}
switch (op.kind) {
case ALL:
return RexUtil.composeConjunction(rexBuilder, comparisons, true);
case NOT_IN:
return rexBuilder.makeCall(
SqlStdOperatorTable.NOT,
RexUtil.composeDisjunction(rexBuilder, comparisons));
case IN:
case SOME:
return RexUtil.composeDisjunction(rexBuilder, comparisons, true);
default:
throw new AssertionError();
}
}
/**
* Ensures that an expression has a given {@link SqlTypeName}, applying a cast if necessary. If
* the expression already has the right type family, returns the expression unchanged.
*/
private RexNode ensureSqlType(RelDataType type, RexNode node) {
if (type.getSqlTypeName() == node.getType().getSqlTypeName()
|| (type.getSqlTypeName() == SqlTypeName.VARCHAR
&& node.getType().getSqlTypeName() == SqlTypeName.CHAR)) {
return node;
}
return rexBuilder.ensureType(type, node, true);
}
/**
* Gets the list size threshold under which {@link #convertInToOr} is used. Lists of this size
* or greater will instead be converted to use a join against an inline table ({@link
* org.apache.calcite.rel.logical.LogicalValues}) rather than a predicate. A threshold of 0
* forces usage of an inline table in all cases; a threshold of Integer.MAX_VALUE forces usage
* of OR in all cases
*
* @return threshold, default {@link #DEFAULT_IN_SUB_QUERY_THRESHOLD}
*/
@Deprecated // to be removed before 2.0
protected int getInSubqueryThreshold() {
return config.getInSubQueryThreshold();
}
/**
* Converts an EXISTS or IN predicate into a join. For EXISTS, the sub-query produces an
* indicator variable, and the result is a relational expression which outer joins that
* indicator to the original query. After performing the outer join, the condition will be TRUE
* if the EXISTS condition holds, NULL otherwise.
*
* @param seek A query, for example 'select * from emp' or 'values (1,2,3)' or '('Foo', 34)'.
* @param subQueryType Whether sub-query is IN, EXISTS or scalar
* @param logic Whether the answer needs to be in full 3-valued logic (TRUE, FALSE, UNKNOWN)
* will be required, or whether we can accept an approximation (say representing UNKNOWN as
* FALSE)
* @param notIn Whether the operation is NOT IN
* @return join expression
*/
private RelOptUtil.Exists convertExists(
SqlNode seek,
RelOptUtil.SubQueryType subQueryType,
RelOptUtil.Logic logic,
boolean notIn,
@Nullable RelDataType targetDataType) {
final SqlValidatorScope seekScope =
(seek instanceof SqlSelect) ? validator().getSelectScope((SqlSelect) seek) : null;
final Blackboard seekBb = createBlackboard(seekScope, null, false);
RelNode seekRel = convertQueryOrInList(seekBb, seek, targetDataType);
requireNonNull(seekRel, () -> "seekRel is null for query " + seek);
return RelOptUtil.createExistsPlan(seekRel, subQueryType, logic, notIn, relBuilder);
}
private @Nullable RelNode convertQueryOrInList(
Blackboard bb, SqlNode seek, @Nullable RelDataType targetRowType) {
// NOTE: Once we start accepting single-row queries as row constructors,
// there will be an ambiguity here for a case like X IN ((SELECT Y FROM
// Z)). The SQL standard resolves the ambiguity by saying that a lone
// select should be interpreted as a table expression, not a row
// expression. The semantic difference is that a table expression can
// return multiple rows.
if (seek instanceof SqlNodeList) {
return convertRowValues(bb, seek, (SqlNodeList) seek, false, targetRowType);
} else {
return convertQueryRecursive(seek, false, null).project();
}
}
private @Nullable RelNode convertRowValues(
Blackboard bb,
SqlNode rowList,
Collection rows,
boolean allowLiteralsOnly,
@Nullable RelDataType targetRowType) {
// NOTE jvs 30-Apr-2006: We combine all rows consisting entirely of
// literals into a single LogicalValues; this gives the optimizer a smaller
// input tree. For everything else (computed expressions, row
// sub-queries), we union each row in as a projection on top of a
// LogicalOneRow.
final ImmutableList.Builder> tupleList = ImmutableList.builder();
final RelDataType listType = validator().getValidatedNodeType(rowList);
final RelDataType rowType;
if (targetRowType != null) {
rowType =
SqlTypeUtil.keepSourceTypeAndTargetNullability(
targetRowType, listType, typeFactory);
} else {
rowType = SqlTypeUtil.promoteToRowType(typeFactory, listType, null);
}
final List unionInputs = new ArrayList<>();
for (SqlNode node : rows) {
SqlBasicCall call;
if (isRowConstructor(node)) {
call = (SqlBasicCall) node;
ImmutableList.Builder tuple = ImmutableList.builder();
for (Ord operand : Ord.zip(call.getOperandList())) {
RexLiteral rexLiteral =
convertLiteralInValuesList(operand.e, bb, rowType, operand.i);
if ((rexLiteral == null) && allowLiteralsOnly) {
return null;
}
if ((rexLiteral == null) || !config.isCreateValuesRel()) {
// fallback to convertRowConstructor
tuple = null;
break;
}
tuple.add(rexLiteral);
}
if (tuple != null) {
tupleList.add(tuple.build());
continue;
}
} else {
RexLiteral rexLiteral = convertLiteralInValuesList(node, bb, rowType, 0);
if ((rexLiteral != null) && config.isCreateValuesRel()) {
tupleList.add(ImmutableList.of(rexLiteral));
continue;
} else {
if ((rexLiteral == null) && allowLiteralsOnly) {
return null;
}
}
// convert "1" to "row(1)"
call = (SqlBasicCall) SqlStdOperatorTable.ROW.createCall(SqlParserPos.ZERO, node);
}
unionInputs.add(convertRowConstructor(bb, call));
}
LogicalValues values = LogicalValues.create(cluster, rowType, tupleList.build());
RelNode resultRel;
if (unionInputs.isEmpty()) {
resultRel = values;
} else {
if (!values.getTuples().isEmpty()) {
unionInputs.add(values);
}
resultRel = LogicalUnion.create(unionInputs, true);
}
leaves.put(resultRel, resultRel.getRowType().getFieldCount());
return resultRel;
}
private @Nullable RexLiteral convertLiteralInValuesList(
@Nullable SqlNode sqlNode, Blackboard bb, RelDataType rowType, int iField) {
if (!(sqlNode instanceof SqlLiteral)) {
return null;
}
RelDataTypeField field = rowType.getFieldList().get(iField);
RelDataType type = field.getType();
if (type.isStruct()) {
// null literals for weird stuff like UDT's need
// special handling during type flattening, so
// don't use LogicalValues for those
return null;
}
return convertLiteral((SqlLiteral) sqlNode, bb, type);
}
private RexLiteral convertLiteral(SqlLiteral sqlLiteral, Blackboard bb, RelDataType type) {
RexNode literalExpr = exprConverter.convertLiteral(bb, sqlLiteral);
if (!(literalExpr instanceof RexLiteral)) {
assert literalExpr.isA(SqlKind.CAST);
RexNode child = ((RexCall) literalExpr).getOperands().get(0);
assert RexLiteral.isNullLiteral(child);
// NOTE jvs 22-Nov-2006: we preserve type info
// in LogicalValues digest, so it's OK to lose it here
return (RexLiteral) child;
}
RexLiteral literal = (RexLiteral) literalExpr;
Comparable value = literal.getValue();
if (SqlTypeUtil.isExactNumeric(type) && SqlTypeUtil.hasScale(type)) {
BigDecimal roundedValue =
NumberUtil.rescaleBigDecimal((BigDecimal) value, type.getScale());
return rexBuilder.makeExactLiteral(roundedValue, type);
}
if ((value instanceof NlsString) && (type.getSqlTypeName() == SqlTypeName.CHAR)) {
// pad fixed character type
NlsString unpadded = (NlsString) value;
return rexBuilder.makeCharLiteral(
new NlsString(
Spaces.padRight(unpadded.getValue(), type.getPrecision()),
unpadded.getCharsetName(),
unpadded.getCollation()));
}
return literal;
}
private static boolean isRowConstructor(SqlNode node) {
if (!(node.getKind() == SqlKind.ROW)) {
return false;
}
SqlCall call = (SqlCall) node;
return call.getOperator().getName().equalsIgnoreCase("row");
}
/**
* Builds a list of all IN or EXISTS operators inside SQL parse tree.
* Does not traverse inside queries.
*
* @param bb blackboard
* @param node the SQL parse tree
* @param logic Whether the answer needs to be in full 3-valued logic (TRUE, FALSE, UNKNOWN)
* will be required, or whether we can accept an approximation (say representing UNKNOWN as
* FALSE)
* @param registerOnlyScalarSubQueries if set to true and the parse tree corresponds to a
* variation of a select node, only register it if it's a scalar sub-query
*/
private void findSubQueries(
Blackboard bb,
SqlNode node,
RelOptUtil.Logic logic,
boolean registerOnlyScalarSubQueries) {
final SqlKind kind = node.getKind();
switch (kind) {
case EXISTS:
case UNIQUE:
case SELECT:
case MULTISET_QUERY_CONSTRUCTOR:
case MULTISET_VALUE_CONSTRUCTOR:
case ARRAY_QUERY_CONSTRUCTOR:
case MAP_QUERY_CONSTRUCTOR:
case CURSOR:
case SET_SEMANTICS_TABLE:
case SCALAR_QUERY:
if (!registerOnlyScalarSubQueries || (kind == SqlKind.SCALAR_QUERY)) {
bb.registerSubQuery(node, RelOptUtil.Logic.TRUE_FALSE);
}
return;
case IN:
break;
case NOT_IN:
case NOT:
logic = logic.negate();
break;
default:
break;
}
if (node instanceof SqlCall) {
switch (kind) {
// Do no change logic for AND, IN and NOT IN expressions;
// but do change logic for OR, NOT and others;
// EXISTS was handled already.
case AND:
case IN:
case NOT_IN:
break;
default:
logic = RelOptUtil.Logic.TRUE_FALSE_UNKNOWN;
break;
}
for (SqlNode operand : ((SqlCall) node).getOperandList()) {
if (operand != null) {
// In the case of an IN expression, locate scalar
// sub-queries so we can convert them to constants
findSubQueries(
bb,
operand,
logic,
kind == SqlKind.IN
|| kind == SqlKind.NOT_IN
|| kind == SqlKind.SOME
|| kind == SqlKind.ALL
|| registerOnlyScalarSubQueries);
}
}
} else if (node instanceof SqlNodeList) {
for (SqlNode child : (SqlNodeList) node) {
findSubQueries(
bb,
child,
logic,
kind == SqlKind.IN
|| kind == SqlKind.NOT_IN
|| kind == SqlKind.SOME
|| kind == SqlKind.ALL
|| registerOnlyScalarSubQueries);
}
}
// Now that we've located any scalar sub-queries inside the IN
// expression, register the IN expression itself. We need to
// register the scalar sub-queries first so they can be converted
// before the IN expression is converted.
switch (kind) {
case IN:
case NOT_IN:
case SOME:
case ALL:
switch (logic) {
case TRUE_FALSE_UNKNOWN:
RelDataType type = validator().getValidatedNodeTypeIfKnown(node);
if (type == null) {
// The node might not be validated if we still don't know type of the
// node.
// Therefore return directly.
return;
} else {
break;
}
case UNKNOWN_AS_FALSE:
logic = RelOptUtil.Logic.TRUE;
break;
default:
break;
}
bb.registerSubQuery(node, logic);
break;
default:
break;
}
}
/**
* Converts an expression from {@link SqlNode} to {@link RexNode} format.
*
* @param node Expression to translate
* @return Converted expression
*/
public RexNode convertExpression(SqlNode node) {
Map nameToTypeMap = Collections.emptyMap();
final ParameterScope scope =
new ParameterScope((SqlValidatorImpl) validator(), nameToTypeMap);
final Blackboard bb = createBlackboard(scope, null, false);
replaceSubQueries(bb, node, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
return bb.convertExpression(node);
}
/**
* Converts an expression from {@link SqlNode} to {@link RexNode} format, mapping identifier
* references to predefined expressions.
*
* @param node Expression to translate
* @param nameToNodeMap map from String to {@link RexNode}; when an {@link SqlIdentifier} is
* encountered, it is used as a key and translated to the corresponding value from this map
* @return Converted expression
*/
public RexNode convertExpression(SqlNode node, Map nameToNodeMap) {
final Map nameToTypeMap = new HashMap<>();
for (Map.Entry entry : nameToNodeMap.entrySet()) {
nameToTypeMap.put(entry.getKey(), entry.getValue().getType());
}
final ParameterScope scope =
new ParameterScope((SqlValidatorImpl) validator(), nameToTypeMap);
final Blackboard bb = createBlackboard(scope, nameToNodeMap, false);
replaceSubQueries(bb, node, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
return bb.convertExpression(node);
}
/**
* Converts a non-standard expression.
*
* This method is an extension-point that derived classes can override. If this method
* returns a null result, the normal expression translation process will proceed. The default
* implementation always returns null.
*
* @param node Expression
* @param bb Blackboard
* @return null to proceed with the usual expression translation process
*/
protected @Nullable RexNode convertExtendedExpression(SqlNode node, Blackboard bb) {
return null;
}
private RexNode convertOver(Blackboard bb, SqlNode node) {
SqlCall call = (SqlCall) node;
SqlCall aggCall = call.operand(0);
boolean ignoreNulls = false;
switch (aggCall.getKind()) {
case IGNORE_NULLS:
ignoreNulls = true;
// fall through
case RESPECT_NULLS:
aggCall = aggCall.operand(0);
break;
default:
break;
}
SqlNode windowOrRef = call.operand(1);
final SqlWindow window = validator().resolveWindow(windowOrRef, bb.scope());
SqlNode sqlLowerBound = window.getLowerBound();
SqlNode sqlUpperBound = window.getUpperBound();
boolean rows = window.isRows();
SqlNodeList orderList = window.getOrderList();
if (!aggCall.getOperator().allowsFraming()) {
// If the operator does not allow framing, bracketing is implicitly
// everything up to the current row.
sqlLowerBound = SqlWindow.createUnboundedPreceding(SqlParserPos.ZERO);
sqlUpperBound = SqlWindow.createCurrentRow(SqlParserPos.ZERO);
if (aggCall.getKind() == SqlKind.ROW_NUMBER) {
// ROW_NUMBER() expects specific kind of framing.
rows = true;
}
} else if (orderList.size() == 0) {
// Without ORDER BY, there must be no bracketing.
sqlLowerBound = SqlWindow.createUnboundedPreceding(SqlParserPos.ZERO);
sqlUpperBound = SqlWindow.createUnboundedFollowing(SqlParserPos.ZERO);
} else if (sqlLowerBound == null && sqlUpperBound == null) {
sqlLowerBound = SqlWindow.createUnboundedPreceding(SqlParserPos.ZERO);
sqlUpperBound = SqlWindow.createCurrentRow(SqlParserPos.ZERO);
} else if (sqlUpperBound == null) {
sqlUpperBound = SqlWindow.createCurrentRow(SqlParserPos.ZERO);
} else if (sqlLowerBound == null) {
sqlLowerBound = SqlWindow.createCurrentRow(SqlParserPos.ZERO);
}
final SqlNodeList partitionList = window.getPartitionList();
final ImmutableList.Builder partitionKeys = ImmutableList.builder();
for (SqlNode partition : partitionList) {
validator().deriveType(bb.scope(), partition);
partitionKeys.add(bb.convertExpression(partition));
}
final RexNode lowerBound =
bb.convertExpression(requireNonNull(sqlLowerBound, "sqlLowerBound"));
final RexNode upperBound =
bb.convertExpression(requireNonNull(sqlUpperBound, "sqlUpperBound"));
if (orderList.size() == 0 && !rows) {
// A logical range requires an ORDER BY clause. Use the implicit
// ordering of this relation. There must be one, otherwise it would
// have failed validation.
orderList = bb.scope().getOrderList();
if (orderList == null) {
throw new AssertionError("Relation should have sort key for implicit ORDER BY");
}
}
final ImmutableList.Builder orderKeys = ImmutableList.builder();
for (SqlNode order : orderList) {
orderKeys.add(
bb.convertSortExpression(
order,
RelFieldCollation.Direction.ASCENDING,
RelFieldCollation.NullDirection.UNSPECIFIED,
bb::sortToRex));
}
try {
Preconditions.checkArgument(bb.window == null, "already in window agg mode");
bb.window = window;
RexNode rexAgg = exprConverter.convertCall(bb, aggCall);
rexAgg = rexBuilder.ensureType(validator().getValidatedNodeType(call), rexAgg, false);
// Walk over the tree and apply 'over' to all agg functions. This is
// necessary because the returned expression is not necessarily a call
// to an agg function. For example, AVG(x) becomes SUM(x) / COUNT(x).
final SqlLiteral q = aggCall.getFunctionQuantifier();
final boolean isDistinct = q != null && q.getValue() == SqlSelectKeyword.DISTINCT;
final RexShuttle visitor =
new HistogramShuttle(
partitionKeys.build(),
orderKeys.build(),
rows,
RexWindowBounds.create(sqlLowerBound, lowerBound),
RexWindowBounds.create(sqlUpperBound, upperBound),
window.isAllowPartial(),
isDistinct,
ignoreNulls);
return rexAgg.accept(visitor);
} finally {
bb.window = null;
}
}
protected void convertFrom(Blackboard bb, @Nullable SqlNode from) {
convertFrom(bb, from, Collections.emptyList());
}
// ----- FLINK MODIFICATION BEGIN -----
private boolean containsQueryHints = false;
/**
* To tell this converter that this SqlNode tree contains query hints and then a query block
* alias will be attached to the root node of the query block.
*
* The `containsQueryHints` is false default to be compatible with previous behavior and then
* planner can reuse some node.
*
*
TODO At present, it is a relatively hacked way
*/
public void containsQueryHints() {
containsQueryHints = true;
}
// ----- FLINK MODIFICATION END -----
/**
* Converts a FROM clause into a relational expression.
*
* @param bb Scope within which to resolve identifiers
* @param from FROM clause of a query. Examples include:
*
* - a single table ("SALES.EMP"),
*
- an aliased table ("EMP AS E"),
*
- a list of tables ("EMP, DEPT"),
*
- an ANSI Join expression ("EMP JOIN DEPT ON EMP.DEPTNO = DEPT.DEPTNO"),
*
- a VALUES clause ("VALUES ('Fred', 20)"),
*
- a query ("(SELECT * FROM EMP WHERE GENDER = 'F')"),
*
- or any combination of the above.
*
*
* @param fieldNames Field aliases, usually come from AS clause, or null
*/
protected void convertFrom(
Blackboard bb, @Nullable SqlNode from, @Nullable List fieldNames) {
if (from == null) {
bb.setRoot(LogicalValues.createOneRow(cluster), false);
return;
}
final SqlCall call;
switch (from.getKind()) {
case AS:
call = (SqlCall) from;
SqlNode firstOperand = call.operand(0);
final List fieldNameList =
call.operandCount() > 2
? SqlIdentifier.simpleNames(Util.skip(call.getOperandList(), 2))
: null;
convertFrom(bb, firstOperand, fieldNameList);
// ----- FLINK MODIFICATION BEGIN -----
// Add a query-block alias hint to distinguish different query levels
// Due to Calcite will expand the whole SQL Rel Node tree that contains query block,
// but sometimes the query block should be perceived such as query hint propagation.
// TODO add query-block alias hint in SqlNode instead of here
if (containsQueryHints) {
RelNode root = bb.root;
if (root instanceof Hintable) {
RelHint queryBlockAliasHint =
RelHint.builder(FlinkHints.HINT_ALIAS)
.hintOption(call.operand(1).toString())
.build();
RelNode newRoot =
((Hintable) root)
.attachHints(
Collections.singletonList(queryBlockAliasHint));
boolean isLeaf = leaves.containsKey(root);
if (isLeaf) {
// remove old root node
leaves.remove(root);
}
bb.setRoot(newRoot, isLeaf);
}
}
// ----- FLINK MODIFICATION END -----
return;
case MATCH_RECOGNIZE:
convertMatchRecognize(bb, (SqlMatchRecognize) from);
return;
case PIVOT:
convertPivot(bb, (SqlPivot) from);
return;
case UNPIVOT:
convertUnpivot(bb, (SqlUnpivot) from);
return;
case WITH_ITEM:
convertFrom(bb, ((SqlWithItem) from).query);
return;
case WITH:
convertFrom(bb, ((SqlWith) from).body);
return;
case TABLESAMPLE:
final List operands = ((SqlCall) from).getOperandList();
SqlSampleSpec sampleSpec =
SqlLiteral.sampleValue(
requireNonNull(operands.get(1), () -> "operand[1] of " + from));
if (sampleSpec instanceof SqlSampleSpec.SqlSubstitutionSampleSpec) {
String sampleName =
((SqlSampleSpec.SqlSubstitutionSampleSpec) sampleSpec).getName();
datasetStack.push(sampleName);
convertFrom(bb, operands.get(0));
datasetStack.pop();
} else if (sampleSpec instanceof SqlSampleSpec.SqlTableSampleSpec) {
SqlSampleSpec.SqlTableSampleSpec tableSampleSpec =
(SqlSampleSpec.SqlTableSampleSpec) sampleSpec;
convertFrom(bb, operands.get(0));
RelOptSamplingParameters params =
new RelOptSamplingParameters(
tableSampleSpec.isBernoulli(),
tableSampleSpec.getSamplePercentage(),
tableSampleSpec.isRepeatable(),
tableSampleSpec.getRepeatableSeed());
bb.setRoot(new Sample(cluster, bb.root(), params), false);
} else {
throw new AssertionError("unknown TABLESAMPLE type: " + sampleSpec);
}
return;
case TABLE_REF:
call = (SqlCall) from;
convertIdentifier(bb, call.operand(0), null, call.operand(1), null);
return;
case IDENTIFIER:
convertIdentifier(bb, (SqlIdentifier) from, null, null, null);
return;
case EXTEND:
call = (SqlCall) from;
final SqlNode operand0 = call.getOperandList().get(0);
final SqlIdentifier id =
operand0.getKind() == SqlKind.TABLE_REF
? ((SqlCall) operand0).operand(0)
: (SqlIdentifier) operand0;
SqlNodeList extendedColumns = (SqlNodeList) call.getOperandList().get(1);
convertIdentifier(bb, id, extendedColumns, null, null);
return;
case SNAPSHOT:
convertTemporalTable(bb, (SqlCall) from);
return;
case JOIN:
convertJoin(bb, (SqlJoin) from);
return;
case SELECT:
case INTERSECT:
case EXCEPT:
case UNION:
final RelNode rel = convertQueryRecursive(from, false, null).project();
bb.setRoot(rel, true);
return;
case VALUES:
convertValuesImpl(bb, (SqlCall) from, null);
if (fieldNames != null) {
bb.setRoot(relBuilder.push(bb.root()).rename(fieldNames).build(), true);
}
return;
case UNNEST:
convertUnnest(bb, (SqlCall) from, fieldNames);
return;
case COLLECTION_TABLE:
call = (SqlCall) from;
// Dig out real call; TABLE() wrapper is just syntactic.
assert call.getOperandList().size() == 1;
final SqlCall call2 = call.operand(0);
convertCollectionTable(bb, call2);
return;
default:
throw new AssertionError("not a join operator " + from);
}
}
private void convertUnnest(Blackboard bb, SqlCall call, @Nullable List fieldNames) {
final List nodes = call.getOperandList();
final SqlUnnestOperator operator = (SqlUnnestOperator) call.getOperator();
for (SqlNode node : nodes) {
replaceSubQueries(bb, node, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
}
final List exprs = new ArrayList<>();
for (Ord node : Ord.zip(nodes)) {
exprs.add(
relBuilder.alias(
bb.convertExpression(node.e), validator().deriveAlias(node.e, node.i)));
}
RelNode child = (null != bb.root) ? bb.root : LogicalValues.createOneRow(cluster);
RelNode uncollect;
if (validator().config().conformance().allowAliasUnnestItems()) {
uncollect =
relBuilder
.push(child)
.project(exprs)
.uncollect(
requireNonNull(fieldNames, "fieldNames"),
operator.withOrdinality)
.build();
} else {
// REVIEW danny 2020-04-26: should we unify the normal field aliases and
// the item aliases?
uncollect =
relBuilder
.push(child)
.project(exprs)
.uncollect(Collections.emptyList(), operator.withOrdinality)
.let(r -> fieldNames == null ? r : r.rename(fieldNames))
.build();
}
bb.setRoot(uncollect, true);
}
protected void convertMatchRecognize(Blackboard bb, SqlMatchRecognize matchRecognize) {
final SqlValidatorNamespace ns = getNamespace(matchRecognize);
final SqlValidatorScope scope = validator().getMatchRecognizeScope(matchRecognize);
final Blackboard matchBb = createBlackboard(scope, null, false);
final RelDataType rowType = ns.getRowType();
// convert inner query, could be a table name or a derived table
SqlNode expr = matchRecognize.getTableRef();
convertFrom(matchBb, expr);
final RelNode input = matchBb.root();
// PARTITION BY
final SqlNodeList partitionList = matchRecognize.getPartitionList();
final ImmutableBitSet partitionKeys = buildPartitionKeys(matchBb, partitionList);
// ORDER BY
// TODO combine with buildCollation method after support NULLS_FIRST/NULLS_LAST
final SqlNodeList orderList = matchRecognize.getOrderList();
final List orderKeys = new ArrayList<>();
for (SqlNode order : orderList) {
final RelFieldCollation.Direction direction;
switch (order.getKind()) {
case DESCENDING:
direction = RelFieldCollation.Direction.DESCENDING;
order = ((SqlCall) order).operand(0);
break;
case NULLS_FIRST:
case NULLS_LAST:
throw new AssertionError();
default:
direction = RelFieldCollation.Direction.ASCENDING;
break;
}
final RelFieldCollation.NullDirection nullDirection =
validator().config().defaultNullCollation().last(desc(direction))
? RelFieldCollation.NullDirection.LAST
: RelFieldCollation.NullDirection.FIRST;
RexNode e = matchBb.convertExpression(order);
orderKeys.add(
new RelFieldCollation(((RexInputRef) e).getIndex(), direction, nullDirection));
}
final RelCollation orders = cluster.traitSet().canonize(RelCollations.of(orderKeys));
// convert pattern
final Set patternVarsSet = new HashSet<>();
SqlNode pattern = matchRecognize.getPattern();
final SqlBasicVisitor<@Nullable RexNode> patternVarVisitor =
new SqlBasicVisitor<@Nullable RexNode>() {
@Override
public RexNode visit(SqlCall call) {
List operands = call.getOperandList();
List newOperands = new ArrayList<>();
for (SqlNode node : operands) {
RexNode arg = requireNonNull(node.accept(this), node::toString);
newOperands.add(arg);
}
return rexBuilder.makeCall(
validator().getUnknownType(), call.getOperator(), newOperands);
}
@Override
public RexNode visit(SqlIdentifier id) {
assert id.isSimple();
patternVarsSet.add(id.getSimple());
return rexBuilder.makeLiteral(id.getSimple());
}
@Override
public RexNode visit(SqlLiteral literal) {
if (literal instanceof SqlNumericLiteral) {
return rexBuilder.makeExactLiteral(
BigDecimal.valueOf(literal.intValue(true)));
} else {
return rexBuilder.makeLiteral(literal.booleanValue());
}
}
};
final RexNode patternNode = pattern.accept(patternVarVisitor);
assert patternNode != null : "pattern is not found in " + pattern;
SqlLiteral interval = matchRecognize.getInterval();
RexNode intervalNode = null;
if (interval != null) {
intervalNode = matchBb.convertLiteral(interval);
}
// convert subset
final SqlNodeList subsets = matchRecognize.getSubsetList();
final Map> subsetMap = new HashMap<>();
for (SqlNode node : subsets) {
List operands = ((SqlCall) node).getOperandList();
SqlIdentifier left = (SqlIdentifier) operands.get(0);
patternVarsSet.add(left.getSimple());
final SqlNodeList rights = (SqlNodeList) operands.get(1);
final TreeSet list = new TreeSet<>(SqlIdentifier.simpleNames(rights));
subsetMap.put(left.getSimple(), list);
}
SqlNode afterMatch = matchRecognize.getAfter();
if (afterMatch == null) {
afterMatch = SqlMatchRecognize.AfterOption.SKIP_TO_NEXT_ROW.symbol(SqlParserPos.ZERO);
}
final RexNode after;
if (afterMatch instanceof SqlCall) {
List operands = ((SqlCall) afterMatch).getOperandList();
SqlOperator operator = ((SqlCall) afterMatch).getOperator();
assert operands.size() == 1;
SqlIdentifier id = (SqlIdentifier) operands.get(0);
assert patternVarsSet.contains(id.getSimple())
: id.getSimple() + " not defined in pattern";
RexNode rex = rexBuilder.makeLiteral(id.getSimple());
after =
rexBuilder.makeCall(
validator().getUnknownType(), operator, ImmutableList.of(rex));
} else {
after = matchBb.convertExpression(afterMatch);
}
matchBb.setPatternVarRef(true);
// convert measures
final ImmutableMap.Builder measureNodes = ImmutableMap.builder();
for (SqlNode measure : matchRecognize.getMeasureList()) {
List operands = ((SqlCall) measure).getOperandList();
String alias = ((SqlIdentifier) operands.get(1)).getSimple();
RexNode rex = matchBb.convertExpression(operands.get(0));
measureNodes.put(alias, rex);
}
// convert definitions
final ImmutableMap.Builder definitionNodes = ImmutableMap.builder();
for (SqlNode def : matchRecognize.getPatternDefList()) {
replaceSubQueries(matchBb, def, RelOptUtil.Logic.UNKNOWN_AS_FALSE);
List operands = ((SqlCall) def).getOperandList();
String alias = ((SqlIdentifier) operands.get(1)).getSimple();
RexNode rex = matchBb.convertExpression(operands.get(0));
definitionNodes.put(alias, rex);
}
final SqlLiteral rowsPerMatch = matchRecognize.getRowsPerMatch();
final boolean allRows =
rowsPerMatch != null
&& rowsPerMatch.getValue() == SqlMatchRecognize.RowsPerMatchOption.ALL_ROWS;
matchBb.setPatternVarRef(false);
final RelFactories.MatchFactory factory = RelFactories.DEFAULT_MATCH_FACTORY;
final RelNode rel =
factory.createMatch(
input,
patternNode,
rowType,
matchRecognize.getStrictStart().booleanValue(),
matchRecognize.getStrictEnd().booleanValue(),
definitionNodes.build(),
measureNodes.build(),
after,
subsetMap,
allRows,
partitionKeys,
orders,
intervalNode);
bb.setRoot(rel, false);
}
protected void convertPivot(Blackboard bb, SqlPivot pivot) {
final SqlValidatorScope scope = validator().getJoinScope(pivot);
final Blackboard pivotBb = createBlackboard(scope, null, false);
// Convert input
convertFrom(pivotBb, pivot.query);
final RelNode input = pivotBb.root();
final RelDataType inputRowType = input.getRowType();
relBuilder.push(input);
// Gather fields.
final AggConverter aggConverter = new AggConverter(pivotBb, (AggregatingSelectScope) null);
final Set usedColumnNames = pivot.usedColumnNames();
// 1. Gather group keys.
inputRowType.getFieldList().stream()
.filter(field -> !usedColumnNames.contains(field.getName()))
.forEach(
field ->
aggConverter.addGroupExpr(
new SqlIdentifier(field.getName(), SqlParserPos.ZERO)));
// 2. Gather axes.
pivot.axisList.forEach(aggConverter::addGroupExpr);
// 3. Gather columns used as arguments to aggregate functions.
pivotBb.agg = aggConverter;
final List<@Nullable String> aggAliasList = new ArrayList<>();
assert aggConverter.aggCalls.size() == 0;
pivot.forEachAgg(
(alias, call) -> {
call.accept(aggConverter);
aggAliasList.add(alias);
assert aggConverter.aggCalls.size() == aggAliasList.size();
});
pivotBb.agg = null;
// Project the fields that we will need.
relBuilder.project(
Pair.left(aggConverter.getPreExprs()), Pair.right(aggConverter.getPreExprs()));
// Build expressions.
// 1. Build group key
final RelBuilder.GroupKey groupKey =
relBuilder.groupKey(
inputRowType.getFieldList().stream()
.filter(field -> !usedColumnNames.contains(field.getName()))
.map(
field ->
aggConverter.addGroupExpr(
new SqlIdentifier(
field.getName(),
SqlParserPos.ZERO)))
.collect(ImmutableBitSet.toImmutableBitSet()));
// 2. Build axes, for example
// FOR (axis1, axis2 ...) IN ...
final List axes = new ArrayList<>();
for (SqlNode axis : pivot.axisList) {
axes.add(relBuilder.field(aggConverter.addGroupExpr(axis)));
}
// 3. Build aggregate expressions, for example
// PIVOT (sum(a) AS alias1, min(b) AS alias2, ... FOR ... IN ...)
final List aggCalls = new ArrayList<>();
Pair.forEach(
aggAliasList,
aggConverter.aggCalls,
(alias, aggregateCall) ->
aggCalls.add(relBuilder.aggregateCall(aggregateCall).as(alias)));
// 4. Build values, for example
// IN ((v11, v12, ...) AS label1, (v21, v22, ...) AS label2, ...)
final ImmutableList.Builder>> valueList =
ImmutableList.builder();
pivot.forEachNameValues(
(alias, nodeList) ->
valueList.add(
Pair.of(
alias,
nodeList.stream()
.map(bb::convertExpression)
.collect(Util.toImmutableList()))));
final RelNode rel = relBuilder.pivot(groupKey, aggCalls, axes, valueList.build()).build();
bb.setRoot(rel, true);
}
protected void convertUnpivot(Blackboard bb, SqlUnpivot unpivot) {
final SqlValidatorScope scope = validator().getJoinScope(unpivot);
final Blackboard unpivotBb = createBlackboard(scope, null, false);
// Convert input
convertFrom(unpivotBb, unpivot.query);
final RelNode input = unpivotBb.root();
relBuilder.push(input);
final List measureNames =
unpivot.measureList.stream()
.map(node -> ((SqlIdentifier) node).getSimple())
.collect(Util.toImmutableList());
final List axisNames =
unpivot.axisList.stream()
.map(node -> ((SqlIdentifier) node).getSimple())
.collect(Util.toImmutableList());
final ImmutableList.Builder, List>> axisMap =
ImmutableList.builder();
unpivot.forEachNameValues(
(nodeList, valueList) -> {
if (valueList == null) {
valueList =
new SqlNodeList(
Collections.nCopies(
axisNames.size(),
SqlLiteral.createCharString(
SqlUnpivot.aliasValue(nodeList),
SqlParserPos.ZERO)),
SqlParserPos.ZERO);
}
final List literals = new ArrayList<>();
Pair.forEach(
valueList,
unpivot.axisList,
(value, axis) -> {
final RelDataType type = validator().getValidatedNodeType(axis);
literals.add(convertLiteral((SqlLiteral) value, bb, type));
});
final List nodes =
nodeList.stream()
.map(unpivotBb::convertExpression)
.collect(Util.toImmutableList());
axisMap.add(Pair.of(literals, nodes));
});
relBuilder.unpivot(unpivot.includeNulls, measureNames, axisNames, axisMap.build());
relBuilder.convert(getNamespace(unpivot).getRowType(), false);
bb.setRoot(relBuilder.build(), true);
}
private void convertIdentifier(
Blackboard bb,
SqlIdentifier id,
@Nullable SqlNodeList extendedColumns,
@Nullable SqlNodeList tableHints,
@Nullable SchemaVersion schemaVersion) {
final SqlValidatorNamespace fromNamespace = getNamespace(id).resolve();
if (fromNamespace.getNode() != null) {
convertFrom(bb, fromNamespace.getNode());
return;
}
final String datasetName = datasetStack.isEmpty() ? null : datasetStack.peek();
final boolean[] usedDataset = {false};
// ----- FLINK MODIFICATION BEGIN -----
RelOptTable table =
SqlValidatorUtil.getRelOptTable(
fromNamespace,
schemaVersion == null
? catalogReader
: new FlinkCalciteCatalogSnapshotReader(
catalogReader,
catalogReader.getTypeFactory(),
schemaVersion),
datasetName,
usedDataset);
// ----- FLINK MODIFICATION END -----
assert table != null : "getRelOptTable returned null for " + fromNamespace;
if (extendedColumns != null && extendedColumns.size() > 0) {
final SqlValidatorTable validatorTable = table.unwrapOrThrow(SqlValidatorTable.class);
final List extendedFields =
SqlValidatorUtil.getExtendedColumns(validator, validatorTable, extendedColumns);
table = table.extend(extendedFields);
}
// Review Danny 2020-01-13: hacky to construct a new table scan
// in order to apply the hint strategies.
final List hints =
hintStrategies.apply(
SqlUtil.getRelHint(hintStrategies, tableHints),
LogicalTableScan.create(cluster, table, ImmutableList.of()));
final RelNode tableRel = toRel(table, hints);
bb.setRoot(tableRel, true);
if (RelOptUtil.isPureOrder(castNonNull(bb.root)) && removeSortInSubQuery(bb.top)) {
bb.setRoot(castNonNull(bb.root).getInput(0), true);
}
if (usedDataset[0]) {
bb.setDataset(datasetName);
}
}
protected void convertCollectionTable(Blackboard bb, SqlCall call) {
final SqlOperator operator = call.getOperator();
if (operator == SqlStdOperatorTable.TABLESAMPLE) {
final String sampleName = SqlLiteral.unchain(call.operand(0)).getValueAs(String.class);
datasetStack.push(sampleName);
SqlCall cursorCall = call.operand(1);
SqlNode query = cursorCall.operand(0);
RelNode converted = convertQuery(query, false, false).rel;
bb.setRoot(converted, false);
datasetStack.pop();
return;
}
replaceSubQueries(bb, call, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
// Expand table macro if possible. It's more efficient than
// LogicalTableFunctionScan.
final SqlCallBinding callBinding =
new SqlCallBinding(bb.scope().getValidator(), bb.scope, call);
if (operator instanceof SqlUserDefinedTableMacro) {
final SqlUserDefinedTableMacro udf = (SqlUserDefinedTableMacro) operator;
final TranslatableTable table = udf.getTable(callBinding);
final RelDataType rowType = table.getRowType(typeFactory);
CalciteSchema schema =
Schemas.subSchema(
catalogReader.getRootSchema(), udf.getNameAsId().skipLast(1).names);
TableExpressionFactory expressionFunction =
clazz ->
Schemas.getTableExpression(
Objects.requireNonNull(schema, "schema").plus(),
Util.last(udf.getNameAsId().names),
table,
clazz);
RelOptTable relOptTable =
RelOptTableImpl.create(
null, rowType, udf.getNameAsId().names, table, expressionFunction);
RelNode converted = toRel(relOptTable, ImmutableList.of());
bb.setRoot(converted, true);
return;
}
Type elementType;
if (operator instanceof SqlUserDefinedTableFunction) {
SqlUserDefinedTableFunction udtf = (SqlUserDefinedTableFunction) operator;
elementType = udtf.getElementType(callBinding);
} else {
elementType = null;
}
RexNode rexCall = bb.convertExpression(call);
final List inputs = bb.retrieveCursors();
Set columnMappings = getColumnMappings(operator);
LogicalTableFunctionScan callRel =
LogicalTableFunctionScan.create(
cluster,
inputs,
rexCall,
elementType,
validator().getValidatedNodeType(call),
columnMappings);
bb.setRoot(callRel, true);
afterTableFunction(bb, call, callRel);
}
protected void afterTableFunction(
SqlToRelConverter.Blackboard bb, SqlCall call, LogicalTableFunctionScan callRel) {}
private void convertTemporalTable(Blackboard bb, SqlCall call) {
final SqlSnapshot snapshot = (SqlSnapshot) call;
final RexNode period = bb.convertExpression(snapshot.getPeriod());
// convert inner query, could be a table name or a derived table
SqlNode expr = snapshot.getTableRef();
// ----- FLINK MODIFICATION BEGIN -----
SqlNode tableRef = snapshot.getTableRef();
// since we have reduced the period of SqlSnapshot in the validate phase, we only need to
// check whether the period is a RexLiteral.
// in most cases, tableRef is a SqlBasicCall and the first operand is a SqlIdentifier.
// when using SQL Hints, tableRef will be a SqlTableRef.
if (((tableRef instanceof SqlBasicCall
&& ((SqlBasicCall) tableRef).operand(0) instanceof SqlIdentifier)
|| (tableRef instanceof SqlTableRef))
&& period instanceof RexLiteral) {
TableConfig tableConfig = ShortcutUtils.unwrapContext(relBuilder).getTableConfig();
ZoneId zoneId = tableConfig.getLocalTimeZone();
TimestampString timestampString =
((RexLiteral) period).getValueAs(TimestampString.class);
checkNotNull(
timestampString,
"The time travel expression %s can not convert to a valid timestamp string. This is a bug. Please file an issue.",
period);
long timeTravelTimestamp =
TimestampData.fromEpochMillis(timestampString.getMillisSinceEpoch())
.toLocalDateTime()
.atZone(zoneId)
.toInstant()
.toEpochMilli();
SqlIdentifier sqlIdentifier =
tableRef instanceof SqlBasicCall
? ((SqlBasicCall) tableRef).operand(0)
: ((SqlTableRef) tableRef).operand(0);
SchemaVersion schemaVersion = TimestampSchemaVersion.of(timeTravelTimestamp);
convertIdentifier(bb, sqlIdentifier, null, null, schemaVersion);
} else {
convertFrom(bb, expr);
}
// ----- FLINK MODIFICATION END -----
final RelNode snapshotRel = relBuilder.push(bb.root()).snapshot(period).build();
bb.setRoot(snapshotRel, false);
}
private static @Nullable Set getColumnMappings(SqlOperator op) {
SqlReturnTypeInference rti = op.getReturnTypeInference();
if (rti == null) {
return null;
}
if (rti instanceof TableFunctionReturnTypeInference) {
TableFunctionReturnTypeInference tfrti = (TableFunctionReturnTypeInference) rti;
return tfrti.getColumnMappings();
} else {
return null;
}
}
/**
* Shuttle that replace outer {@link RexInputRef} with {@link RexFieldAccess}, and adjust {@code
* offset} to each inner {@link RexInputRef} in the lateral join condition.
*/
private static class RexAccessShuttle extends RexShuttle {
private final RexBuilder builder;
private final RexCorrelVariable rexCorrel;
private final BitSet varCols = new BitSet();
RexAccessShuttle(RexBuilder builder, RexCorrelVariable rexCorrel) {
this.builder = builder;
this.rexCorrel = rexCorrel;
}
@Override
public RexNode visitInputRef(RexInputRef input) {
int i = input.getIndex() - rexCorrel.getType().getFieldCount();
if (i < 0) {
varCols.set(input.getIndex());
return builder.makeFieldAccess(rexCorrel, input.getIndex());
}
return builder.makeInputRef(input.getType(), i);
}
}
protected RelNode createJoin(
Blackboard bb,
RelNode leftRel,
RelNode rightRel,
RexNode joinCond,
JoinRelType joinType) {
assert joinCond != null;
final CorrelationUse p = getCorrelationUse(bb, rightRel);
if (p != null) {
RelNode innerRel = p.r;
ImmutableBitSet requiredCols = p.requiredColumns;
if (!joinCond.isAlwaysTrue()) {
final RelFactories.FilterFactory factory = RelFactories.DEFAULT_FILTER_FACTORY;
final RexCorrelVariable rexCorrel =
(RexCorrelVariable) rexBuilder.makeCorrel(leftRel.getRowType(), p.id);
final RexAccessShuttle shuttle = new RexAccessShuttle(rexBuilder, rexCorrel);
// Replace outer RexInputRef with RexFieldAccess,
// and push lateral join predicate into inner child
final RexNode newCond = joinCond.accept(shuttle);
innerRel = factory.createFilter(p.r, newCond, ImmutableSet.of());
requiredCols = ImmutableBitSet.fromBitSet(shuttle.varCols).union(p.requiredColumns);
}
return LogicalCorrelate.create(
leftRel, innerRel, ImmutableList.of(), p.id, requiredCols, joinType);
}
final RelNode node =
relBuilder.push(leftRel).push(rightRel).join(joinType, joinCond).build();
// If join conditions are pushed down, update the leaves.
if (node instanceof Project) {
final Join newJoin = (Join) node.getInputs().get(0);
if (leaves.containsKey(leftRel)) {
leaves.put(newJoin.getLeft(), leaves.get(leftRel));
}
if (leaves.containsKey(rightRel)) {
leaves.put(newJoin.getRight(), leaves.get(rightRel));
}
}
return node;
}
private @Nullable CorrelationUse getCorrelationUse(Blackboard bb, final RelNode r0) {
final Set correlatedVariables = RelOptUtil.getVariablesUsed(r0);
if (correlatedVariables.isEmpty()) {
return null;
}
final ImmutableBitSet.Builder requiredColumns = ImmutableBitSet.builder();
final List correlNames = new ArrayList<>();
// All correlations must refer the same namespace since correlation
// produces exactly one correlation source.
// The same source might be referenced by different variables since
// DeferredLookups are not de-duplicated at create time.
SqlValidatorNamespace prevNs = null;
for (CorrelationId correlName : correlatedVariables) {
DeferredLookup lookup =
requireNonNull(
mapCorrelToDeferred.get(correlName),
() -> "correlation variable is not found: " + correlName);
RexFieldAccess fieldAccess = lookup.getFieldAccess(correlName);
String originalRelName = lookup.getOriginalRelName();
String originalFieldName = fieldAccess.getField().getName();
final SqlNameMatcher nameMatcher = bb.getValidator().getCatalogReader().nameMatcher();
final SqlValidatorScope.ResolvedImpl resolved = new SqlValidatorScope.ResolvedImpl();
lookup.bb
.scope()
.resolve(ImmutableList.of(originalRelName), nameMatcher, false, resolved);
assert resolved.count() == 1;
final SqlValidatorScope.Resolve resolve = resolved.only();
final SqlValidatorNamespace foundNs = resolve.namespace;
final RelDataType rowType = resolve.rowType();
final int childNamespaceIndex = resolve.path.steps().get(0).i;
final SqlValidatorScope ancestorScope = resolve.scope;
boolean correlInCurrentScope = bb.scope().isWithin(ancestorScope);
if (!correlInCurrentScope) {
continue;
}
if (prevNs == null) {
prevNs = foundNs;
} else {
assert prevNs == foundNs
: "All correlation variables should resolve"
+ " to the same namespace."
+ " Prev ns="
+ prevNs
+ ", new ns="
+ foundNs;
}
int namespaceOffset = 0;
if (childNamespaceIndex > 0) {
// If not the first child, need to figure out the width
// of output types from all the preceding namespaces
assert ancestorScope instanceof ListScope;
List children = ((ListScope) ancestorScope).getChildren();
for (int i = 0; i < childNamespaceIndex; i++) {
SqlValidatorNamespace child = children.get(i);
namespaceOffset += child.getRowType().getFieldCount();
}
}
RexFieldAccess topLevelFieldAccess = fieldAccess;
while (topLevelFieldAccess.getReferenceExpr() instanceof RexFieldAccess) {
topLevelFieldAccess = (RexFieldAccess) topLevelFieldAccess.getReferenceExpr();
}
final RelDataTypeField field =
rowType.getFieldList()
.get(topLevelFieldAccess.getField().getIndex() - namespaceOffset);
int pos = namespaceOffset + field.getIndex();
assert field.getType() == topLevelFieldAccess.getField().getType();
assert pos != -1;
// bb.root is an aggregate and only projects group by
// keys.
Map exprProjection = bb.mapRootRelToFieldProjection.get(bb.root);
if (exprProjection != null) {
// sub-query can reference group by keys projected from
// the root of the outer relation.
Integer projection = exprProjection.get(pos);
if (projection != null) {
pos = projection;
} else {
// correl not grouped
throw new AssertionError(
"Identifier '"
+ originalRelName
+ "."
+ originalFieldName
+ "' is not a group expr");
}
}
requiredColumns.set(pos);
correlNames.add(correlName);
}
if (correlNames.isEmpty()) {
// None of the correlating variables originated in this scope.
return null;
}
RelNode r = r0;
if (correlNames.size() > 1) {
// The same table was referenced more than once.
// So we deduplicate.
r =
DeduplicateCorrelateVariables.go(
rexBuilder, correlNames.get(0), Util.skip(correlNames), r0);
// Add new node to leaves.
leaves.put(r, r.getRowType().getFieldCount());
}
return new CorrelationUse(correlNames.get(0), requiredColumns.build(), r);
}
/**
* Determines whether a sub-query is non-correlated. Note that a non-correlated sub-query can
* contain correlated references, provided those references do not reference select statements
* that are parents of the sub-query.
*
* @param subq the sub-query
* @param bb blackboard used while converting the sub-query, i.e., the blackboard of the parent
* query of this sub-query
* @return true if the sub-query is non-correlated
*/
private boolean isSubQueryNonCorrelated(RelNode subq, Blackboard bb) {
Set correlatedVariables = RelOptUtil.getVariablesUsed(subq);
for (CorrelationId correlName : correlatedVariables) {
DeferredLookup lookup =
requireNonNull(
mapCorrelToDeferred.get(correlName),
() -> "correlation variable is not found: " + correlName);
String originalRelName = lookup.getOriginalRelName();
final SqlNameMatcher nameMatcher =
lookup.bb.scope().getValidator().getCatalogReader().nameMatcher();
final SqlValidatorScope.ResolvedImpl resolved = new SqlValidatorScope.ResolvedImpl();
lookup.bb
.scope()
.resolve(ImmutableList.of(originalRelName), nameMatcher, false, resolved);
SqlValidatorScope ancestorScope = resolved.only().scope;
// If the correlated reference is in a scope that's "above" the
// sub-query, then this is a correlated sub-query.
SqlValidatorScope parentScope = bb.scope;
do {
if (ancestorScope == parentScope) {
return false;
}
if (parentScope instanceof DelegatingScope) {
parentScope = ((DelegatingScope) parentScope).getParent();
} else {
break;
}
} while (parentScope != null);
}
return true;
}
/**
* Returns a list of fields to be prefixed to each relational expression.
*
* @return List of system fields
*/
protected List getSystemFields() {
return Collections.emptyList();
}
private void convertJoin(Blackboard bb, SqlJoin join) {
SqlValidator validator = validator();
final SqlValidatorScope scope = validator.getJoinScope(join);
final Blackboard fromBlackboard = createBlackboard(scope, null, false);
SqlNode left = join.getLeft();
SqlNode right = join.getRight();
final SqlValidatorScope leftScope =
Util.first(
validator.getJoinScope(left), ((DelegatingScope) bb.scope()).getParent());
final Blackboard leftBlackboard = createBlackboard(leftScope, null, false);
final SqlValidatorScope rightScope =
Util.first(
validator.getJoinScope(right), ((DelegatingScope) bb.scope()).getParent());
final Blackboard rightBlackboard = createBlackboard(rightScope, null, false);
convertFrom(leftBlackboard, left);
final RelNode leftRel = requireNonNull(leftBlackboard.root, "leftBlackboard.root");
convertFrom(rightBlackboard, right);
final RelNode tempRightRel = requireNonNull(rightBlackboard.root, "rightBlackboard.root");
final JoinConditionType conditionType = join.getConditionType();
final RexNode condition;
final RelNode rightRel;
if (join.isNatural()) {
condition = convertNaturalCondition(getNamespace(left), getNamespace(right));
rightRel = tempRightRel;
} else {
switch (conditionType) {
case NONE:
condition = rexBuilder.makeLiteral(true);
rightRel = tempRightRel;
break;
case USING:
condition =
convertUsingCondition(join, getNamespace(left), getNamespace(right));
rightRel = tempRightRel;
break;
case ON:
Pair conditionAndRightNode =
convertOnCondition(fromBlackboard, join, leftRel, tempRightRel);
condition = conditionAndRightNode.left;
rightRel = conditionAndRightNode.right;
break;
default:
throw Util.unexpected(conditionType);
}
}
final RelNode joinRel =
createJoin(
fromBlackboard,
leftRel,
rightRel,
condition,
convertJoinType(join.getJoinType()));
relBuilder.push(joinRel);
final RelNode newProjectRel = relBuilder.project(relBuilder.fields()).build();
bb.setRoot(newProjectRel, false);
}
private RexNode convertNaturalCondition(
SqlValidatorNamespace leftNamespace, SqlValidatorNamespace rightNamespace) {
final List columnList =
SqlValidatorUtil.deriveNaturalJoinColumnList(
catalogReader.nameMatcher(),
leftNamespace.getRowType(),
rightNamespace.getRowType());
return convertUsing(leftNamespace, rightNamespace, columnList);
}
private RexNode convertUsingCondition(
SqlJoin join,
SqlValidatorNamespace leftNamespace,
SqlValidatorNamespace rightNamespace) {
final SqlNodeList list =
(SqlNodeList)
requireNonNull(join.getCondition(), () -> "getCondition for join " + join);
return convertUsing(
leftNamespace,
rightNamespace,
ImmutableList.copyOf(SqlIdentifier.simpleNames(list)));
}
/**
* This currently does not expand correlated full outer joins correctly. Replaying on the right
* side to correctly support left joins multiplicities.
*
*
*
*
* SELECT *
* FROM t1
* LEFT JOIN t2 ON
* EXIST(SELECT t3.c3 WHERE t1.c1 = t3.c1 AND t2.c2 = t3.c2)
* AND NOT (t2.t2 = 2)
*
*
*
*
* Given the de-correlated query produces:
*
*
*
*
* t1.c1 | t2.c2
* ------+------
* 1 | 1
* 1 | 2
*
*
*
*
* If correlated query was replayed on the left side, then an extra rows would be emitted for
* every {code t1.c1 = 1}, where it failed to join to right side due to {code NOT(t2.t2 = 2)}.
* However, if the query is joined on the right, side multiplicity is maintained.
*/
private Pair convertOnCondition(
Blackboard bb, SqlJoin join, RelNode leftRel, RelNode rightRel) {
SqlNode condition =
requireNonNull(join.getCondition(), () -> "getCondition for join " + join);
bb.setRoot(ImmutableList.of(leftRel, rightRel));
replaceSubQueries(bb, condition, RelOptUtil.Logic.UNKNOWN_AS_FALSE);
final RelNode newRightRel =
bb.root == null || bb.registered.size() == 0 ? rightRel : bb.reRegister(rightRel);
bb.setRoot(ImmutableList.of(leftRel, newRightRel));
RexNode conditionExp = bb.convertExpression(condition);
if (conditionExp instanceof RexInputRef && newRightRel != rightRel) {
int leftFieldCount = leftRel.getRowType().getFieldCount();
List rightFieldList = newRightRel.getRowType().getFieldList();
int rightFieldCount = newRightRel.getRowType().getFieldCount();
conditionExp =
rexBuilder.makeInputRef(
rightFieldList.get(rightFieldCount - 1).getType(),
leftFieldCount + rightFieldCount - 1);
}
return Pair.of(conditionExp, newRightRel);
}
/**
* Returns an expression for matching columns of a USING clause or inferred from NATURAL JOIN.
* "a JOIN b USING (x, y)" becomes "a.x = b.x AND a.y = b.y". Returns null if the column list is
* empty.
*
* @param leftNamespace Namespace of left input to join
* @param rightNamespace Namespace of right input to join
* @param nameList List of column names to join on
* @return Expression to match columns from name list, or true if name list is empty
*/
private RexNode convertUsing(
SqlValidatorNamespace leftNamespace,
SqlValidatorNamespace rightNamespace,
List nameList) {
final SqlNameMatcher nameMatcher = catalogReader.nameMatcher();
final List list = new ArrayList<>();
for (String name : nameList) {
List operands = new ArrayList<>();
int offset = 0;
for (SqlValidatorNamespace n : ImmutableList.of(leftNamespace, rightNamespace)) {
final RelDataType rowType = n.getRowType();
final RelDataTypeField field = nameMatcher.field(rowType, name);
assert field != null
: "field " + name + " is not found in " + rowType + " with " + nameMatcher;
operands.add(rexBuilder.makeInputRef(field.getType(), offset + field.getIndex()));
offset += rowType.getFieldList().size();
}
list.add(rexBuilder.makeCall(SqlStdOperatorTable.EQUALS, operands));
}
return RexUtil.composeConjunction(rexBuilder, list);
}
private static JoinRelType convertJoinType(JoinType joinType) {
switch (joinType) {
case COMMA:
case INNER:
case CROSS:
return JoinRelType.INNER;
case FULL:
return JoinRelType.FULL;
case LEFT:
return JoinRelType.LEFT;
case RIGHT:
return JoinRelType.RIGHT;
default:
throw Util.unexpected(joinType);
}
}
/**
* Converts the SELECT, GROUP BY and HAVING clauses of an aggregate query.
*
* This method extracts SELECT, GROUP BY and HAVING clauses, and creates an {@link
* AggConverter}, then delegates to {@link #createAggImpl}. Derived class may override this
* method to change any of those clauses or specify a different {@link AggConverter}.
*
* @param bb Scope within which to resolve identifiers
* @param select Query
* @param orderExprList Additional expressions needed to implement ORDER BY
*/
protected void convertAgg(Blackboard bb, SqlSelect select, List orderExprList) {
assert bb.root != null : "precondition: child != null";
SqlNodeList groupList = select.getGroup();
SqlNodeList selectList = select.getSelectList();
SqlNode having = select.getHaving();
final AggConverter aggConverter = new AggConverter(bb, select);
createAggImpl(bb, aggConverter, selectList, groupList, having, orderExprList);
}
protected final void createAggImpl(
Blackboard bb,
final AggConverter aggConverter,
SqlNodeList selectList,
@Nullable SqlNodeList groupList,
@Nullable SqlNode having,
List orderExprList) {
// Find aggregate functions in SELECT and HAVING clause
final AggregateFinder aggregateFinder = new AggregateFinder();
selectList.accept(aggregateFinder);
if (having != null) {
having.accept(aggregateFinder);
}
// first replace the sub-queries inside the aggregates
// because they will provide input rows to the aggregates.
replaceSubQueries(bb, aggregateFinder.list, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
// also replace sub-queries inside filters in the aggregates
replaceSubQueries(bb, aggregateFinder.filterList, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
// also replace sub-queries inside ordering spec in the aggregates
replaceSubQueries(bb, aggregateFinder.orderList, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
// If group-by clause is missing, pretend that it has zero elements.
if (groupList == null) {
groupList = SqlNodeList.EMPTY;
}
replaceSubQueries(bb, groupList, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
// register the group exprs
// build a map to remember the projections from the top scope to the
// output of the current root.
//
// Calcite allows expressions, not just column references in
// group by list. This is not SQL 2003 compliant, but hey.
final AggregatingSelectScope scope =
requireNonNull(aggConverter.aggregatingSelectScope, "aggregatingSelectScope");
final AggregatingSelectScope.Resolved r = scope.resolved.get();
for (SqlNode groupExpr : r.groupExprList) {
aggConverter.addGroupExpr(groupExpr);
}
final RexNode havingExpr;
final List> projects = new ArrayList<>();
try {
Preconditions.checkArgument(bb.agg == null, "already in agg mode");
bb.agg = aggConverter;
// convert the select and having expressions, so that the
// agg converter knows which aggregations are required
selectList.accept(aggConverter);
// Assert we don't have dangling items left in the stack
assert !aggConverter.inOver;
for (SqlNode expr : orderExprList) {
expr.accept(aggConverter);
assert !aggConverter.inOver;
}
if (having != null) {
having.accept(aggConverter);
assert !aggConverter.inOver;
}
// compute inputs to the aggregator
List> preExprs = aggConverter.getPreExprs();
if (preExprs.size() == 0) {
// Special case for COUNT(*), where we can end up with no inputs
// at all. The rest of the system doesn't like 0-tuples, so we
// select a dummy constant here.
final RexNode zero = rexBuilder.makeExactLiteral(BigDecimal.ZERO);
preExprs = ImmutableList.of(Pair.of(zero, null));
}
final RelNode inputRel = bb.root();
// Project the expressions required by agg and having.
bb.setRoot(
relBuilder
.push(inputRel)
.projectNamed(Pair.left(preExprs), Pair.right(preExprs), false)
.build(),
false);
bb.mapRootRelToFieldProjection.put(bb.root(), r.groupExprProjection);
// REVIEW jvs 31-Oct-2007: doesn't the declaration of
// monotonicity here assume sort-based aggregation at
// the physical level?
// Tell bb which of group columns are sorted.
bb.columnMonotonicities.clear();
for (SqlNode groupItem : groupList) {
bb.columnMonotonicities.add(bb.scope().getMonotonicity(groupItem));
}
// Add the aggregator
bb.setRoot(
createAggregate(
bb, r.groupSet, r.groupSets.asList(), aggConverter.getAggCalls()),
false);
bb.mapRootRelToFieldProjection.put(bb.root(), r.groupExprProjection);
// Replace sub-queries in having here and modify having to use
// the replaced expressions
if (having != null) {
SqlNode newHaving = pushDownNotForIn(bb.scope(), having);
replaceSubQueries(bb, newHaving, RelOptUtil.Logic.UNKNOWN_AS_FALSE);
havingExpr = bb.convertExpression(newHaving);
} else {
havingExpr = relBuilder.literal(true);
}
// Now convert the other sub-queries in the select list.
// This needs to be done separately from the sub-query inside
// any aggregate in the select list, and after the aggregate rel
// is allocated.
replaceSubQueries(bb, selectList, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
// Now sub-queries in the entire select list have been converted.
// Convert the select expressions to get the final list to be
// projected.
int k = 0;
// For select expressions, use the field names previously assigned
// by the validator. If we derive afresh, we might generate names
// like "EXPR$2" that don't match the names generated by the
// validator. This is especially the case when there are system
// fields; system fields appear in the relnode's rowtype but do not
// (yet) appear in the validator type.
final SelectScope selectScope = SqlValidatorUtil.getEnclosingSelectScope(bb.scope);
assert selectScope != null;
final SqlValidatorNamespace selectNamespace = getNamespaceOrNull(selectScope.getNode());
assert selectNamespace != null : "selectNamespace must not be null for " + selectScope;
final List names = selectNamespace.getRowType().getFieldNames();
int sysFieldCount = selectList.size() - names.size();
for (SqlNode expr : selectList) {
projects.add(
Pair.of(
bb.convertExpression(expr),
k < sysFieldCount
? castNonNull(validator().deriveAlias(expr, k++))
: names.get(k++ - sysFieldCount)));
}
for (SqlNode expr : orderExprList) {
projects.add(
Pair.of(
bb.convertExpression(expr),
castNonNull(validator().deriveAlias(expr, k++))));
}
} finally {
bb.agg = null;
}
// implement HAVING (we have already checked that it is non-trivial)
relBuilder.push(bb.root());
if (havingExpr != null) {
relBuilder.filter(havingExpr);
}
// implement the SELECT list
relBuilder.project(Pair.left(projects), Pair.right(projects)).rename(Pair.right(projects));
bb.setRoot(relBuilder.build(), false);
// Tell bb which of group columns are sorted.
bb.columnMonotonicities.clear();
for (SqlNode selectItem : selectList) {
bb.columnMonotonicities.add(bb.scope().getMonotonicity(selectItem));
}
}
/**
* Creates an Aggregate.
*
* In case the aggregate rel changes the order in which it projects fields, the
* groupExprProjection parameter is provided, and the implementation of this method may
* modify it.
*
*
The sortedCount parameter is the number of expressions known to be monotonic.
* These expressions must be on the leading edge of the grouping keys. The default
* implementation of this method ignores this parameter.
*
* @param bb Blackboard
* @param groupSet Bit set of ordinals of grouping columns
* @param groupSets Grouping sets
* @param aggCalls Array of calls to aggregate functions
* @return LogicalAggregate
*/
protected RelNode createAggregate(
Blackboard bb,
ImmutableBitSet groupSet,
ImmutableList groupSets,
List aggCalls) {
relBuilder.push(bb.root());
final RelBuilder.GroupKey groupKey = relBuilder.groupKey(groupSet, groupSets);
return relBuilder.aggregate(groupKey, aggCalls).build();
}
public RexDynamicParam convertDynamicParam(final SqlDynamicParam dynamicParam) {
// REVIEW jvs 8-Jan-2005: dynamic params may be encountered out of
// order. Should probably cross-check with the count from the parser
// at the end and make sure they all got filled in. Why doesn't List
// have a resize() method?!? Make this a utility.
while (dynamicParam.getIndex() >= dynamicParamSqlNodes.size()) {
dynamicParamSqlNodes.add(null);
}
dynamicParamSqlNodes.set(dynamicParam.getIndex(), dynamicParam);
return rexBuilder.makeDynamicParam(
getDynamicParamType(dynamicParam.getIndex()), dynamicParam.getIndex());
}
/**
* Creates a list of collations required to implement the ORDER BY clause, if there is one.
* Populates extraOrderExprs with any sort expressions which are not in the select
* clause.
*
* @param bb Scope within which to resolve identifiers
* @param select Select clause. Never null, because we invent a dummy SELECT if ORDER BY is
* applied to a set operation (UNION etc.)
* @param orderList Order by clause, may be null
* @param extraOrderExprs Sort expressions which are not in the select clause (output)
* @param collationList List of collations (output)
*/
protected void gatherOrderExprs(
Blackboard bb,
SqlSelect select,
@Nullable SqlNodeList orderList,
List extraOrderExprs,
List collationList) {
// TODO: add validation rules to SqlValidator also
assert bb.root != null : "precondition: child != null";
assert select != null;
if (orderList == null) {
return;
}
if (removeSortInSubQuery(bb.top)) {
SqlNode offset = select.getOffset();
if ((offset == null
|| (offset instanceof SqlLiteral
&& Objects.equals(
((SqlLiteral) offset).bigDecimalValue(),
BigDecimal.ZERO)))
&& select.getFetch() == null) {
return;
}
}
for (SqlNode orderItem : orderList) {
collationList.add(
convertOrderItem(
select,
orderItem,
extraOrderExprs,
RelFieldCollation.Direction.ASCENDING,
RelFieldCollation.NullDirection.UNSPECIFIED));
}
}
protected RelFieldCollation convertOrderItem(
SqlSelect select,
SqlNode orderItem,
List extraExprs,
RelFieldCollation.Direction direction,
RelFieldCollation.NullDirection nullDirection) {
assert select != null;
// Handle DESC keyword, e.g. 'select a, b from t order by a desc'.
switch (orderItem.getKind()) {
case DESCENDING:
return convertOrderItem(
select,
((SqlCall) orderItem).operand(0),
extraExprs,
RelFieldCollation.Direction.DESCENDING,
nullDirection);
case NULLS_FIRST:
return convertOrderItem(
select,
((SqlCall) orderItem).operand(0),
extraExprs,
direction,
RelFieldCollation.NullDirection.FIRST);
case NULLS_LAST:
return convertOrderItem(
select,
((SqlCall) orderItem).operand(0),
extraExprs,
direction,
RelFieldCollation.NullDirection.LAST);
default:
break;
}
SqlNode converted = validator().expandOrderExpr(select, orderItem);
switch (nullDirection) {
case UNSPECIFIED:
nullDirection =
validator().config().defaultNullCollation().last(desc(direction))
? RelFieldCollation.NullDirection.LAST
: RelFieldCollation.NullDirection.FIRST;
break;
default:
break;
}
// Scan the select list and order exprs for an identical expression.
final SelectScope selectScope =
requireNonNull(
validator().getRawSelectScope(select),
() -> "getRawSelectScope is not found for " + select);
int ordinal = -1;
List expandedSelectList = selectScope.getExpandedSelectList();
for (SqlNode selectItem : requireNonNull(expandedSelectList, "expandedSelectList")) {
++ordinal;
if (converted.equalsDeep(stripAs(selectItem), Litmus.IGNORE)) {
return new RelFieldCollation(ordinal, direction, nullDirection);
}
}
for (SqlNode extraExpr : extraExprs) {
++ordinal;
if (converted.equalsDeep(extraExpr, Litmus.IGNORE)) {
return new RelFieldCollation(ordinal, direction, nullDirection);
}
}
// TODO: handle collation sequence
// TODO: flag expressions as non-standard
extraExprs.add(converted);
return new RelFieldCollation(ordinal + 1, direction, nullDirection);
}
private static boolean desc(RelFieldCollation.Direction direction) {
switch (direction) {
case DESCENDING:
case STRICTLY_DESCENDING:
return true;
default:
return false;
}
}
@Deprecated // to be removed before 2.0
protected boolean enableDecorrelation() {
// disable sub-query decorrelation when needed.
// e.g. if outer joins are not supported.
return config.isDecorrelationEnabled();
}
protected RelNode decorrelateQuery(RelNode rootRel) {
return RelDecorrelator.decorrelateQuery(rootRel, relBuilder);
}
/**
* Returns whether to trim unused fields as part of the conversion process.
*
* @return Whether to trim unused fields
*/
@Deprecated // to be removed before 2.0
public boolean isTrimUnusedFields() {
return config.isTrimUnusedFields();
}
/**
* Recursively converts a query to a relational expression.
*
* @param query Query
* @param top Whether this query is the top-level query of the statement
* @param targetRowType Target row type, or null
* @return Relational expression
*/
protected RelRoot convertQueryRecursive(
SqlNode query, boolean top, @Nullable RelDataType targetRowType) {
final SqlKind kind = query.getKind();
switch (kind) {
case SELECT:
return RelRoot.of(convertSelect((SqlSelect) query, top), kind);
case INSERT:
return RelRoot.of(convertInsert((SqlInsert) query), kind);
case DELETE:
return RelRoot.of(convertDelete((SqlDelete) query), kind);
case UPDATE:
return RelRoot.of(convertUpdate((SqlUpdate) query), kind);
case MERGE:
return RelRoot.of(convertMerge((SqlMerge) query), kind);
case UNION:
case INTERSECT:
case EXCEPT:
return RelRoot.of(convertSetOp((SqlCall) query), kind);
case WITH:
return convertWith((SqlWith) query, top);
case VALUES:
return RelRoot.of(convertValues((SqlCall) query, targetRowType), kind);
default:
throw new AssertionError("not a query: " + query);
}
}
/**
* Converts a set operation (UNION, INTERSECT, MINUS) into relational expressions.
*
* @param call Call to set operator
* @return Relational expression
*/
protected RelNode convertSetOp(SqlCall call) {
final RelNode left = convertQueryRecursive(call.operand(0), false, null).project();
final RelNode right = convertQueryRecursive(call.operand(1), false, null).project();
switch (call.getKind()) {
case UNION:
return LogicalUnion.create(ImmutableList.of(left, right), all(call));
case INTERSECT:
return LogicalIntersect.create(ImmutableList.of(left, right), all(call));
case EXCEPT:
return LogicalMinus.create(ImmutableList.of(left, right), all(call));
default:
throw Util.unexpected(call.getKind());
}
}
private static boolean all(SqlCall call) {
return ((SqlSetOperator) call.getOperator()).isAll();
}
protected RelNode convertInsert(SqlInsert call) {
RelOptTable targetTable = getTargetTable(call);
final RelDataType targetRowType = validator().getValidatedNodeType(call);
assert targetRowType != null;
RelNode sourceRel = convertQueryRecursive(call.getSource(), true, targetRowType).project();
RelNode massagedRel = convertColumnList(call, sourceRel);
return createModify(targetTable, massagedRel);
}
/** Creates a relational expression to modify a table or modifiable view. */
private RelNode createModify(RelOptTable targetTable, RelNode source) {
final ModifiableTable modifiableTable = targetTable.unwrap(ModifiableTable.class);
if (modifiableTable != null && modifiableTable == targetTable.unwrap(Table.class)) {
return modifiableTable.toModificationRel(
cluster,
targetTable,
catalogReader,
source,
LogicalTableModify.Operation.INSERT,
null,
null,
false);
}
final ModifiableView modifiableView = targetTable.unwrap(ModifiableView.class);
if (modifiableView != null) {
final Table delegateTable = modifiableView.getTable();
final RelDataType delegateRowType = delegateTable.getRowType(typeFactory);
final RelOptTable delegateRelOptTable =
RelOptTableImpl.create(
null, delegateRowType, delegateTable, modifiableView.getTablePath());
final RelNode newSource =
createSource(targetTable, source, modifiableView, delegateRowType);
return createModify(delegateRelOptTable, newSource);
}
return LogicalTableModify.create(
targetTable,
catalogReader,
source,
LogicalTableModify.Operation.INSERT,
null,
null,
false);
}
/**
* Wraps a relational expression in the projects and filters implied by a {@link
* ModifiableView}.
*
* The input relational expression is suitable for inserting into the view, and the returned
* relational expression is suitable for inserting into its delegate table.
*
*
In principle, the delegate table of a view might be another modifiable view, and if so,
* the process can be repeated.
*/
private RelNode createSource(
RelOptTable targetTable,
RelNode source,
ModifiableView modifiableView,
RelDataType delegateRowType) {
final ImmutableIntList mapping = modifiableView.getColumnMapping();
assert mapping.size() == targetTable.getRowType().getFieldCount();
// For columns represented in the mapping, the expression is just a field
// reference.
final Map projectMap = new HashMap<>();
final List filters = new ArrayList<>();
for (int i = 0; i < mapping.size(); i++) {
int target = mapping.get(i);
if (target >= 0) {
projectMap.put(target, RexInputRef.of(i, source.getRowType()));
}
}
// For columns that are not in the mapping, and have a constraint of the
// form "column = value", the expression is the literal "value".
//
// If a column has multiple constraints, the extra ones will become a
// filter.
final RexNode constraint = modifiableView.getConstraint(rexBuilder, delegateRowType);
RelOptUtil.inferViewPredicates(projectMap, filters, constraint);
final List> projects = new ArrayList<>();
for (RelDataTypeField field : delegateRowType.getFieldList()) {
RexNode node = projectMap.get(field.getIndex());
if (node == null) {
node = rexBuilder.makeNullLiteral(field.getType());
}
projects.add(
Pair.of(rexBuilder.ensureType(field.getType(), node, false), field.getName()));
}
return relBuilder
.push(source)
.projectNamed(Pair.left(projects), Pair.right(projects), false)
.filter(filters)
.build();
}
private RelOptTable.ToRelContext createToRelContext(List hints) {
return ViewExpanders.toRelContext(viewExpander, cluster, hints);
}
public RelNode toRel(final RelOptTable table, final List hints) {
final RelNode scan = table.toRel(createToRelContext(hints));
final InitializerExpressionFactory ief =
table.maybeUnwrap(InitializerExpressionFactory.class)
.orElse(NullInitializerExpressionFactory.INSTANCE);
boolean hasVirtualFields =
table.getRowType().getFieldList().stream()
.anyMatch(
f ->
ief.generationStrategy(table, f.getIndex())
== ColumnStrategy.VIRTUAL);
if (hasVirtualFields) {
final RexNode sourceRef = rexBuilder.makeRangeReference(scan);
final Blackboard bb =
createInsertBlackboard(table, sourceRef, table.getRowType().getFieldNames());
final List list = new ArrayList<>();
for (RelDataTypeField f : table.getRowType().getFieldList()) {
final ColumnStrategy strategy = ief.generationStrategy(table, f.getIndex());
switch (strategy) {
case VIRTUAL:
list.add(ief.newColumnDefaultValue(table, f.getIndex(), bb));
break;
default:
list.add(
rexBuilder.makeInputRef(
scan, RelOptTableImpl.realOrdinal(table, f.getIndex())));
}
}
relBuilder.push(scan);
relBuilder.project(list);
final RelNode project = relBuilder.build();
BiFunction postConversionHook =
ief.postExpressionConversionHook();
if (postConversionHook != null) {
return postConversionHook.apply(bb, project);
} else {
return project;
}
}
return scan;
}
protected RelOptTable getTargetTable(SqlNode call) {
final SqlValidatorNamespace targetNs = getNamespace(call);
SqlValidatorNamespace namespace;
if (targetNs.isWrapperFor(SqlValidatorImpl.DmlNamespace.class)) {
namespace = targetNs.unwrap(SqlValidatorImpl.DmlNamespace.class);
} else {
namespace = targetNs.resolve();
}
RelOptTable table = SqlValidatorUtil.getRelOptTable(namespace, catalogReader, null, null);
return requireNonNull(table, "no table found for " + call);
}
/**
* Creates a source for an INSERT statement.
*
* If the column list is not specified, source expressions match target columns in order.
*
*
If the column list is specified, Source expressions are mapped to target columns by name
* via targetColumnList, and may not cover the entire target table. So, we'll make up a full
* row, using a combination of default values and the source expressions provided.
*
* @param call Insert expression
* @param source Source relational expression
* @return Converted INSERT statement
*/
protected RelNode convertColumnList(final SqlInsert call, RelNode source) {
RelDataType sourceRowType = source.getRowType();
final RexNode sourceRef = rexBuilder.makeRangeReference(sourceRowType, 0, false);
final List targetColumnNames = new ArrayList<>();
final List columnExprs = new ArrayList<>();
collectInsertTargets(call, sourceRef, targetColumnNames, columnExprs);
final RelOptTable targetTable = getTargetTable(call);
final RelDataType targetRowType = RelOptTableImpl.realRowType(targetTable);
final List targetFields = targetRowType.getFieldList();
final List<@Nullable RexNode> sourceExps =
new ArrayList<>(Collections.nCopies(targetFields.size(), null));
final List<@Nullable String> fieldNames =
new ArrayList<>(Collections.nCopies(targetFields.size(), null));
final InitializerExpressionFactory initializerFactory =
getInitializerFactory(getNamespace(call).getTable());
// Walk the name list and place the associated value in the
// expression list according to the ordinal value returned from
// the table construct, leaving nulls in the list for columns
// that are not referenced.
final SqlNameMatcher nameMatcher = catalogReader.nameMatcher();
for (Pair p : Pair.zip(targetColumnNames, columnExprs)) {
RelDataTypeField field = nameMatcher.field(targetRowType, p.left);
assert field != null : "column " + p.left + " not found";
sourceExps.set(field.getIndex(), p.right);
}
// Lazily create a blackboard that contains all non-generated columns.
final Supplier bb =
() -> createInsertBlackboard(targetTable, sourceRef, targetColumnNames);
// Walk the expression list and get default values for any columns
// that were not supplied in the statement. Get field names too.
for (int i = 0; i < targetFields.size(); ++i) {
final RelDataTypeField field = targetFields.get(i);
final String fieldName = field.getName();
fieldNames.set(i, fieldName);
RexNode sourceExpression = sourceExps.get(i);
if (sourceExpression == null || sourceExpression.getKind() == SqlKind.DEFAULT) {
sourceExpression =
initializerFactory.newColumnDefaultValue(targetTable, i, bb.get());
// bare nulls are dangerous in the wrong hands
sourceExpression = castNullLiteralIfNeeded(sourceExpression, field.getType());
sourceExps.set(i, sourceExpression);
}
}
// sourceExps should not contain nulls (see the loop above)
@SuppressWarnings("assignment.type.incompatible")
List nonNullExprs = sourceExps;
return relBuilder.push(source).projectNamed(nonNullExprs, fieldNames, false).build();
}
/**
* Creates a blackboard for translating the expressions of generated columns in an INSERT
* statement.
*/
private Blackboard createInsertBlackboard(
RelOptTable targetTable, RexNode sourceRef, List targetColumnNames) {
final Map nameToNodeMap = new HashMap<>();
int j = 0;
// Assign expressions for non-generated columns.
final List strategies = targetTable.getColumnStrategies();
final List targetFields = targetTable.getRowType().getFieldNames();
for (String targetColumnName : targetColumnNames) {
final int i = targetFields.indexOf(targetColumnName);
switch (strategies.get(i)) {
case STORED:
case VIRTUAL:
break;
default:
nameToNodeMap.put(targetColumnName, rexBuilder.makeFieldAccess(sourceRef, j++));
}
}
return createBlackboard(null, nameToNodeMap, false);
}
private static InitializerExpressionFactory getInitializerFactory(
@Nullable SqlValidatorTable validatorTable) {
// We might unwrap a null instead of a InitializerExpressionFactory.
final Table table = unwrap(validatorTable, Table.class);
if (table != null) {
InitializerExpressionFactory f = unwrap(table, InitializerExpressionFactory.class);
if (f != null) {
return f;
}
}
return NullInitializerExpressionFactory.INSTANCE;
}
private static @Nullable T unwrap(@Nullable Object o, Class clazz) {
if (o instanceof Wrapper) {
return ((Wrapper) o).unwrap(clazz);
}
return null;
}
private RexNode castNullLiteralIfNeeded(RexNode node, RelDataType type) {
if (!RexLiteral.isNullLiteral(node)) {
return node;
}
return rexBuilder.makeCast(type, node);
}
/**
* Given an INSERT statement, collects the list of names to be populated and the expressions to
* put in them.
*
* @param call Insert statement
* @param sourceRef Expression representing a row from the source relational expression
* @param targetColumnNames List of target column names, to be populated
* @param columnExprs List of expressions, to be populated
*/
protected void collectInsertTargets(
SqlInsert call,
final RexNode sourceRef,
final List targetColumnNames,
List columnExprs) {
final RelOptTable targetTable = getTargetTable(call);
final RelDataType tableRowType = targetTable.getRowType();
SqlNodeList targetColumnList = call.getTargetColumnList();
if (targetColumnList == null) {
if (validator().config().conformance().isInsertSubsetColumnsAllowed()) {
final RelDataType targetRowType =
typeFactory.createStructType(
tableRowType
.getFieldList()
.subList(0, sourceRef.getType().getFieldCount()));
targetColumnNames.addAll(targetRowType.getFieldNames());
} else {
targetColumnNames.addAll(tableRowType.getFieldNames());
}
} else {
for (int i = 0; i < targetColumnList.size(); i++) {
SqlIdentifier id = (SqlIdentifier) targetColumnList.get(i);
RelDataTypeField field =
SqlValidatorUtil.getTargetField(
tableRowType, typeFactory, id, catalogReader, targetTable);
assert field != null : "column " + id.toString() + " not found";
targetColumnNames.add(field.getName());
}
}
final Blackboard bb = createInsertBlackboard(targetTable, sourceRef, targetColumnNames);
// Next, assign expressions for generated columns.
final List strategies = targetTable.getColumnStrategies();
for (String columnName : targetColumnNames) {
final int i = tableRowType.getFieldNames().indexOf(columnName);
final RexNode expr;
switch (strategies.get(i)) {
case STORED:
final InitializerExpressionFactory f =
targetTable
.maybeUnwrap(InitializerExpressionFactory.class)
.orElse(NullInitializerExpressionFactory.INSTANCE);
expr = f.newColumnDefaultValue(targetTable, i, bb);
break;
case VIRTUAL:
expr = null;
break;
default:
expr = requireNonNull(bb.nameToNodeMap, "nameToNodeMap").get(columnName);
}
// expr is nullable, however, all the nulls will be removed in the loop below
columnExprs.add(castNonNull(expr));
}
// Remove virtual columns from the list.
for (int i = 0; i < targetColumnNames.size(); i++) {
if (columnExprs.get(i) == null) {
columnExprs.remove(i);
targetColumnNames.remove(i);
--i;
}
}
}
private RelNode convertDelete(SqlDelete call) {
RelOptTable targetTable = getTargetTable(call);
RelNode sourceRel =
convertSelect(
requireNonNull(call.getSourceSelect(), () -> "sourceSelect for " + call),
false);
return LogicalTableModify.create(
targetTable,
catalogReader,
sourceRel,
LogicalTableModify.Operation.DELETE,
null,
null,
false);
}
private RelNode convertUpdate(SqlUpdate call) {
final SqlValidatorScope scope =
validator()
.getWhereScope(
requireNonNull(
call.getSourceSelect(), () -> "sourceSelect for " + call));
Blackboard bb = createBlackboard(scope, null, false);
replaceSubQueries(bb, call, RelOptUtil.Logic.TRUE_FALSE_UNKNOWN);
RelOptTable targetTable = getTargetTable(call);
// convert update column list from SqlIdentifier to String
final List targetColumnNameList = new ArrayList<>();
final RelDataType targetRowType = targetTable.getRowType();
for (SqlNode node : call.getTargetColumnList()) {
SqlIdentifier id = (SqlIdentifier) node;
RelDataTypeField field =
SqlValidatorUtil.getTargetField(
targetRowType, typeFactory, id, catalogReader, targetTable);
assert field != null : "column " + id.toString() + " not found";
targetColumnNameList.add(field.getName());
}
RelNode sourceRel =
convertSelect(
requireNonNull(call.getSourceSelect(), () -> "sourceSelect for " + call),
false);
bb.setRoot(sourceRel, false);
ImmutableList.Builder rexNodeSourceExpressionListBuilder = ImmutableList.builder();
for (SqlNode n : call.getSourceExpressionList()) {
RexNode rn = bb.convertExpression(n);
rexNodeSourceExpressionListBuilder.add(rn);
}
return LogicalTableModify.create(
targetTable,
catalogReader,
sourceRel,
LogicalTableModify.Operation.UPDATE,
targetColumnNameList,
rexNodeSourceExpressionListBuilder.build(),
false);
}
private RelNode convertMerge(SqlMerge call) {
RelOptTable targetTable = getTargetTable(call);
// convert update column list from SqlIdentifier to String
final List targetColumnNameList = new ArrayList<>();
final RelDataType targetRowType = targetTable.getRowType();
SqlUpdate updateCall = call.getUpdateCall();
if (updateCall != null) {
for (SqlNode targetColumn : updateCall.getTargetColumnList()) {
SqlIdentifier id = (SqlIdentifier) targetColumn;
RelDataTypeField field =
SqlValidatorUtil.getTargetField(
targetRowType, typeFactory, id, catalogReader, targetTable);
assert field != null : "column " + id.toString() + " not found";
targetColumnNameList.add(field.getName());
}
}
// replace the projection of the source select with a
// projection that contains the following:
// 1) the expressions corresponding to the new insert row (if there is
// an insert)
// 2) all columns from the target table (if there is an update)
// 3) the set expressions in the update call (if there is an update)
// first, convert the merge's source select to construct the columns
// from the target table and the set expressions in the update call
RelNode mergeSourceRel =
convertSelect(
requireNonNull(call.getSourceSelect(), () -> "sourceSelect for " + call),
false);
// then, convert the insert statement so we can get the insert
// values expressions
SqlInsert insertCall = call.getInsertCall();
int nLevel1Exprs = 0;
List level1InsertExprs = null;
List level2InsertExprs = null;
if (insertCall != null) {
RelNode insertRel = convertInsert(insertCall);
// if there are 2 level of projections in the insert source, combine
// them into a single project; level1 refers to the topmost project;
// the level1 projection contains references to the level2
// expressions, except in the case where no target expression was
// provided, in which case, the expression is the default value for
// the column; or if the expressions directly map to the source
// table
level1InsertExprs = ((LogicalProject) insertRel.getInput(0)).getProjects();
if (insertRel.getInput(0).getInput(0) instanceof LogicalProject) {
level2InsertExprs =
((LogicalProject) insertRel.getInput(0).getInput(0)).getProjects();
}
nLevel1Exprs = level1InsertExprs.size();
}
LogicalJoin join = (LogicalJoin) mergeSourceRel.getInput(0);
int nSourceFields = join.getLeft().getRowType().getFieldCount();
final List projects = new ArrayList<>();
for (int level1Idx = 0; level1Idx < nLevel1Exprs; level1Idx++) {
requireNonNull(level1InsertExprs, "level1InsertExprs");
if ((level2InsertExprs != null)
&& (level1InsertExprs.get(level1Idx) instanceof RexInputRef)) {
int level2Idx = ((RexInputRef) level1InsertExprs.get(level1Idx)).getIndex();
projects.add(level2InsertExprs.get(level2Idx));
} else {
projects.add(level1InsertExprs.get(level1Idx));
}
}
if (updateCall != null) {
final LogicalProject project = (LogicalProject) mergeSourceRel;
projects.addAll(Util.skip(project.getProjects(), nSourceFields));
}
relBuilder.push(join).project(projects);
return LogicalTableModify.create(
targetTable,
catalogReader,
relBuilder.build(),
LogicalTableModify.Operation.MERGE,
targetColumnNameList,
null,
false);
}
/**
* Converts an identifier into an expression in a given scope. For example, the "empno" in
* "select empno from emp join dept" becomes "emp.empno".
*/
private RexNode convertIdentifier(Blackboard bb, SqlIdentifier identifier) {
// first check for reserved identifiers like CURRENT_USER
final SqlCall call = bb.getValidator().makeNullaryCall(identifier);
if (call != null) {
return bb.convertExpression(call);
}
String pv = null;
if (bb.isPatternVarRef && identifier.names.size() > 1) {
pv = identifier.names.get(0);
}
final SqlQualified qualified;
if (bb.scope != null) {
qualified = bb.scope.fullyQualify(identifier);
} else {
qualified = SqlQualified.create(null, 1, null, identifier);
}
final Pair> e0 =
bb.lookupExp(qualified);
RexNode e = e0.left;
for (String name : qualified.suffix()) {
if (e == e0.left && e0.right != null) {
e = e0.right.apply(e, name);
} else {
final boolean caseSensitive = true; // name already fully-qualified
if (identifier.isStar() && bb.scope instanceof MatchRecognizeScope) {
e = rexBuilder.makeFieldAccess(e, 0);
} else {
e = rexBuilder.makeFieldAccess(e, name, caseSensitive);
}
}
}
if (e instanceof RexInputRef) {
// adjust the type to account for nulls introduced by outer joins
e = adjustInputRef(bb, (RexInputRef) e);
if (pv != null) {
e = RexPatternFieldRef.of(pv, (RexInputRef) e);
}
}
if (e0.left instanceof RexCorrelVariable) {
assert e instanceof RexFieldAccess;
final RexNode prev =
bb.mapCorrelateToRex.put(((RexCorrelVariable) e0.left).id, (RexFieldAccess) e);
assert prev == null;
}
return e;
}
/**
* Adjusts the type of a reference to an input field to account for nulls introduced by outer
* joins; and adjusts the offset to match the physical implementation.
*
* @param bb Blackboard
* @param inputRef Input ref
* @return Adjusted input ref
*/
protected RexNode adjustInputRef(Blackboard bb, RexInputRef inputRef) {
RelDataTypeField field = bb.getRootField(inputRef);
if (field != null) {
if (!SqlTypeUtil.equalSansNullability(
typeFactory, field.getType(), inputRef.getType())) {
return inputRef;
}
return rexBuilder.makeInputRef(field.getType(), inputRef.getIndex());
}
return inputRef;
}
/**
* Converts a row constructor into a relational expression.
*
* @param bb Blackboard
* @param rowConstructor Row constructor expression
* @return Relational expression which returns a single row.
*/
private RelNode convertRowConstructor(Blackboard bb, SqlCall rowConstructor) {
Preconditions.checkArgument(isRowConstructor(rowConstructor));
final List operands = rowConstructor.getOperandList();
return convertMultisets(operands, bb);
}
private RelNode convertCursor(Blackboard bb, SubQuery subQuery) {
final SqlCall cursorCall = (SqlCall) subQuery.node;
assert cursorCall.operandCount() == 1;
SqlNode query = cursorCall.operand(0);
RelNode converted = convertQuery(query, false, false).rel;
int iCursor = bb.cursors.size();
bb.cursors.add(converted);
subQuery.expr = new RexInputRef(iCursor, converted.getRowType());
return converted;
}
private RelNode convertMultisets(final List operands, Blackboard bb) {
// NOTE: Wael 2/04/05: this implementation is not the most efficient in
// terms of planning since it generates XOs that can be reduced.
final List