diff options
author | Cheng Lian <lian.cs.zju@gmail.com> | 2014-10-09 18:25:06 -0700 |
---|---|---|
committer | Michael Armbrust <michael@databricks.com> | 2014-10-09 18:25:06 -0700 |
commit | edf02da389f75df5a42465d41f035d6b65599848 (patch) | |
tree | 1530a6ae85e804ff8ed0498e69a971bc569ee5ca | |
parent | 363baacaded56047bcc63276d729ab911e0336cf (diff) | |
download | spark-edf02da389f75df5a42465d41f035d6b65599848.tar.gz spark-edf02da389f75df5a42465d41f035d6b65599848.tar.bz2 spark-edf02da389f75df5a42465d41f035d6b65599848.zip |
[SPARK-3654][SQL] Unifies SQL and HiveQL parsers
This PR is a follow up of #2590, and tries to introduce a top level SQL parser entry point for all SQL dialects supported by Spark SQL.
A top level parser `SparkSQLParser` is introduced to handle the syntaxes that all SQL dialects should recognize (e.g. `CACHE TABLE`, `UNCACHE TABLE` and `SET`, etc.). For all the syntaxes this parser doesn't recognize directly, it fallbacks to a specified function that tries to parse arbitrary input to a `LogicalPlan`. This function is typically another parser combinator like `SqlParser`. DDL syntaxes introduced in #2475 can be moved to here.
The `ExtendedHiveQlParser` now only handle Hive specific extensions.
Also took the chance to refactor/reformat `SqlParser` for better readability.
Author: Cheng Lian <lian.cs.zju@gmail.com>
Closes #2698 from liancheng/gen-sql-parser and squashes the following commits:
ceada76 [Cheng Lian] Minor styling fixes
9738934 [Cheng Lian] Minor refactoring, removes optional trailing ";" in the parser
bb2ab12 [Cheng Lian] SET property value can be empty string
ce8860b [Cheng Lian] Passes test suites
e86968e [Cheng Lian] Removes debugging code
8bcace5 [Cheng Lian] Replaces digit.+ to rep1(digit) (Scala style checking doesn't like it)
d15d54f [Cheng Lian] Unifies SQL and HiveQL parsers
12 files changed, 414 insertions, 401 deletions
diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SparkSQLParser.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SparkSQLParser.scala new file mode 100644 index 0000000000..04467342e6 --- /dev/null +++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SparkSQLParser.scala @@ -0,0 +1,186 @@ +/* + * 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 + +import scala.language.implicitConversions +import scala.util.parsing.combinator.lexical.StdLexical +import scala.util.parsing.combinator.syntactical.StandardTokenParsers +import scala.util.parsing.combinator.{PackratParsers, RegexParsers} +import scala.util.parsing.input.CharArrayReader.EofCh + +import org.apache.spark.sql.catalyst.plans.logical._ + +private[sql] abstract class AbstractSparkSQLParser + extends StandardTokenParsers with PackratParsers { + + def apply(input: String): LogicalPlan = phrase(start)(new lexical.Scanner(input)) match { + case Success(plan, _) => plan + case failureOrError => sys.error(failureOrError.toString) + } + + protected case class Keyword(str: String) + + protected def start: Parser[LogicalPlan] + + // Returns the whole input string + protected lazy val wholeInput: Parser[String] = new Parser[String] { + def apply(in: Input): ParseResult[String] = + Success(in.source.toString, in.drop(in.source.length())) + } + + // Returns the rest of the input string that are not parsed yet + protected lazy val restInput: Parser[String] = new Parser[String] { + def apply(in: Input): ParseResult[String] = + Success( + in.source.subSequence(in.offset, in.source.length()).toString, + in.drop(in.source.length())) + } +} + +class SqlLexical(val keywords: Seq[String]) extends StdLexical { + case class FloatLit(chars: String) extends Token { + override def toString = chars + } + + reserved ++= keywords.flatMap(w => allCaseVersions(w)) + + delimiters += ( + "@", "*", "+", "-", "<", "=", "<>", "!=", "<=", ">=", ">", "/", "(", ")", + ",", ";", "%", "{", "}", ":", "[", "]", "." + ) + + override lazy val token: Parser[Token] = + ( identChar ~ (identChar | digit).* ^^ + { case first ~ rest => processIdent((first :: rest).mkString) } + | rep1(digit) ~ ('.' ~> digit.*).? ^^ { + case i ~ None => NumericLit(i.mkString) + case i ~ Some(d) => FloatLit(i.mkString + "." + d.mkString) + } + | '\'' ~> chrExcept('\'', '\n', EofCh).* <~ '\'' ^^ + { case chars => StringLit(chars mkString "") } + | '"' ~> chrExcept('"', '\n', EofCh).* <~ '"' ^^ + { case chars => StringLit(chars mkString "") } + | EofCh ^^^ EOF + | '\'' ~> failure("unclosed string literal") + | '"' ~> failure("unclosed string literal") + | delim + | failure("illegal character") + ) + + override def identChar = letter | elem('_') + + override def whitespace: Parser[Any] = + ( whitespaceChar + | '/' ~ '*' ~ comment + | '/' ~ '/' ~ chrExcept(EofCh, '\n').* + | '#' ~ chrExcept(EofCh, '\n').* + | '-' ~ '-' ~ chrExcept(EofCh, '\n').* + | '/' ~ '*' ~ failure("unclosed comment") + ).* + + /** Generate all variations of upper and lower case of a given string */ + def allCaseVersions(s: String, prefix: String = ""): Stream[String] = { + if (s == "") { + Stream(prefix) + } else { + allCaseVersions(s.tail, prefix + s.head.toLower) ++ + allCaseVersions(s.tail, prefix + s.head.toUpper) + } + } +} + +/** + * The top level Spark SQL parser. This parser recognizes syntaxes that are available for all SQL + * dialects supported by Spark SQL, and delegates all the other syntaxes to the `fallback` parser. + * + * @param fallback A function that parses an input string to a logical plan + */ +private[sql] class SparkSQLParser(fallback: String => LogicalPlan) extends AbstractSparkSQLParser { + + // A parser for the key-value part of the "SET [key = [value ]]" syntax + private object SetCommandParser extends RegexParsers { + private val key: Parser[String] = "(?m)[^=]+".r + + private val value: Parser[String] = "(?m).*$".r + + private val pair: Parser[LogicalPlan] = + (key ~ ("=".r ~> value).?).? ^^ { + case None => SetCommand(None) + case Some(k ~ v) => SetCommand(Some(k.trim -> v.map(_.trim))) + } + + def apply(input: String): LogicalPlan = parseAll(pair, input) match { + case Success(plan, _) => plan + case x => sys.error(x.toString) + } + } + + protected val AS = Keyword("AS") + protected val CACHE = Keyword("CACHE") + protected val LAZY = Keyword("LAZY") + protected val SET = Keyword("SET") + protected val TABLE = Keyword("TABLE") + protected val SOURCE = Keyword("SOURCE") + protected val UNCACHE = Keyword("UNCACHE") + + protected implicit def asParser(k: Keyword): Parser[String] = + lexical.allCaseVersions(k.str).map(x => x : Parser[String]).reduce(_ | _) + + private val reservedWords: Seq[String] = + this + .getClass + .getMethods + .filter(_.getReturnType == classOf[Keyword]) + .map(_.invoke(this).asInstanceOf[Keyword].str) + + override val lexical = new SqlLexical(reservedWords) + + override protected lazy val start: Parser[LogicalPlan] = + cache | uncache | set | shell | source | others + + private lazy val cache: Parser[LogicalPlan] = + CACHE ~> LAZY.? ~ (TABLE ~> ident) ~ (AS ~> restInput).? ^^ { + case isLazy ~ tableName ~ plan => + CacheTableCommand(tableName, plan.map(fallback), isLazy.isDefined) + } + + private lazy val uncache: Parser[LogicalPlan] = + UNCACHE ~ TABLE ~> ident ^^ { + case tableName => UncacheTableCommand(tableName) + } + + private lazy val set: Parser[LogicalPlan] = + SET ~> restInput ^^ { + case input => SetCommandParser(input) + } + + private lazy val shell: Parser[LogicalPlan] = + "!" ~> restInput ^^ { + case input => ShellCommand(input.trim) + } + + private lazy val source: Parser[LogicalPlan] = + SOURCE ~> restInput ^^ { + case input => SourceCommand(input.trim) + } + + private lazy val others: Parser[LogicalPlan] = + wholeInput ^^ { + case input => fallback(input) + } +} diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SqlParser.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SqlParser.scala index 4662f585cf..b4d606d37e 100755 --- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SqlParser.scala +++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/SqlParser.scala @@ -18,10 +18,6 @@ package org.apache.spark.sql.catalyst import scala.language.implicitConversions -import scala.util.parsing.combinator.lexical.StdLexical -import scala.util.parsing.combinator.syntactical.StandardTokenParsers -import scala.util.parsing.combinator.PackratParsers -import scala.util.parsing.input.CharArrayReader.EofCh import org.apache.spark.sql.catalyst.analysis._ import org.apache.spark.sql.catalyst.expressions._ @@ -39,31 +35,7 @@ import org.apache.spark.sql.catalyst.types._ * This is currently included mostly for illustrative purposes. Users wanting more complete support * for a SQL like language should checkout the HiveQL support in the sql/hive sub-project. */ -class SqlParser extends StandardTokenParsers with PackratParsers { - - def apply(input: String): LogicalPlan = { - // Special-case out set commands since the value fields can be - // complex to handle without RegexParsers. Also this approach - // is clearer for the several possible cases of set commands. - if (input.trim.toLowerCase.startsWith("set")) { - input.trim.drop(3).split("=", 2).map(_.trim) match { - case Array("") => // "set" - SetCommand(None, None) - case Array(key) => // "set key" - SetCommand(Some(key), None) - case Array(key, value) => // "set key=value" - SetCommand(Some(key), Some(value)) - } - } else { - phrase(query)(new lexical.Scanner(input)) match { - case Success(r, x) => r - case x => sys.error(x.toString) - } - } - } - - protected case class Keyword(str: String) - +class SqlParser extends AbstractSparkSQLParser { protected implicit def asParser(k: Keyword): Parser[String] = lexical.allCaseVersions(k.str).map(x => x : Parser[String]).reduce(_ | _) @@ -100,7 +72,6 @@ class SqlParser extends StandardTokenParsers with PackratParsers { protected val IS = Keyword("IS") protected val JOIN = Keyword("JOIN") protected val LAST = Keyword("LAST") - protected val LAZY = Keyword("LAZY") protected val LEFT = Keyword("LEFT") protected val LIKE = Keyword("LIKE") protected val LIMIT = Keyword("LIMIT") @@ -128,7 +99,6 @@ class SqlParser extends StandardTokenParsers with PackratParsers { protected val THEN = Keyword("THEN") protected val TIMESTAMP = Keyword("TIMESTAMP") protected val TRUE = Keyword("TRUE") - protected val UNCACHE = Keyword("UNCACHE") protected val UNION = Keyword("UNION") protected val UPPER = Keyword("UPPER") protected val WHEN = Keyword("WHEN") @@ -136,7 +106,8 @@ class SqlParser extends StandardTokenParsers with PackratParsers { // Use reflection to find the reserved words defined in this class. protected val reservedWords = - this.getClass + this + .getClass .getMethods .filter(_.getReturnType == classOf[Keyword]) .map(_.invoke(this).asInstanceOf[Keyword].str) @@ -150,86 +121,68 @@ class SqlParser extends StandardTokenParsers with PackratParsers { } } - protected lazy val query: Parser[LogicalPlan] = ( - select * ( - UNION ~ ALL ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Union(q1, q2) } | - INTERSECT ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Intersect(q1, q2) } | - EXCEPT ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Except(q1, q2)} | - UNION ~ opt(DISTINCT) ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Distinct(Union(q1, q2)) } + protected lazy val start: Parser[LogicalPlan] = + ( select * + ( UNION ~ ALL ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Union(q1, q2) } + | INTERSECT ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Intersect(q1, q2) } + | EXCEPT ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Except(q1, q2)} + | UNION ~ DISTINCT.? ^^^ { (q1: LogicalPlan, q2: LogicalPlan) => Distinct(Union(q1, q2)) } ) - | insert | cache | unCache - ) + | insert + ) protected lazy val select: Parser[LogicalPlan] = - SELECT ~> opt(DISTINCT) ~ projections ~ - opt(from) ~ opt(filter) ~ - opt(grouping) ~ - opt(having) ~ - opt(orderBy) ~ - opt(limit) <~ opt(";") ^^ { - case d ~ p ~ r ~ f ~ g ~ h ~ o ~ l => - val base = r.getOrElse(NoRelation) - val withFilter = f.map(f => Filter(f, base)).getOrElse(base) - val withProjection = - g.map {g => - Aggregate(g, assignAliases(p), withFilter) - }.getOrElse(Project(assignAliases(p), withFilter)) - val withDistinct = d.map(_ => Distinct(withProjection)).getOrElse(withProjection) - val withHaving = h.map(h => Filter(h, withDistinct)).getOrElse(withDistinct) - val withOrder = o.map(o => Sort(o, withHaving)).getOrElse(withHaving) - val withLimit = l.map { l => Limit(l, withOrder) }.getOrElse(withOrder) - withLimit - } + SELECT ~> DISTINCT.? ~ + repsep(projection, ",") ~ + (FROM ~> relations).? ~ + (WHERE ~> expression).? ~ + (GROUP ~ BY ~> rep1sep(expression, ",")).? ~ + (HAVING ~> expression).? ~ + (ORDER ~ BY ~> ordering).? ~ + (LIMIT ~> expression).? ^^ { + case d ~ p ~ r ~ f ~ g ~ h ~ o ~ l => + val base = r.getOrElse(NoRelation) + val withFilter = f.map(f => Filter(f, base)).getOrElse(base) + val withProjection = g + .map(Aggregate(_, assignAliases(p), withFilter)) + .getOrElse(Project(assignAliases(p), withFilter)) + val withDistinct = d.map(_ => Distinct(withProjection)).getOrElse(withProjection) + val withHaving = h.map(Filter(_, withDistinct)).getOrElse(withDistinct) + val withOrder = o.map(Sort(_, withHaving)).getOrElse(withHaving) + val withLimit = l.map(Limit(_, withOrder)).getOrElse(withOrder) + withLimit + } protected lazy val insert: Parser[LogicalPlan] = - INSERT ~> opt(OVERWRITE) ~ inTo ~ select <~ opt(";") ^^ { - case o ~ r ~ s => - val overwrite: Boolean = o.getOrElse("") == "OVERWRITE" - InsertIntoTable(r, Map[String, Option[String]](), s, overwrite) - } - - protected lazy val cache: Parser[LogicalPlan] = - CACHE ~> opt(LAZY) ~ (TABLE ~> ident) ~ opt(AS ~> select) <~ opt(";") ^^ { - case isLazy ~ tableName ~ plan => - CacheTableCommand(tableName, plan, isLazy.isDefined) - } - - protected lazy val unCache: Parser[LogicalPlan] = - UNCACHE ~ TABLE ~> ident <~ opt(";") ^^ { - case tableName => UncacheTableCommand(tableName) + INSERT ~> OVERWRITE.? ~ (INTO ~> relation) ~ select ^^ { + case o ~ r ~ s => InsertIntoTable(r, Map.empty[String, Option[String]], s, o.isDefined) } - protected lazy val projections: Parser[Seq[Expression]] = repsep(projection, ",") - protected lazy val projection: Parser[Expression] = - expression ~ (opt(AS) ~> opt(ident)) ^^ { - case e ~ None => e - case e ~ Some(a) => Alias(e, a)() + expression ~ (AS.? ~> ident.?) ^^ { + case e ~ a => a.fold(e)(Alias(e, _)()) } - protected lazy val from: Parser[LogicalPlan] = FROM ~> relations - - protected lazy val inTo: Parser[LogicalPlan] = INTO ~> relation - // Based very loosely on the MySQL Grammar. // http://dev.mysql.com/doc/refman/5.0/en/join.html protected lazy val relations: Parser[LogicalPlan] = - relation ~ "," ~ relation ^^ { case r1 ~ _ ~ r2 => Join(r1, r2, Inner, None) } | - relation + ( relation ~ ("," ~> relation) ^^ { case r1 ~ r2 => Join(r1, r2, Inner, None) } + | relation + ) protected lazy val relation: Parser[LogicalPlan] = - joinedRelation | - relationFactor + joinedRelation | relationFactor protected lazy val relationFactor: Parser[LogicalPlan] = - ident ~ (opt(AS) ~> opt(ident)) ^^ { - case tableName ~ alias => UnresolvedRelation(None, tableName, alias) - } | - "(" ~> query ~ ")" ~ opt(AS) ~ ident ^^ { case s ~ _ ~ _ ~ a => Subquery(a, s) } + ( ident ~ (opt(AS) ~> opt(ident)) ^^ { + case tableName ~ alias => UnresolvedRelation(None, tableName, alias) + } + | ("(" ~> start <~ ")") ~ (AS.? ~> ident) ^^ { case s ~ a => Subquery(a, s) } + ) protected lazy val joinedRelation: Parser[LogicalPlan] = - relationFactor ~ opt(joinType) ~ JOIN ~ relationFactor ~ opt(joinConditions) ^^ { - case r1 ~ jt ~ _ ~ r2 ~ cond => + relationFactor ~ joinType.? ~ (JOIN ~> relationFactor) ~ joinConditions.? ^^ { + case r1 ~ jt ~ r2 ~ cond => Join(r1, r2, joinType = jt.getOrElse(Inner), cond) } @@ -237,160 +190,145 @@ class SqlParser extends StandardTokenParsers with PackratParsers { ON ~> expression protected lazy val joinType: Parser[JoinType] = - INNER ^^^ Inner | - LEFT ~ SEMI ^^^ LeftSemi | - LEFT ~ opt(OUTER) ^^^ LeftOuter | - RIGHT ~ opt(OUTER) ^^^ RightOuter | - FULL ~ opt(OUTER) ^^^ FullOuter - - protected lazy val filter: Parser[Expression] = WHERE ~ expression ^^ { case _ ~ e => e } - - protected lazy val orderBy: Parser[Seq[SortOrder]] = - ORDER ~> BY ~> ordering + ( INNER ^^^ Inner + | LEFT ~ SEMI ^^^ LeftSemi + | LEFT ~ OUTER.? ^^^ LeftOuter + | RIGHT ~ OUTER.? ^^^ RightOuter + | FULL ~ OUTER.? ^^^ FullOuter + ) protected lazy val ordering: Parser[Seq[SortOrder]] = - rep1sep(singleOrder, ",") | - rep1sep(expression, ",") ~ opt(direction) ^^ { - case exps ~ None => exps.map(SortOrder(_, Ascending)) - case exps ~ Some(d) => exps.map(SortOrder(_, d)) - } + ( rep1sep(singleOrder, ",") + | rep1sep(expression, ",") ~ direction.? ^^ { + case exps ~ d => exps.map(SortOrder(_, d.getOrElse(Ascending))) + } + ) protected lazy val singleOrder: Parser[SortOrder] = - expression ~ direction ^^ { case e ~ o => SortOrder(e,o) } + expression ~ direction ^^ { case e ~ o => SortOrder(e, o) } protected lazy val direction: Parser[SortDirection] = - ASC ^^^ Ascending | - DESC ^^^ Descending - - protected lazy val grouping: Parser[Seq[Expression]] = - GROUP ~> BY ~> rep1sep(expression, ",") - - protected lazy val having: Parser[Expression] = - HAVING ~> expression - - protected lazy val limit: Parser[Expression] = - LIMIT ~> expression + ( ASC ^^^ Ascending + | DESC ^^^ Descending + ) - protected lazy val expression: Parser[Expression] = orExpression + protected lazy val expression: Parser[Expression] = + orExpression protected lazy val orExpression: Parser[Expression] = - andExpression * (OR ^^^ { (e1: Expression, e2: Expression) => Or(e1,e2) }) + andExpression * (OR ^^^ { (e1: Expression, e2: Expression) => Or(e1, e2) }) protected lazy val andExpression: Parser[Expression] = - comparisonExpression * (AND ^^^ { (e1: Expression, e2: Expression) => And(e1,e2) }) + comparisonExpression * (AND ^^^ { (e1: Expression, e2: Expression) => And(e1, e2) }) protected lazy val comparisonExpression: Parser[Expression] = - termExpression ~ "=" ~ termExpression ^^ { case e1 ~ _ ~ e2 => EqualTo(e1, e2) } | - termExpression ~ "<" ~ termExpression ^^ { case e1 ~ _ ~ e2 => LessThan(e1, e2) } | - termExpression ~ "<=" ~ termExpression ^^ { case e1 ~ _ ~ e2 => LessThanOrEqual(e1, e2) } | - termExpression ~ ">" ~ termExpression ^^ { case e1 ~ _ ~ e2 => GreaterThan(e1, e2) } | - termExpression ~ ">=" ~ termExpression ^^ { case e1 ~ _ ~ e2 => GreaterThanOrEqual(e1, e2) } | - termExpression ~ "!=" ~ termExpression ^^ { case e1 ~ _ ~ e2 => Not(EqualTo(e1, e2)) } | - termExpression ~ "<>" ~ termExpression ^^ { case e1 ~ _ ~ e2 => Not(EqualTo(e1, e2)) } | - termExpression ~ BETWEEN ~ termExpression ~ AND ~ termExpression ^^ { - case e ~ _ ~ el ~ _ ~ eu => And(GreaterThanOrEqual(e, el), LessThanOrEqual(e, eu)) - } | - termExpression ~ RLIKE ~ termExpression ^^ { case e1 ~ _ ~ e2 => RLike(e1, e2) } | - termExpression ~ REGEXP ~ termExpression ^^ { case e1 ~ _ ~ e2 => RLike(e1, e2) } | - termExpression ~ LIKE ~ termExpression ^^ { case e1 ~ _ ~ e2 => Like(e1, e2) } | - termExpression ~ IN ~ "(" ~ rep1sep(termExpression, ",") <~ ")" ^^ { - case e1 ~ _ ~ _ ~ e2 => In(e1, e2) - } | - termExpression ~ NOT ~ IN ~ "(" ~ rep1sep(termExpression, ",") <~ ")" ^^ { - case e1 ~ _ ~ _ ~ _ ~ e2 => Not(In(e1, e2)) - } | - termExpression <~ IS ~ NULL ^^ { case e => IsNull(e) } | - termExpression <~ IS ~ NOT ~ NULL ^^ { case e => IsNotNull(e) } | - NOT ~> termExpression ^^ {e => Not(e)} | - termExpression + ( termExpression ~ ("=" ~> termExpression) ^^ { case e1 ~ e2 => EqualTo(e1, e2) } + | termExpression ~ ("<" ~> termExpression) ^^ { case e1 ~ e2 => LessThan(e1, e2) } + | termExpression ~ ("<=" ~> termExpression) ^^ { case e1 ~ e2 => LessThanOrEqual(e1, e2) } + | termExpression ~ (">" ~> termExpression) ^^ { case e1 ~ e2 => GreaterThan(e1, e2) } + | termExpression ~ (">=" ~> termExpression) ^^ { case e1 ~ e2 => GreaterThanOrEqual(e1, e2) } + | termExpression ~ ("!=" ~> termExpression) ^^ { case e1 ~ e2 => Not(EqualTo(e1, e2)) } + | termExpression ~ ("<>" ~> termExpression) ^^ { case e1 ~ e2 => Not(EqualTo(e1, e2)) } + | termExpression ~ (BETWEEN ~> termExpression) ~ (AND ~> termExpression) ^^ { + case e ~ el ~ eu => And(GreaterThanOrEqual(e, el), LessThanOrEqual(e, eu)) + } + | termExpression ~ (RLIKE ~> termExpression) ^^ { case e1 ~ e2 => RLike(e1, e2) } + | termExpression ~ (REGEXP ~> termExpression) ^^ { case e1 ~ e2 => RLike(e1, e2) } + | termExpression ~ (LIKE ~> termExpression) ^^ { case e1 ~ e2 => Like(e1, e2) } + | termExpression ~ (IN ~ "(" ~> rep1sep(termExpression, ",")) <~ ")" ^^ { + case e1 ~ e2 => In(e1, e2) + } + | termExpression ~ (NOT ~ IN ~ "(" ~> rep1sep(termExpression, ",")) <~ ")" ^^ { + case e1 ~ e2 => Not(In(e1, e2)) + } + | termExpression <~ IS ~ NULL ^^ { case e => IsNull(e) } + | termExpression <~ IS ~ NOT ~ NULL ^^ { case e => IsNotNull(e) } + | NOT ~> termExpression ^^ {e => Not(e)} + | termExpression + ) protected lazy val termExpression: Parser[Expression] = - productExpression * ( - "+" ^^^ { (e1: Expression, e2: Expression) => Add(e1,e2) } | - "-" ^^^ { (e1: Expression, e2: Expression) => Subtract(e1,e2) } ) + productExpression * + ( "+" ^^^ { (e1: Expression, e2: Expression) => Add(e1, e2) } + | "-" ^^^ { (e1: Expression, e2: Expression) => Subtract(e1, e2) } + ) protected lazy val productExpression: Parser[Expression] = - baseExpression * ( - "*" ^^^ { (e1: Expression, e2: Expression) => Multiply(e1,e2) } | - "/" ^^^ { (e1: Expression, e2: Expression) => Divide(e1,e2) } | - "%" ^^^ { (e1: Expression, e2: Expression) => Remainder(e1,e2) } - ) + baseExpression * + ( "*" ^^^ { (e1: Expression, e2: Expression) => Multiply(e1, e2) } + | "/" ^^^ { (e1: Expression, e2: Expression) => Divide(e1, e2) } + | "%" ^^^ { (e1: Expression, e2: Expression) => Remainder(e1, e2) } + ) protected lazy val function: Parser[Expression] = - SUM ~> "(" ~> expression <~ ")" ^^ { case exp => Sum(exp) } | - SUM ~> "(" ~> DISTINCT ~> expression <~ ")" ^^ { case exp => SumDistinct(exp) } | - COUNT ~> "(" ~ "*" <~ ")" ^^ { case _ => Count(Literal(1)) } | - COUNT ~> "(" ~ expression <~ ")" ^^ { case dist ~ exp => Count(exp) } | - COUNT ~> "(" ~> DISTINCT ~> expression <~ ")" ^^ { case exp => CountDistinct(exp :: Nil) } | - APPROXIMATE ~> COUNT ~> "(" ~> DISTINCT ~> expression <~ ")" ^^ { - case exp => ApproxCountDistinct(exp) - } | - APPROXIMATE ~> "(" ~> floatLit ~ ")" ~ COUNT ~ "(" ~ DISTINCT ~ expression <~ ")" ^^ { - case s ~ _ ~ _ ~ _ ~ _ ~ e => ApproxCountDistinct(e, s.toDouble) - } | - FIRST ~> "(" ~> expression <~ ")" ^^ { case exp => First(exp) } | - LAST ~> "(" ~> expression <~ ")" ^^ { case exp => Last(exp) } | - AVG ~> "(" ~> expression <~ ")" ^^ { case exp => Average(exp) } | - MIN ~> "(" ~> expression <~ ")" ^^ { case exp => Min(exp) } | - MAX ~> "(" ~> expression <~ ")" ^^ { case exp => Max(exp) } | - UPPER ~> "(" ~> expression <~ ")" ^^ { case exp => Upper(exp) } | - LOWER ~> "(" ~> expression <~ ")" ^^ { case exp => Lower(exp) } | - IF ~> "(" ~> expression ~ "," ~ expression ~ "," ~ expression <~ ")" ^^ { - case c ~ "," ~ t ~ "," ~ f => If(c,t,f) - } | - CASE ~> expression.? ~ (WHEN ~> expression ~ (THEN ~> expression)).* ~ - (ELSE ~> expression).? <~ END ^^ { - case casePart ~ altPart ~ elsePart => - val altExprs = altPart.flatMap { - case we ~ te => - Seq(casePart.fold(we)(EqualTo(_, we)), te) + ( SUM ~> "(" ~> expression <~ ")" ^^ { case exp => Sum(exp) } + | SUM ~> "(" ~> DISTINCT ~> expression <~ ")" ^^ { case exp => SumDistinct(exp) } + | COUNT ~ "(" ~> "*" <~ ")" ^^ { case _ => Count(Literal(1)) } + | COUNT ~ "(" ~> expression <~ ")" ^^ { case exp => Count(exp) } + | COUNT ~> "(" ~> DISTINCT ~> expression <~ ")" ^^ { case exp => CountDistinct(exp :: Nil) } + | APPROXIMATE ~ COUNT ~ "(" ~ DISTINCT ~> expression <~ ")" ^^ + { case exp => ApproxCountDistinct(exp) } + | APPROXIMATE ~> "(" ~> floatLit ~ ")" ~ COUNT ~ "(" ~ DISTINCT ~ expression <~ ")" ^^ + { case s ~ _ ~ _ ~ _ ~ _ ~ e => ApproxCountDistinct(e, s.toDouble) } + | FIRST ~ "(" ~> expression <~ ")" ^^ { case exp => First(exp) } + | LAST ~ "(" ~> expression <~ ")" ^^ { case exp => Last(exp) } + | AVG ~ "(" ~> expression <~ ")" ^^ { case exp => Average(exp) } + | MIN ~ "(" ~> expression <~ ")" ^^ { case exp => Min(exp) } + | MAX ~ "(" ~> expression <~ ")" ^^ { case exp => Max(exp) } + | UPPER ~ "(" ~> expression <~ ")" ^^ { case exp => Upper(exp) } + | LOWER ~ "(" ~> expression <~ ")" ^^ { case exp => Lower(exp) } + | IF ~ "(" ~> expression ~ ("," ~> expression) ~ ("," ~> expression) <~ ")" ^^ + { case c ~ t ~ f => If(c, t, f) } + | CASE ~> expression.? ~ (WHEN ~> expression ~ (THEN ~> expression)).* ~ + (ELSE ~> expression).? <~ END ^^ { + case casePart ~ altPart ~ elsePart => + val altExprs = altPart.flatMap { case whenExpr ~ thenExpr => + Seq(casePart.fold(whenExpr)(EqualTo(_, whenExpr)), thenExpr) + } + CaseWhen(altExprs ++ elsePart.toList) } - CaseWhen(altExprs ++ elsePart.toList) - } | - (SUBSTR | SUBSTRING) ~> "(" ~> expression ~ "," ~ expression <~ ")" ^^ { - case s ~ "," ~ p => Substring(s,p,Literal(Integer.MAX_VALUE)) - } | - (SUBSTR | SUBSTRING) ~> "(" ~> expression ~ "," ~ expression ~ "," ~ expression <~ ")" ^^ { - case s ~ "," ~ p ~ "," ~ l => Substring(s,p,l) - } | - SQRT ~> "(" ~> expression <~ ")" ^^ { case exp => Sqrt(exp) } | - ABS ~> "(" ~> expression <~ ")" ^^ { case exp => Abs(exp) } | - ident ~ "(" ~ repsep(expression, ",") <~ ")" ^^ { - case udfName ~ _ ~ exprs => UnresolvedFunction(udfName, exprs) - } + | (SUBSTR | SUBSTRING) ~ "(" ~> expression ~ ("," ~> expression) <~ ")" ^^ + { case s ~ p => Substring(s, p, Literal(Integer.MAX_VALUE)) } + | (SUBSTR | SUBSTRING) ~ "(" ~> expression ~ ("," ~> expression) ~ ("," ~> expression) <~ ")" ^^ + { case s ~ p ~ l => Substring(s, p, l) } + | SQRT ~ "(" ~> expression <~ ")" ^^ { case exp => Sqrt(exp) } + | ABS ~ "(" ~> expression <~ ")" ^^ { case exp => Abs(exp) } + | ident ~ ("(" ~> repsep(expression, ",")) <~ ")" ^^ + { case udfName ~ exprs => UnresolvedFunction(udfName, exprs) } + ) protected lazy val cast: Parser[Expression] = - CAST ~> "(" ~> expression ~ AS ~ dataType <~ ")" ^^ { case exp ~ _ ~ t => Cast(exp, t) } + CAST ~ "(" ~> expression ~ (AS ~> dataType) <~ ")" ^^ { case exp ~ t => Cast(exp, t) } protected lazy val literal: Parser[Literal] = - numericLit ^^ { - case i if i.toLong > Int.MaxValue => Literal(i.toLong) - case i => Literal(i.toInt) - } | - NULL ^^^ Literal(null, NullType) | - floatLit ^^ {case f => Literal(f.toDouble) } | - stringLit ^^ {case s => Literal(s, StringType) } + ( numericLit ^^ { + case i if i.toLong > Int.MaxValue => Literal(i.toLong) + case i => Literal(i.toInt) + } + | NULL ^^^ Literal(null, NullType) + | floatLit ^^ {case f => Literal(f.toDouble) } + | stringLit ^^ {case s => Literal(s, StringType) } + ) protected lazy val floatLit: Parser[String] = elem("decimal", _.isInstanceOf[lexical.FloatLit]) ^^ (_.chars) protected lazy val baseExpression: PackratParser[Expression] = - expression ~ "[" ~ expression <~ "]" ^^ { - case base ~ _ ~ ordinal => GetItem(base, ordinal) - } | - (expression <~ ".") ~ ident ^^ { - case base ~ fieldName => GetField(base, fieldName) - } | - TRUE ^^^ Literal(true, BooleanType) | - FALSE ^^^ Literal(false, BooleanType) | - cast | - "(" ~> expression <~ ")" | - function | - "-" ~> literal ^^ UnaryMinus | - dotExpressionHeader | - ident ^^ UnresolvedAttribute | - "*" ^^^ Star(None) | - literal + ( expression ~ ("[" ~> expression <~ "]") ^^ + { case base ~ ordinal => GetItem(base, ordinal) } + | (expression <~ ".") ~ ident ^^ + { case base ~ fieldName => GetField(base, fieldName) } + | TRUE ^^^ Literal(true, BooleanType) + | FALSE ^^^ Literal(false, BooleanType) + | cast + | "(" ~> expression <~ ")" + | function + | "-" ~> literal ^^ UnaryMinus + | dotExpressionHeader + | ident ^^ UnresolvedAttribute + | "*" ^^^ Star(None) + | literal + ) protected lazy val dotExpressionHeader: Parser[Expression] = (ident <~ ".") ~ ident ~ rep("." ~> ident) ^^ { @@ -400,55 +338,3 @@ class SqlParser extends StandardTokenParsers with PackratParsers { protected lazy val dataType: Parser[DataType] = STRING ^^^ StringType | TIMESTAMP ^^^ TimestampType } - -class SqlLexical(val keywords: Seq[String]) extends StdLexical { - case class FloatLit(chars: String) extends Token { - override def toString = chars - } - - reserved ++= keywords.flatMap(w => allCaseVersions(w)) - - delimiters += ( - "@", "*", "+", "-", "<", "=", "<>", "!=", "<=", ">=", ">", "/", "(", ")", - ",", ";", "%", "{", "}", ":", "[", "]", "." - ) - - override lazy val token: Parser[Token] = ( - identChar ~ rep( identChar | digit ) ^^ - { case first ~ rest => processIdent(first :: rest mkString "") } - | rep1(digit) ~ opt('.' ~> rep(digit)) ^^ { - case i ~ None => NumericLit(i mkString "") - case i ~ Some(d) => FloatLit(i.mkString("") + "." + d.mkString("")) - } - | '\'' ~ rep( chrExcept('\'', '\n', EofCh) ) ~ '\'' ^^ - { case '\'' ~ chars ~ '\'' => StringLit(chars mkString "") } - | '\"' ~ rep( chrExcept('\"', '\n', EofCh) ) ~ '\"' ^^ - { case '\"' ~ chars ~ '\"' => StringLit(chars mkString "") } - | EofCh ^^^ EOF - | '\'' ~> failure("unclosed string literal") - | '\"' ~> failure("unclosed string literal") - | delim - | failure("illegal character") - ) - - override def identChar = letter | elem('_') - - override def whitespace: Parser[Any] = rep( - whitespaceChar - | '/' ~ '*' ~ comment - | '/' ~ '/' ~ rep( chrExcept(EofCh, '\n') ) - | '#' ~ rep( chrExcept(EofCh, '\n') ) - | '-' ~ '-' ~ rep( chrExcept(EofCh, '\n') ) - | '/' ~ '*' ~ failure("unclosed comment") - ) - - /** Generate all variations of upper and lower case of a given string */ - def allCaseVersions(s: String, prefix: String = ""): Stream[String] = { - if (s == "") { - Stream(prefix) - } else { - allCaseVersions(s.tail, prefix + s.head.toLower) ++ - allCaseVersions(s.tail, prefix + s.head.toUpper) - } - } -} diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/commands.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/commands.scala index 9a3848cfc6..b8ba2ee428 100644 --- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/commands.scala +++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/commands.scala @@ -39,9 +39,9 @@ case class NativeCommand(cmd: String) extends Command { } /** - * Commands of the form "SET (key) (= value)". + * Commands of the form "SET [key [= value] ]". */ -case class SetCommand(key: Option[String], value: Option[String]) extends Command { +case class SetCommand(kv: Option[(String, Option[String])]) extends Command { override def output = Seq( AttributeReference("", StringType, nullable = false)()) } @@ -81,3 +81,14 @@ case class DescribeCommand( AttributeReference("data_type", StringType, nullable = false)(), AttributeReference("comment", StringType, nullable = false)()) } + +/** + * Returned for the "! shellCommand" command + */ +case class ShellCommand(cmd: String) extends Command + + +/** + * Returned for the "SOURCE file" command + */ +case class SourceCommand(filePath: String) extends Command diff --git a/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala b/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala index 014e1e2826..23e7b2d270 100644 --- a/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala +++ b/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala @@ -66,12 +66,17 @@ class SQLContext(@transient val sparkContext: SparkContext) @transient protected[sql] lazy val analyzer: Analyzer = new Analyzer(catalog, functionRegistry, caseSensitive = true) + @transient protected[sql] val optimizer = Optimizer + @transient - protected[sql] val parser = new catalyst.SqlParser + protected[sql] val sqlParser = { + val fallback = new catalyst.SqlParser + new catalyst.SparkSQLParser(fallback(_)) + } - protected[sql] def parseSql(sql: String): LogicalPlan = parser(sql) + protected[sql] def parseSql(sql: String): LogicalPlan = sqlParser(sql) protected[sql] def executeSql(sql: String): this.QueryExecution = executePlan(parseSql(sql)) protected[sql] def executePlan(plan: LogicalPlan): this.QueryExecution = new this.QueryExecution { val logical = plan } diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala index bbf17b9fad..4f1af7234d 100644 --- a/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala +++ b/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala @@ -304,8 +304,8 @@ private[sql] abstract class SparkStrategies extends QueryPlanner[SparkPlan] { case class CommandStrategy(context: SQLContext) extends Strategy { def apply(plan: LogicalPlan): Seq[SparkPlan] = plan match { - case logical.SetCommand(key, value) => - Seq(execution.SetCommand(key, value, plan.output)(context)) + case logical.SetCommand(kv) => + Seq(execution.SetCommand(kv, plan.output)(context)) case logical.ExplainCommand(logicalPlan, extended) => Seq(execution.ExplainCommand(logicalPlan, plan.output, extended)(context)) case logical.CacheTableCommand(tableName, optPlan, isLazy) => diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/commands.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/commands.scala index d49633c24a..5859eba408 100644 --- a/sql/core/src/main/scala/org/apache/spark/sql/execution/commands.scala +++ b/sql/core/src/main/scala/org/apache/spark/sql/execution/commands.scala @@ -48,29 +48,28 @@ trait Command { * :: DeveloperApi :: */ @DeveloperApi -case class SetCommand( - key: Option[String], value: Option[String], output: Seq[Attribute])( +case class SetCommand(kv: Option[(String, Option[String])], output: Seq[Attribute])( @transient context: SQLContext) extends LeafNode with Command with Logging { - override protected lazy val sideEffectResult: Seq[Row] = (key, value) match { - // Set value for key k. - case (Some(k), Some(v)) => - if (k == SQLConf.Deprecated.MAPRED_REDUCE_TASKS) { + override protected lazy val sideEffectResult: Seq[Row] = kv match { + // Set value for the key. + case Some((key, Some(value))) => + if (key == SQLConf.Deprecated.MAPRED_REDUCE_TASKS) { logWarning(s"Property ${SQLConf.Deprecated.MAPRED_REDUCE_TASKS} is deprecated, " + s"automatically converted to ${SQLConf.SHUFFLE_PARTITIONS} instead.") - context.setConf(SQLConf.SHUFFLE_PARTITIONS, v) - Seq(Row(s"${SQLConf.SHUFFLE_PARTITIONS}=$v")) + context.setConf(SQLConf.SHUFFLE_PARTITIONS, value) + Seq(Row(s"${SQLConf.SHUFFLE_PARTITIONS}=$value")) } else { - context.setConf(k, v) - Seq(Row(s"$k=$v")) + context.setConf(key, value) + Seq(Row(s"$key=$value")) } - // Query the value bound to key k. - case (Some(k), _) => + // Query the value bound to the key. + case Some((key, None)) => // TODO (lian) This is just a workaround to make the Simba ODBC driver work. // Should remove this once we get the ODBC driver updated. - if (k == "-v") { + if (key == "-v") { val hiveJars = Seq( "hive-exec-0.12.0.jar", "hive-service-0.12.0.jar", @@ -84,23 +83,20 @@ case class SetCommand( Row("system:java.class.path=" + hiveJars), Row("system:sun.java.command=shark.SharkServer2")) } else { - if (k == SQLConf.Deprecated.MAPRED_REDUCE_TASKS) { + if (key == SQLConf.Deprecated.MAPRED_REDUCE_TASKS) { logWarning(s"Property ${SQLConf.Deprecated.MAPRED_REDUCE_TASKS} is deprecated, " + s"showing ${SQLConf.SHUFFLE_PARTITIONS} instead.") Seq(Row(s"${SQLConf.SHUFFLE_PARTITIONS}=${context.numShufflePartitions}")) } else { - Seq(Row(s"$k=${context.getConf(k, "<undefined>")}")) + Seq(Row(s"$key=${context.getConf(key, "<undefined>")}")) } } // Query all key-value pairs that are set in the SQLConf of the context. - case (None, None) => + case _ => context.getAllConfs.map { case (k, v) => Row(s"$k=$v") }.toSeq - - case _ => - throw new IllegalArgumentException() } override def otherCopyArgs = context :: Nil diff --git a/sql/core/src/test/scala/org/apache/spark/sql/CachedTableSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/CachedTableSuite.scala index 1e624f9700..c87ded81fd 100644 --- a/sql/core/src/test/scala/org/apache/spark/sql/CachedTableSuite.scala +++ b/sql/core/src/test/scala/org/apache/spark/sql/CachedTableSuite.scala @@ -69,7 +69,7 @@ class CachedTableSuite extends QueryTest { test("calling .unpersist() should drop in-memory columnar cache") { table("testData").cache() table("testData").count() - table("testData").unpersist(true) + table("testData").unpersist(blocking = true) assertCached(table("testData"), 0) } diff --git a/sql/core/src/test/scala/org/apache/spark/sql/SQLQuerySuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/SQLQuerySuite.scala index 79de1bb855..a94022c0cf 100644 --- a/sql/core/src/test/scala/org/apache/spark/sql/SQLQuerySuite.scala +++ b/sql/core/src/test/scala/org/apache/spark/sql/SQLQuerySuite.scala @@ -42,7 +42,6 @@ class SQLQuerySuite extends QueryTest with BeforeAndAfterAll { TimeZone.setDefault(origZone) } - test("SPARK-3176 Added Parser of SQL ABS()") { checkAnswer( sql("SELECT ABS(-1.3)"), @@ -61,7 +60,6 @@ class SQLQuerySuite extends QueryTest with BeforeAndAfterAll { 4) } - test("SPARK-2041 column name equals tablename") { checkAnswer( sql("SELECT tableName FROM tableName"), @@ -694,6 +692,6 @@ class SQLQuerySuite extends QueryTest with BeforeAndAfterAll { test("SPARK-3813 CASE WHEN a THEN b [WHEN c THEN d]* [ELSE e] END") { checkAnswer( - sql("SELECT CASE WHEN key=1 THEN 1 ELSE 2 END FROM testData WHERE key = 1 group by key"), 1) + sql("SELECT CASE WHEN key = 1 THEN 1 ELSE 2 END FROM testData WHERE key = 1 group by key"), 1) } } diff --git a/sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/server/SparkSQLOperationManager.scala b/sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/server/SparkSQLOperationManager.scala index 910174a153..accf61576b 100644 --- a/sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/server/SparkSQLOperationManager.scala +++ b/sql/hive-thriftserver/src/main/scala/org/apache/spark/sql/hive/thriftserver/server/SparkSQLOperationManager.scala @@ -172,7 +172,7 @@ private[thriftserver] class SparkSQLOperationManager(hiveContext: HiveContext) result = hiveContext.sql(statement) logDebug(result.queryExecution.toString()) result.queryExecution.logical match { - case SetCommand(Some(key), Some(value)) if (key == SQLConf.THRIFTSERVER_POOL) => + case SetCommand(Some((SQLConf.THRIFTSERVER_POOL, Some(value)))) => sessionToActivePool(parentSession) = value logInfo(s"Setting spark.scheduler.pool=$value for future statements in this session.") case _ => diff --git a/sql/hive/src/main/scala/org/apache/spark/sql/hive/ExtendedHiveQlParser.scala b/sql/hive/src/main/scala/org/apache/spark/sql/hive/ExtendedHiveQlParser.scala index c5844e92ea..430ffb2998 100644 --- a/sql/hive/src/main/scala/org/apache/spark/sql/hive/ExtendedHiveQlParser.scala +++ b/sql/hive/src/main/scala/org/apache/spark/sql/hive/ExtendedHiveQlParser.scala @@ -18,118 +18,50 @@ package org.apache.spark.sql.hive import scala.language.implicitConversions -import scala.util.parsing.combinator.syntactical.StandardTokenParsers -import scala.util.parsing.combinator.PackratParsers + import org.apache.spark.sql.catalyst.plans.logical._ -import org.apache.spark.sql.catalyst.SqlLexical +import org.apache.spark.sql.catalyst.{AbstractSparkSQLParser, SqlLexical} /** - * A parser that recognizes all HiveQL constructs together with several Spark SQL specific - * extensions like CACHE TABLE and UNCACHE TABLE. + * A parser that recognizes all HiveQL constructs together with Spark SQL specific extensions. */ -private[hive] class ExtendedHiveQlParser extends StandardTokenParsers with PackratParsers { - - def apply(input: String): LogicalPlan = { - // Special-case out set commands since the value fields can be - // complex to handle without RegexParsers. Also this approach - // is clearer for the several possible cases of set commands. - if (input.trim.toLowerCase.startsWith("set")) { - input.trim.drop(3).split("=", 2).map(_.trim) match { - case Array("") => // "set" - SetCommand(None, None) - case Array(key) => // "set key" - SetCommand(Some(key), None) - case Array(key, value) => // "set key=value" - SetCommand(Some(key), Some(value)) - } - } else if (input.trim.startsWith("!")) { - ShellCommand(input.drop(1)) - } else { - phrase(query)(new lexical.Scanner(input)) match { - case Success(r, x) => r - case x => sys.error(x.toString) - } - } - } - - protected case class Keyword(str: String) - - protected val ADD = Keyword("ADD") - protected val AS = Keyword("AS") - protected val CACHE = Keyword("CACHE") - protected val DFS = Keyword("DFS") - protected val FILE = Keyword("FILE") - protected val JAR = Keyword("JAR") - protected val LAZY = Keyword("LAZY") - protected val SET = Keyword("SET") - protected val SOURCE = Keyword("SOURCE") - protected val TABLE = Keyword("TABLE") - protected val UNCACHE = Keyword("UNCACHE") - +private[hive] class ExtendedHiveQlParser extends AbstractSparkSQLParser { protected implicit def asParser(k: Keyword): Parser[String] = lexical.allCaseVersions(k.str).map(x => x : Parser[String]).reduce(_ | _) - protected def allCaseConverse(k: String): Parser[String] = - lexical.allCaseVersions(k).map(x => x : Parser[String]).reduce(_ | _) + protected val ADD = Keyword("ADD") + protected val DFS = Keyword("DFS") + protected val FILE = Keyword("FILE") + protected val JAR = Keyword("JAR") - protected val reservedWords = - this.getClass + private val reservedWords = + this + .getClass .getMethods .filter(_.getReturnType == classOf[Keyword]) .map(_.invoke(this).asInstanceOf[Keyword].str) override val lexical = new SqlLexical(reservedWords) - protected lazy val query: Parser[LogicalPlan] = - cache | uncache | addJar | addFile | dfs | source | hiveQl + protected lazy val start: Parser[LogicalPlan] = dfs | addJar | addFile | hiveQl protected lazy val hiveQl: Parser[LogicalPlan] = restInput ^^ { - case statement => HiveQl.createPlan(statement.trim()) - } - - // Returns the whole input string - protected lazy val wholeInput: Parser[String] = new Parser[String] { - def apply(in: Input) = - Success(in.source.toString, in.drop(in.source.length())) - } - - // Returns the rest of the input string that are not parsed yet - protected lazy val restInput: Parser[String] = new Parser[String] { - def apply(in: Input) = - Success( - in.source.subSequence(in.offset, in.source.length).toString, - in.drop(in.source.length())) - } - - protected lazy val cache: Parser[LogicalPlan] = - CACHE ~> opt(LAZY) ~ (TABLE ~> ident) ~ opt(AS ~> hiveQl) ^^ { - case isLazy ~ tableName ~ plan => - CacheTableCommand(tableName, plan, isLazy.isDefined) - } - - protected lazy val uncache: Parser[LogicalPlan] = - UNCACHE ~ TABLE ~> ident ^^ { - case tableName => UncacheTableCommand(tableName) + case statement => HiveQl.createPlan(statement.trim) } - protected lazy val addJar: Parser[LogicalPlan] = - ADD ~ JAR ~> restInput ^^ { - case jar => AddJar(jar.trim()) + protected lazy val dfs: Parser[LogicalPlan] = + DFS ~> wholeInput ^^ { + case command => NativeCommand(command.trim) } - protected lazy val addFile: Parser[LogicalPlan] = + private lazy val addFile: Parser[LogicalPlan] = ADD ~ FILE ~> restInput ^^ { - case file => AddFile(file.trim()) + case input => AddFile(input.trim) } - protected lazy val dfs: Parser[LogicalPlan] = - DFS ~> wholeInput ^^ { - case command => NativeCommand(command.trim()) - } - - protected lazy val source: Parser[LogicalPlan] = - SOURCE ~> restInput ^^ { - case file => SourceCommand(file.trim()) + private lazy val addJar: Parser[LogicalPlan] = + ADD ~ JAR ~> restInput ^^ { + case input => AddJar(input.trim) } } diff --git a/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveQl.scala b/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveQl.scala index 32c9175f18..98a46a31e1 100644 --- a/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveQl.scala +++ b/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveQl.scala @@ -21,6 +21,7 @@ import org.apache.hadoop.hive.ql.lib.Node import org.apache.hadoop.hive.ql.parse._ import org.apache.hadoop.hive.ql.plan.PlanUtils +import org.apache.spark.sql.catalyst.SparkSQLParser import org.apache.spark.sql.catalyst.analysis._ import org.apache.spark.sql.catalyst.expressions._ import org.apache.spark.sql.catalyst.plans._ @@ -38,10 +39,6 @@ import scala.collection.JavaConversions._ */ private[hive] case object NativePlaceholder extends Command -private[hive] case class ShellCommand(cmd: String) extends Command - -private[hive] case class SourceCommand(filePath: String) extends Command - private[hive] case class AddFile(filePath: String) extends Command private[hive] case class AddJar(path: String) extends Command @@ -126,9 +123,11 @@ private[hive] object HiveQl { "TOK_CREATETABLE", "TOK_DESCTABLE" ) ++ nativeCommands - - // It parses hive sql query along with with several Spark SQL specific extensions - protected val hiveSqlParser = new ExtendedHiveQlParser + + protected val hqlParser = { + val fallback = new ExtendedHiveQlParser + new SparkSQLParser(fallback(_)) + } /** * A set of implicit transformations that allow Hive ASTNodes to be rewritten by transformations @@ -218,7 +217,7 @@ private[hive] object HiveQl { def getAst(sql: String): ASTNode = ParseUtils.findRootNonNullToken((new ParseDriver).parse(sql)) /** Returns a LogicalPlan for a given HiveQL string. */ - def parseSql(sql: String): LogicalPlan = hiveSqlParser(sql) + def parseSql(sql: String): LogicalPlan = hqlParser(sql) /** Creates LogicalPlan for a given HiveQL string. */ def createPlan(sql: String) = { diff --git a/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveStrategies.scala b/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveStrategies.scala index 508d8239c7..5c66322f1e 100644 --- a/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveStrategies.scala +++ b/sql/hive/src/main/scala/org/apache/spark/sql/hive/HiveStrategies.scala @@ -167,10 +167,10 @@ private[hive] trait HiveStrategies { database.get, tableName, query, - InsertIntoHiveTable(_: MetastoreRelation, - Map(), - query, - true)(hiveContext)) :: Nil + InsertIntoHiveTable(_: MetastoreRelation, + Map(), + query, + overwrite = true)(hiveContext)) :: Nil case _ => Nil } } |