[KYUUBI #280] Align Operation GET_SCHEMAS * behavior with Hive #292

![pan3793](https://badgen.net/badge/Hello/pan3793/green) [![PR 292](https://badgen.net/badge/Preview/PR%20292/blue)](https://github.com/yaooqinn/kyuubi/pull/292) ![Bug](https://badgen.net/badge/Label/Bug/) [&#10088;?&#10089;](https://pullrequestbadge.com/?utm_medium=github&utm_source=yaooqinn&utm_campaign=badge_info)<!-- PR-BADGE: PLEASE DO NOT REMOVE THIS COMMENT -->

<!--
Thanks for sending a pull request!  Here are some tips for you:
  1. If this is your first time, please read our contributor guidelines: https://kyuubi.readthedocs.io/en/latest/community/contributions.html
-->

<!-- replace ${issue ID} with the actual issue id -->
Fixes #280
closes #292

<!--
Please clarify why the changes are needed. For instance,
  1. If you add a feature, you can talk about the user case of it.
  2. If you fix a bug, you can clarify why it is a bug.
-->
Before this PR, Kyuubi treat Operation GET_SCHEMAS * as an invalid Operation then cause HUE list databases failed, but HiveServer2 will return all databases on same request.

- [x] Add some test cases that check the changes thoroughly including negative and positive cases if possible
- [x] Add screenshots for manual tests if appropriate
- [x] [Run test](https://kyuubi.readthedocs.io/en/latest/tools/testing.html#running-tests) locally before make a pull request

After PR:
![201610862635_ pic](https://user-images.githubusercontent.com/26535726/104832256-032fc600-58cb-11eb-839b-6ece5328245a.jpg)
Squashed commit of the following:

commit f56fdc3fe35b4aa5a7b6f09cbbaa023ec713e2fe
Author: Cheng Pan <379377944@qq.com>
Date:   Sun Jan 17 23:27:03 2021 +0800

    [KYUUBI-280] code style and ut

commit ff74dbaf3e6d5606e5a2171e6dcc3628aa631eea
Author: Cheng Pan <379377944@qq.com>
Date:   Sun Jan 17 23:08:59 2021 +0800

    [KYUUBI-280] remove comments and handle schema * in other place

commit a8bcf98cd70e36903ece654e1c37d08e868b1c5b
Author: Cheng Pan <379377944@qq.com>
Date:   Sun Jan 17 13:47:24 2021 +0800

    [KYUUBI-280] Align Operation GET_SCHEMAS * behavior with Hive
This commit is contained in:
Cheng Pan 2021-01-18 11:21:59 +08:00 committed by Kent Yao
parent 3a147f834b
commit 6f8564ce59
No known key found for this signature in database
GPG Key ID: A4F0BE81C89B595B
5 changed files with 37 additions and 25 deletions

View File

@ -21,6 +21,7 @@ import java.util.regex.Pattern
import scala.collection.mutable.ArrayBuffer
import org.apache.commons.lang3.StringUtils
import org.apache.spark.sql.{Row, SparkSession}
import org.apache.spark.sql.types.{ArrayType, BinaryType, BooleanType, ByteType, CalendarIntervalType, DataType, DateType, DecimalType, DoubleType, FloatType, IntegerType, LongType, MapType, NullType, NumericType, ShortType, StringType, StructField, StructType, TimestampType}
@ -187,7 +188,9 @@ class GetColumns(
val gviews = new ArrayBuffer[Row]()
val globalTmpDb = catalog.globalTempViewManager.database
if (Pattern.compile(schemaPattern).matcher(globalTmpDb).matches()) {
if (StringUtils.isEmpty(schemaName) || schemaName == "*"
|| Pattern.compile(convertSchemaPattern(schemaName, false))
.matcher(globalTmpDb).matches()) {
catalog.globalTempViewManager.listViewNames(tablePattern).foreach { v =>
catalog.globalTempViewManager.get(v).foreach { plan =>
plan.schema.zipWithIndex

View File

@ -19,6 +19,7 @@ package org.apache.kyuubi.engine.spark.operation
import java.util.regex.Pattern
import org.apache.commons.lang3.StringUtils
import org.apache.spark.sql.{Row, SparkSession}
import org.apache.spark.sql.types.StructType
@ -44,8 +45,9 @@ class GetSchemas(spark: SparkSession, session: Session, catalogName: String, sch
val schemaPattern = convertSchemaPattern(schema)
val databases = spark.sessionState.catalog.listDatabases(schemaPattern)
val globalTmpViewDb = spark.sessionState.catalog.globalTempViewManager.database
if (Pattern.compile(convertSchemaPattern(schema, false))
.matcher(globalTmpViewDb).matches()) {
if (StringUtils.isEmpty(schema) || schema == "*"
|| Pattern.compile(convertSchemaPattern(schema, false))
.matcher(globalTmpViewDb).matches()) {
iter = (databases :+ globalTmpViewDb).map(Row(_, "")).toList.iterator
} else {
iter = databases.map(Row(_, "")).toList.iterator

View File

@ -19,6 +19,7 @@ package org.apache.kyuubi.engine.spark.operation
import java.util.regex.Pattern
import org.apache.commons.lang3.StringUtils
import org.apache.spark.sql.{Row, SparkSession}
import org.apache.spark.sql.catalyst.catalog.CatalogTableType
import org.apache.spark.sql.types.StructType
@ -51,7 +52,7 @@ class GetTables(
.add(TABLE_NAME, "string", nullable = true, "Table name.")
.add(TABLE_TYPE, "string", nullable = true, "The table type, e.g. \"TABLE\", \"VIEW\"")
.add(REMARKS, "string", nullable = true, "Comments about the table.")
.add("TYPE_CAT", "string", nullable = true, "The types catalog." )
.add("TYPE_CAT", "string", nullable = true, "The types catalog.")
.add("TYPE_SCHEM", "string", nullable = true, "the types schema (may be null)")
.add("TYPE_NAME", "string", nullable = true, "Type name.")
.add("SELF_REFERENCING_COL_NAME", "string", nullable = true,
@ -82,8 +83,9 @@ class GetTables(
val views = if (matched(CatalogTableType.VIEW)) {
val globalTempViewDb = catalog.globalTempViewManager.database
(if (Pattern.compile(convertSchemaPattern(schema, datanucleusFormat = false))
.matcher(globalTempViewDb).matches()) {
(if (StringUtils.isEmpty(schema) || schema == "*"
|| Pattern.compile(convertSchemaPattern(schema, datanucleusFormat = false))
.matcher(globalTempViewDb).matches()) {
catalog.listTables(globalTempViewDb, tablePattern, includeLocalTempViews = true)
} else {
catalog.listLocalTempViews(tablePattern)

View File

@ -84,8 +84,8 @@ class SparkOperationSuite extends WithSparkSQLEngine {
val metaData = statement.getConnection.getMetaData
Seq("%", null, ".*", "c.*") foreach { pattern =>
val rowSet = metaData.getColumns("", dftSchema, tableName, pattern)
Seq("%", null, ".*", "c.*") foreach { columnPattern =>
val rowSet = metaData.getColumns("", dftSchema, tableName, columnPattern)
import java.sql.Types._
val expectedJavaTypes = Seq(BOOLEAN, TINYINT, SMALLINT, INTEGER, BIGINT, FLOAT, DOUBLE,
@ -137,8 +137,8 @@ class SparkOperationSuite extends WithSparkSQLEngine {
assert(pos === 18, "all columns should have been verified")
}
val e = intercept[HiveSQLException](metaData.getColumns(null, "*", null, null))
assert(e.getCause.getMessage contains "Dangling meta character '*' near index 0\n*\n^")
val rowSet = metaData.getColumns(null, "*", "not_exist", "not_exist")
assert(!rowSet.next())
val e1 = intercept[HiveSQLException](metaData.getColumns(null, null, null, "*"))
assert(e1.getCause.getMessage contains "Dangling meta character '*' near index 0\n*\n^")

View File

@ -106,7 +106,7 @@ trait JDBCTests extends KyuubiFunSuite {
dbs.foreach(db => statement.execute(s"CREATE DATABASE IF NOT EXISTS $db"))
val metaData = statement.getConnection.getMetaData
Seq("", "%", null, ".*", "_*", "_%", ".%") foreach { pattern =>
Seq("", "*", "%", null, ".*", "_*", "_%", ".%") foreach { pattern =>
checkResult(metaData.getSchemas(null, pattern), dbs ++ dbDflts)
}
@ -120,9 +120,6 @@ trait JDBCTests extends KyuubiFunSuite {
checkResult(metaData.getSchemas(null, "db1"), Seq("db1"))
checkResult(metaData.getSchemas(null, "db_not_exist"), Seq.empty)
val e = intercept[HiveSQLException](metaData.getSchemas(null, "*"))
assert(e.getCause.getMessage contains "Dangling meta character '*' near index 0\n*\n^")
}
}
@ -166,25 +163,33 @@ trait JDBCTests extends KyuubiFunSuite {
}
assert(i === 4)
val rs3 = metaData.getTables(null, null, "table%", Array("VIEW"))
assert(!rs3.next())
val rs3 = metaData.getTables(null, "*", "*", Array("VIEW"))
i = 2
while(rs3.next()) {
assert(rs3.getString(TABLE_NAME) == tables(i))
i += 1
}
assert(i === 4)
val rs4 = metaData.getTables(null, null, "table%", Array("TABLE"))
val rs4 = metaData.getTables(null, null, "table%", Array("VIEW"))
assert(!rs4.next())
val rs5 = metaData.getTables(null, "*", "table%", Array("VIEW"))
assert(!rs5.next())
val rs6 = metaData.getTables(null, null, "table%", Array("TABLE"))
i = 0
while(rs4.next()) {
assert(rs4.getString(TABLE_NAME) == tables(i))
while(rs6.next()) {
assert(rs6.getString(TABLE_NAME) == tables(i))
i += 1
}
assert(i === 2)
val rs5 = metaData.getTables(null, "default", "%", Array("VIEW"))
val rs7 = metaData.getTables(null, "default", "%", Array("VIEW"))
i = 2
while(rs5.next()) {
assert(rs5.getString(TABLE_NAME) == view_test)
while(rs7.next()) {
assert(rs7.getString(TABLE_NAME) == view_test)
}
val e = intercept[HiveSQLException](metaData.getTables(null, "*", null, null))
assert(e.getCause.getMessage contains "Dangling meta character '*' near index 0\n*\n^")
}
}