Github user hvanhovell commented on a diff in the pull request:
https://github.com/apache/spark/pull/11557#discussion_r55185301
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/ng/AstBuilder.scala
---
@@ -0,0 +1,1128 @@
+/*
+ * 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.spark.sql.catalyst.parser.ng
+
+import java.sql.{Date, Timestamp}
+
+import scala.collection.JavaConverters._
+
+import org.antlr.v4.runtime.{ParserRuleContext, Token}
+import org.antlr.v4.runtime.tree.{ParseTree, TerminalNode}
+
+import org.apache.spark.Logging
+import org.apache.spark.sql.AnalysisException
+import org.apache.spark.sql.catalyst.{InternalRow, TableIdentifier}
+import org.apache.spark.sql.catalyst.analysis._
+import org.apache.spark.sql.catalyst.expressions._
+import org.apache.spark.sql.catalyst.parser.ParseUtils
+import org.apache.spark.sql.catalyst.parser.ng.SqlBaseParser._
+import org.apache.spark.sql.catalyst.plans._
+import org.apache.spark.sql.catalyst.plans.logical._
+import org.apache.spark.sql.catalyst.trees.CurrentOrigin
+import org.apache.spark.sql.types._
+import org.apache.spark.unsafe.types.CalendarInterval
+import org.apache.spark.util.random.RandomSampler
+
+/**
+ * The AstBuilder converts an ANTLR4 ParseTree into a catalyst Expression,
LogicalPlan or
+ * TableIdentifier.
+ */
+class AstBuilder extends SqlBaseBaseVisitor[AnyRef] with Logging {
+ import AstBuilder._
+
+ protected def typedVisit[T](ctx: ParseTree): T = {
+ ctx.accept(this).asInstanceOf[T]
+ }
+
+ override def visitSingleStatement(ctx: SingleStatementContext):
LogicalPlan = withOrigin(ctx) {
+ visit(ctx.statement).asInstanceOf[LogicalPlan]
+ }
+
+ override def visitSingleExpression(ctx: SingleExpressionContext):
Expression = withOrigin(ctx) {
+ visitNamedExpression(ctx.namedExpression)
+ }
+
+ override def visitSingleTableIdentifier(
+ ctx: SingleTableIdentifierContext): TableIdentifier =
withOrigin(ctx) {
+ visitTableIdentifier(ctx.tableIdentifier)
+ }
+
+ /*
********************************************************************************************
+ * Plan parsing
+ *
********************************************************************************************
*/
+ protected def plan(tree: ParserRuleContext): LogicalPlan =
typedVisit(tree)
+
+ /**
+ * Create a plan for a SHOW FUNCTIONS command.
+ */
+ override def visitShowFunctions(ctx: ShowFunctionsContext): LogicalPlan
= withOrigin(ctx) {
+ import ctx._
+ if (qualifiedName != null) {
+ val names =
qualifiedName().identifier().asScala.map(_.getText).toList
+ names match {
+ case db :: name :: Nil =>
+ ShowFunctions(Some(db), Some(name))
+ case name :: Nil =>
+ ShowFunctions(None, Some(name))
+ case _ =>
+ notSupported("SHOW FUNCTIONS unsupported name", ctx)
+ }
+ } else if (pattern != null) {
+ ShowFunctions(None, Some(unquote(pattern.getText)))
+ } else {
+ ShowFunctions(None, None)
+ }
+ }
+
+ /**
+ * Create a plan for a DESCRIBE FUNCTION command.
+ */
+ override def visitDescribeFunction(ctx: DescribeFunctionContext):
LogicalPlan = withOrigin(ctx) {
+ val functionName =
ctx.qualifiedName().identifier().asScala.map(_.getText).mkString(".")
+ DescribeFunction(functionName, ctx.EXTENDED != null)
+ }
+
+ /**
+ * Create a top-level plan with Common Table Expressions.
+ */
+ override def visitQuery(ctx: QueryContext): LogicalPlan =
withOrigin(ctx) {
+ val query = plan(ctx.queryNoWith)
+
+ // Apply CTEs
+ query.optional(ctx.ctes) {
+ val ctes = ctx.ctes.namedQuery.asScala.map {
+ case nCtx =>
+ val namedQuery = visitNamedQuery(nCtx)
+ (namedQuery.alias, namedQuery)
+ }.toMap
+ With(query, ctes)
+ }
+ }
+
+ /**
+ * Create a named logical plan.
+ *
+ * This is only used for Common Table Expressions.
+ */
+ override def visitNamedQuery(ctx: NamedQueryContext): SubqueryAlias =
withOrigin(ctx) {
+ SubqueryAlias(ctx.name.getText, plan(ctx.queryNoWith))
+ }
+
+ /**
+ * Create a logical plan which allows for multiple inserts using one
'from' statement. These
+ * queries have the following SQL form:
+ * {{{
+ * [WITH cte...]?
+ * FROM src
+ * [INSERT INTO tbl1 SELECT *]+
+ * }}}
+ * For example:
+ * {{{
+ * FROM db.tbl1 A
+ * INSERT INTO dbo.tbl1 SELECT * WHERE A.value = 10 LIMIT 5
+ * INSERT INTO dbo.tbl2 SELECT * WHERE A.value = 12
+ * }}}
+ * This (Hive) feature cannot be combined with set-operators.
+ */
+ override def visitMultiInsertQuery(ctx: MultiInsertQueryContext):
LogicalPlan = withOrigin(ctx) {
+ val from = visitFromClause(ctx.fromClause)
+
+ // Build the insert clauses.
+ val inserts = ctx.multiInsertQueryBody.asScala.map {
+ body =>
+ assert(body.querySpecification.fromClause == null,
+ "Multi-Insert queries cannot have a FROM clause in their
individual SELECT statements",
+ body)
+
+ withQuerySpecification(body.querySpecification, from).
+ // Add organization statements.
+ optionalMap(body.queryOrganization)(withQueryOrganization).
+ // Add insert.
+ optionalMap(body.insertInto())(withInsertInto)
+ }
+
+ // If there are multiple INSERTS just UNION them together into one
query.
+ inserts match {
+ case Seq(query) => query
+ case queries => Union(queries)
+ }
+ }
+
+ /**
+ * Create a logical plan for a regular (single-insert) query.
+ */
+ override def visitSingleInsertQuery(
+ ctx: SingleInsertQueryContext): LogicalPlan = withOrigin(ctx) {
+ plan(ctx.queryTerm).
+ // Add organization statements.
+ optionalMap(ctx.queryOrganization)(withQueryOrganization).
+ // Add insert.
+ optionalMap(ctx.insertInto())(withInsertInto)
+ }
+
+ /**
+ * Add an INSERT INTO [TABLE]/INSERT OVERWRITE TABLE operation to the
logical plan.
+ */
+ private def withInsertInto(
+ ctx: InsertIntoContext,
+ query: LogicalPlan): LogicalPlan = withOrigin(ctx) {
+ val tableIdent = visitTableIdentifier(ctx.tableIdentifier)
+ val partitionKeys = Option(ctx.partitionSpec).toSeq.flatMap {
+ _.partitionVal.asScala.map {
+ pVal => (pVal.identifier.getText, Option(pVal.constant).map(c =>
unquote(c.getText)))
+ }
+ }.toMap
+
+ InsertIntoTable(
+ UnresolvedRelation(tableIdent, None),
+ partitionKeys,
+ query,
+ ctx.OVERWRITE != null,
+ ctx.EXISTS != null)
+ }
+
+ /**
+ * Add ORDER BY/SORT BY/CLUSTER BY/DISTRIBUTE BY/LIMIT/WINDOWS clauses
to the logical plan.
+ */
+ private def withQueryOrganization(
+ ctx: QueryOrganizationContext,
+ query: LogicalPlan): LogicalPlan = withOrigin(ctx) {
+ import ctx._
+
+ // Handle ORDER BY, SORT BY, DISTRIBUTE BY, and CLUSTER BY clause.
+ val withOrder = if (
+ !order.isEmpty && sort.isEmpty && distributeBy.isEmpty &&
clusterBy.isEmpty) {
+ // ORDER BY ...
+ Sort(order.asScala.map(visitSortItem), global = true, query)
+ } else if (order.isEmpty && !sort.isEmpty && distributeBy.isEmpty &&
clusterBy.isEmpty) {
+ // SORT BY ...
+ Sort(sort.asScala.map(visitSortItem), global = false, query)
+ } else if (order.isEmpty && sort.isEmpty && !distributeBy.isEmpty &&
clusterBy.isEmpty) {
+ // DISTRIBUTE BY ...
+ RepartitionByExpression(expressionList(distributeBy), query)
+ } else if (order.isEmpty && !sort.isEmpty && !distributeBy.isEmpty &&
clusterBy.isEmpty) {
+ // SORT BY ... DISTRIBUTE BY ...
+ Sort(
+ sort.asScala.map(visitSortItem),
+ global = false,
+ RepartitionByExpression(expressionList(distributeBy), query))
+ } else if (order.isEmpty && sort.isEmpty && distributeBy.isEmpty &&
!clusterBy.isEmpty) {
+ // CLUSTER BY ...
+ val expressions = expressionList(clusterBy)
+ Sort(
+ expressions.map(SortOrder(_, Ascending)),
+ global = false,
+ RepartitionByExpression(expressions, query))
+ } else if (order.isEmpty && sort.isEmpty && distributeBy.isEmpty &&
clusterBy.isEmpty) {
+ // [EMPTY]
+ query
+ } else {
+ notSupported("Combination of ORDER BY/SORT BY/DISTRIBUTE BY/CLUSTER
BY is not supported", ctx)
+ }
+
+ // LIMIT
+ val withLimit = withOrder.optional(limit) {
+ Limit(typedVisit(limit), withOrder)
+ }
+
+ // WINDOWS
+ withLimit.optionalMap(windows)(withWindows)
+ }
+
+ /**
+ * Create a logical plan using a query specification.
+ */
+ override def visitQuerySpecification(
+ ctx: QuerySpecificationContext): LogicalPlan = withOrigin(ctx) {
+ val from = OneRowRelation.optional(ctx.fromClause) {
+ visitFromClause(ctx.fromClause)
+ }
+ withQuerySpecification(ctx, from)
+ }
+
+ /**
+ * Add a query specification to a logical plan. The query specification
is the core of the logical
+ * plan, this is where sourcing (FROM clause), transforming (SELECT
TRANSFORM/MAP/REDUCE),
+ * projection (SELECT), aggregation (GROUP BY ... HAVING ...) and
filtering (WHERE) takes place.
+ *
+ * Note that query hints are ignored (both by the parser and the
builder).
+ */
+ private def withQuerySpecification(
+ ctx: QuerySpecificationContext,
+ relation: LogicalPlan): LogicalPlan = withOrigin(ctx) {
+ import ctx._
+
+ // WHERE
+ val withFilter = relation.optional(where) {
+ Filter(expression(where), relation)
+ }
+
+ // Expressions.
+ val expressions = namedExpression.asScala.map(visit).map {
+ case e: Expression => UnresolvedAlias(e)
+ }
+
+ // Create either a transform or a regular query.
+ kind.getType match {
+ case SqlBaseParser.MAP | SqlBaseParser.REDUCE |
SqlBaseParser.TRANSFORM =>
+ // Transform
+
+ // Create the attributes.
+ val attributes = if (colTypeList != null) {
+ // Typed return columns.
+ visitColTypeList(colTypeList).toAttributes
+ } else if (columnAliasList != null) {
+ // Untyped return columns.
+ visitColumnAliasList(columnAliasList).map { name =>
+ AttributeReference(name, StringType, nullable = true)()
+ }
+ } else {
+ Seq.empty
+ }
+
+ // Create the transform.
+ ScriptTransformation(
+ expressions,
+ unquote(script.getText),
+ attributes,
+ withFilter,
+ withScriptIOSchema(inRowFormat, outRowFormat, outRecordReader))
+
+ case SqlBaseParser.SELECT =>
+ // Regular select
+
+ // Add lateral views.
+ val withLateralView =
ctx.lateralView.asScala.foldLeft(withFilter)(withGenerate)
+
+ // Add aggregation with having or a project.
+ val withProject = if (aggregation != null) {
+ withAggregation(aggregation, expressions,
withLateralView).optionalMap(having) {
+ case (h, p) => Filter(expression(h), p)
--- End diff --
Note to self: Group By and Having clause should not be pipelined, a user is
allowed to define a global aggregate and filter that...
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
