[SPARK-18922][TESTS] Fix more path-related test failures on Windows

## What changes were proposed in this pull request?

This PR proposes to fix the test failures due to different format of paths on Windows.

Failed tests are as below:

```
ColumnExpressionSuite:
- input_file_name, input_file_block_start, input_file_block_length - FileScanRDD *** FAILED *** (187 milliseconds)
  "file:///C:/projects/spark/target/tmp/spark-0b21b963-6cfa-411c-8d6f-e6a5e1e73bce/part-00001-c083a03a-e55e-4b05-9073-451de352d006.snappy.parquet" did not contain "C:\projects\spark\target\tmp\spark-0b21b963-6cfa-411c-8d6f-e6a5e1e73bce" (ColumnExpressionSuite.scala:545)

- input_file_name, input_file_block_start, input_file_block_length - HadoopRDD *** FAILED *** (172 milliseconds)
  "file:/C:/projects/spark/target/tmp/spark-5d0afa94-7c2f-463b-9db9-2e8403e2bc5f/part-00000-f6530138-9ad3-466d-ab46-0eeb6f85ed0b.txt" did not contain "C:\projects\spark\target\tmp\spark-5d0afa94-7c2f-463b-9db9-2e8403e2bc5f" (ColumnExpressionSuite.scala:569)

- input_file_name, input_file_block_start, input_file_block_length - NewHadoopRDD *** FAILED *** (156 milliseconds)
  "file:/C:/projects/spark/target/tmp/spark-a894c7df-c74d-4d19-82a2-a04744cb3766/part-00000-29674e3f-3fcf-4327-9b04-4dab1d46338d.txt" did not contain "C:\projects\spark\target\tmp\spark-a894c7df-c74d-4d19-82a2-a04744cb3766" (ColumnExpressionSuite.scala:598)
```

```
DataStreamReaderWriterSuite:
- source metadataPath *** FAILED *** (62 milliseconds)
  org.mockito.exceptions.verification.junit.ArgumentsAreDifferent: Argument(s) are different! Wanted:
streamSourceProvider.createSource(
    org.apache.spark.sql.SQLContext3b04133b,
    "C:\projects\spark\target\tmp\streaming.metadata-b05db6ae-c8dc-4ce4-b0d9-1eb8c84876c0/sources/0",
    None,
    "org.apache.spark.sql.streaming.test",
    Map()
);
-> at org.apache.spark.sql.streaming.test.DataStreamReaderWriterSuite$$anonfun$12.apply$mcV$sp(DataStreamReaderWriterSuite.scala:374)
Actual invocation has different arguments:
streamSourceProvider.createSource(
    org.apache.spark.sql.SQLContext3b04133b,
    "/C:/projects/spark/target/tmp/streaming.metadata-b05db6ae-c8dc-4ce4-b0d9-1eb8c84876c0/sources/0",
    None,
    "org.apache.spark.sql.streaming.test",
    Map()
);
```

```
GlobalTempViewSuite:
- CREATE GLOBAL TEMP VIEW USING *** FAILED *** (110 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark  arget mpspark-960398ba-a0a1-45f6-a59a-d98533f9f519;
```

```
CreateTableAsSelectSuite:
- CREATE TABLE USING AS SELECT *** FAILED *** (0 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- create a table, drop it and create another one with the same name *** FAILED *** (16 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- create table using as select - with partitioned by *** FAILED *** (0 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- create table using as select - with non-zero buckets *** FAILED *** (0 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string
```

```
HiveMetadataCacheSuite:
- partitioned table is cached when partition pruning is true *** FAILED *** (532 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- partitioned table is cached when partition pruning is false *** FAILED *** (297 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

```
MultiDatabaseSuite:
- createExternalTable() to non-default database - with USE *** FAILED *** (954 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark  arget mpspark-0839d9a7-5e29-467a-9e3e-3e4cd618ee09;

- createExternalTable() to non-default database - without USE *** FAILED *** (500 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark  arget mpspark-c7e24d73-1d8f-45e8-ab7d-53a83087aec3;

 - invalid database name and table names *** FAILED *** (31 milliseconds)
   "Path does not exist: file:/C:projectsspark  arget mpspark-15a2a494-3483-4876-80e5-ec396e704b77;" did not contain "`t:a` is not a valid name for tables/databases. Valid names only contain alphabet characters, numbers and _." (MultiDatabaseSuite.scala:296)
```

```
OrcQuerySuite:
 - SPARK-8501: Avoids discovery schema from empty ORC files *** FAILED *** (15 milliseconds)
   org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

 - Verify the ORC conversion parameter: CONVERT_METASTORE_ORC *** FAILED *** (78 milliseconds)
   org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

 - converted ORC table supports resolving mixed case field *** FAILED *** (297 milliseconds)
   org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

```
HadoopFsRelationTest - JsonHadoopFsRelationSuite, OrcHadoopFsRelationSuite, ParquetHadoopFsRelationSuite, SimpleTextHadoopFsRelationSuite:
 - Locality support for FileScanRDD *** FAILED *** (15 milliseconds)
   java.lang.IllegalArgumentException: Wrong FS: file://C:\projects\spark\target\tmp\spark-383d1f13-8783-47fd-964d-9c75e5eec50f, expected: file:///
```

```
HiveQuerySuite:
- CREATE TEMPORARY FUNCTION *** FAILED *** (0 milliseconds)
   java.net.MalformedURLException: For input string: "%5Cprojects%5Cspark%5Csql%5Chive%5Ctarget%5Cscala-2.11%5Ctest-classes%5CTestUDTF.jar"

 - ADD FILE command *** FAILED *** (500 milliseconds)
   java.net.URISyntaxException: Illegal character in opaque part at index 2: C:\projects\spark\sql\hive\target\scala-2.11\test-classes\data\files\v1.txt

 - ADD JAR command 2 *** FAILED *** (110 milliseconds)
   org.apache.spark.sql.AnalysisException: LOAD DATA input path does not exist: C:projectssparksqlhive  argetscala-2.11 est-classesdatafilessample.json;
```

```
PruneFileSourcePartitionsSuite:
 - PruneFileSourcePartitions should not change the output of LogicalRelation *** FAILED *** (15 milliseconds)
   org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

```
HiveCommandSuite:
 - LOAD DATA LOCAL *** FAILED *** (109 milliseconds)
   org.apache.spark.sql.AnalysisException: LOAD DATA input path does not exist: C:projectssparksqlhive  argetscala-2.11 est-classesdatafilesemployee.dat;

 - LOAD DATA *** FAILED *** (93 milliseconds)
   java.net.URISyntaxException: Illegal character in opaque part at index 15: C:projectsspark arget mpemployee.dat7496657117354281006.tmp

 - Truncate Table *** FAILED *** (78 milliseconds)
   org.apache.spark.sql.AnalysisException: LOAD DATA input path does not exist: C:projectssparksqlhive  argetscala-2.11 est-classesdatafilesemployee.dat;
```

```
HiveExternalCatalogBackwardCompatibilitySuite:
- make sure we can read table created by old version of Spark *** FAILED *** (0 milliseconds)
  "[/C:/projects/spark/target/tmp/]spark-0554d859-74e1-..." did not equal "[C:\projects\spark\target\tmp\]spark-0554d859-74e1-..." (HiveExternalCatalogBackwardCompatibilitySuite.scala:213)
  org.scalatest.exceptions.TestFailedException

- make sure we can alter table location created by old version of Spark *** FAILED *** (110 milliseconds)
  java.net.URISyntaxException: Illegal character in opaque part at index 15: C:projectsspark	arget	mpspark-0e9b2c5f-49a1-4e38-a32a-c0ab1813a79f
```

```
ExternalCatalogSuite:
- create/drop/rename partitions should create/delete/rename the directory *** FAILED *** (610 milliseconds)
  java.net.URISyntaxException: Illegal character in opaque part at index 2: C:\projects\spark\target\tmp\spark-4c24f010-18df-437b-9fed-990c6f9adece
```

```
SQLQuerySuite:
- describe functions - temporary user defined functions *** FAILED *** (16 milliseconds)
  java.net.URISyntaxException: Illegal character in opaque part at index 22: C:projectssparksqlhive	argetscala-2.11	est-classesTestUDTF.jar

- specifying database name for a temporary table is not allowed *** FAILED *** (125 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-a34c9814-a483-43f2-be29-37f616b6df91;
```

```
PartitionProviderCompatibilitySuite:
- convert partition provider to hive with repair table *** FAILED *** (281 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-ee5fc96d-8c7d-4ebf-8571-a1d62736473e;

- when partition management is enabled, new tables have partition provider hive *** FAILED *** (187 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-803ad4d6-3e8c-498d-9ca5-5cda5d9b2a48;

- when partition management is disabled, new tables have no partition provider *** FAILED *** (172 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-c9fda9e2-4020-465f-8678-52cd72d0a58f;

- when partition management is disabled, we preserve the old behavior even for new tables *** FAILED *** (203 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget
mpspark-f4a518a6-c49d-43d3-b407-0ddd76948e13;

- insert overwrite partition of legacy datasource table *** FAILED *** (188 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-f4a518a6-c49d-43d3-b407-0ddd76948e79;

- insert overwrite partition of new datasource table overwrites just partition *** FAILED *** (219 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-6ba3a88d-6f6c-42c5-a9f4-6d924a0616ff;

- SPARK-18544 append with saveAsTable - partition management true *** FAILED *** (173 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-cd234a6d-9cb4-4d1d-9e51-854ae9543bbd;

- SPARK-18635 special chars in partition values - partition management true *** FAILED *** (2 seconds, 967 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-18635 special chars in partition values - partition management false *** FAILED *** (62 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-18659 insert overwrite table with lowercase - partition management true *** FAILED *** (63 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-18544 append with saveAsTable - partition management false *** FAILED *** (266 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-18659 insert overwrite table files - partition management false *** FAILED *** (63 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-18659 insert overwrite table with lowercase - partition management false *** FAILED *** (78 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- sanity check table setup *** FAILED *** (31 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- insert into partial dynamic partitions *** FAILED *** (47 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- insert into fully dynamic partitions *** FAILED *** (62 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- insert into static partition *** FAILED *** (78 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- overwrite partial dynamic partitions *** FAILED *** (63 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- overwrite fully dynamic partitions *** FAILED *** (47 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- overwrite static partition *** FAILED *** (63 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

```
MetastoreDataSourcesSuite:
- check change without refresh *** FAILED *** (203 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-00713fe4-ca04-448c-bfc7-6c5e9a2ad2a1;

- drop, change, recreate *** FAILED *** (78 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-2030a21b-7d67-4385-a65b-bb5e2bed4861;

- SPARK-15269 external data source table creation *** FAILED *** (78 milliseconds)
  org.apache.spark.sql.AnalysisException: Path does not exist: file:/C:projectsspark	arget	mpspark-4d50fd4a-14bc-41d6-9232-9554dd233f86;

- CTAS *** FAILED *** (109 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- CTAS with IF NOT EXISTS *** FAILED *** (109 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- CTAS: persisted partitioned bucketed data source table *** FAILED *** (0 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- SPARK-15025: create datasource table with path with select *** FAILED *** (16 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- CTAS: persisted partitioned data source table *** FAILED *** (47 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string
```

```
HiveMetastoreCatalogSuite:
- Persist non-partitioned parquet relation into metastore as managed table using CTAS *** FAILED *** (16 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string

- Persist non-partitioned orc relation into metastore as managed table using CTAS *** FAILED *** (16 milliseconds)
  java.lang.IllegalArgumentException: Can not create a Path from an empty string
```

```
HiveUDFSuite:
- SPARK-11522 select input_file_name from non-parquet table *** FAILED *** (16 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

```
QueryPartitionSuite:
- SPARK-13709: reading partitioned Avro table with nested schema *** FAILED *** (250 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

```
ParquetHiveCompatibilitySuite:
- simple primitives *** FAILED *** (16 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-10177 timestamp *** FAILED *** (0 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- array *** FAILED *** (16 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- map *** FAILED *** (16 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- struct *** FAILED *** (0 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);

- SPARK-16344: array of struct with a single field named 'array_element' *** FAILED *** (15 milliseconds)
  org.apache.spark.sql.AnalysisException: org.apache.hadoop.hive.ql.metadata.HiveException: MetaException(message:java.lang.IllegalArgumentException: Can not create a Path from an empty string);
```

## How was this patch tested?

Manually tested via AppVeyor.

```
ColumnExpressionSuite:
- input_file_name, input_file_block_start, input_file_block_length - FileScanRDD (234 milliseconds)
- input_file_name, input_file_block_start, input_file_block_length - HadoopRDD (235 milliseconds)
- input_file_name, input_file_block_start, input_file_block_length - NewHadoopRDD (203 milliseconds)
```

```
DataStreamReaderWriterSuite:
- source metadataPath (63 milliseconds)
```

```
GlobalTempViewSuite:
 - CREATE GLOBAL TEMP VIEW USING (436 milliseconds)
```

```
CreateTableAsSelectSuite:
- CREATE TABLE USING AS SELECT (171 milliseconds)
- create a table, drop it and create another one with the same name (422 milliseconds)
- create table using as select - with partitioned by (141 milliseconds)
- create table using as select - with non-zero buckets (125 milliseconds)
```

```
HiveMetadataCacheSuite:
- partitioned table is cached when partition pruning is true (3 seconds, 211 milliseconds)
- partitioned table is cached when partition pruning is false (1 second, 781 milliseconds)
```

```
MultiDatabaseSuite:
 - createExternalTable() to non-default database - with USE (797 milliseconds)
 - createExternalTable() to non-default database - without USE (640 milliseconds)
 - invalid database name and table names (62 milliseconds)
```

```
OrcQuerySuite:
 - SPARK-8501: Avoids discovery schema from empty ORC files (703 milliseconds)
 - Verify the ORC conversion parameter: CONVERT_METASTORE_ORC (750 milliseconds)
 - converted ORC table supports resolving mixed case field (625 milliseconds)
```

```
HadoopFsRelationTest - JsonHadoopFsRelationSuite, OrcHadoopFsRelationSuite, ParquetHadoopFsRelationSuite, SimpleTextHadoopFsRelationSuite:
 - Locality support for FileScanRDD (296 milliseconds)
```

```
HiveQuerySuite:
 - CREATE TEMPORARY FUNCTION (125 milliseconds)
 - ADD FILE command (250 milliseconds)
 - ADD JAR command 2 (609 milliseconds)
```

```
PruneFileSourcePartitionsSuite:
- PruneFileSourcePartitions should not change the output of LogicalRelation (359 milliseconds)
```

```
HiveCommandSuite:
 - LOAD DATA LOCAL (1 second, 829 milliseconds)
 - LOAD DATA (1 second, 735 milliseconds)
 - Truncate Table (1 second, 641 milliseconds)
```

```
HiveExternalCatalogBackwardCompatibilitySuite:
 - make sure we can read table created by old version of Spark (32 milliseconds)
 - make sure we can alter table location created by old version of Spark (125 milliseconds)
 - make sure we can rename table created by old version of Spark (281 milliseconds)
```

```
ExternalCatalogSuite:
- create/drop/rename partitions should create/delete/rename the directory (625 milliseconds)
```

```
SQLQuerySuite:
- describe functions - temporary user defined functions (31 milliseconds)
- specifying database name for a temporary table is not allowed (390 milliseconds)
```

```
PartitionProviderCompatibilitySuite:
 - convert partition provider to hive with repair table (813 milliseconds)
 - when partition management is enabled, new tables have partition provider hive (562 milliseconds)
 - when partition management is disabled, new tables have no partition provider (344 milliseconds)
 - when partition management is disabled, we preserve the old behavior even for new tables (422 milliseconds)
 - insert overwrite partition of legacy datasource table (750 milliseconds)
 - SPARK-18544 append with saveAsTable - partition management true (985 milliseconds)
 - SPARK-18635 special chars in partition values - partition management true (3 seconds, 328 milliseconds)
 - SPARK-18635 special chars in partition values - partition management false (2 seconds, 891 milliseconds)
 - SPARK-18659 insert overwrite table with lowercase - partition management true (750 milliseconds)
 - SPARK-18544 append with saveAsTable - partition management false (656 milliseconds)
 - SPARK-18659 insert overwrite table files - partition management false (922 milliseconds)
 - SPARK-18659 insert overwrite table with lowercase - partition management false (469 milliseconds)
 - sanity check table setup (937 milliseconds)
 - insert into partial dynamic partitions (2 seconds, 985 milliseconds)
 - insert into fully dynamic partitions (1 second, 937 milliseconds)
 - insert into static partition (1 second, 578 milliseconds)
 - overwrite partial dynamic partitions (7 seconds, 561 milliseconds)
 - overwrite fully dynamic partitions (1 second, 766 milliseconds)
 - overwrite static partition (1 second, 797 milliseconds)
```

```
MetastoreDataSourcesSuite:
 - check change without refresh (610 milliseconds)
 - drop, change, recreate (437 milliseconds)
 - SPARK-15269 external data source table creation (297 milliseconds)
 - CTAS with IF NOT EXISTS (437 milliseconds)
 - CTAS: persisted partitioned bucketed data source table (422 milliseconds)
 - SPARK-15025: create datasource table with path with select (265 milliseconds)
 - CTAS (438 milliseconds)
 - CTAS with IF NOT EXISTS (469 milliseconds)
 - CTAS: persisted partitioned bucketed data source table (406 milliseconds)
```

```
HiveMetastoreCatalogSuite:
 - Persist non-partitioned parquet relation into metastore as managed table using CTAS (406 milliseconds)
 - Persist non-partitioned orc relation into metastore as managed table using CTAS (313 milliseconds)
```

```
HiveUDFSuite:
 - SPARK-11522 select input_file_name from non-parquet table (3 seconds, 144 milliseconds)
```

```
QueryPartitionSuite:
 - SPARK-13709: reading partitioned Avro table with nested schema (1 second, 67 milliseconds)
```

```
ParquetHiveCompatibilitySuite:
 - simple primitives (745 milliseconds)
 - SPARK-10177 timestamp (375 milliseconds)
 - array (407 milliseconds)
 - map (409 milliseconds)
 - struct (437 milliseconds)
 - SPARK-16344: array of struct with a single field named 'array_element' (391 milliseconds)
```

Author: hyukjinkwon <gurwls223@gmail.com>

Closes #16397 from HyukjinKwon/SPARK-18922-paths.

4 files changed, 19 insertions, 18 deletions

diff --git a/sql/core/src/test/scala/org/apache/spark/sql/ColumnExpressionSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/ColumnExpressionSuite.scala
index b0339a88fb..ee280a313c 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/ColumnExpressionSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/ColumnExpressionSuite.scala
@@ -542,7 +542,7 @@ class ColumnExpressionSuite extends QueryTest with SharedSQLContext {
       val q = spark.read.parquet(dir.getCanonicalPath).select(
         input_file_name(), expr("input_file_block_start()"), expr("input_file_block_length()"))
       val firstRow = q.head()
-      assert(firstRow.getString(0).contains(dir.getCanonicalPath))
+      assert(firstRow.getString(0).contains(dir.toURI.getPath))
       assert(firstRow.getLong(1) == 0)
       assert(firstRow.getLong(2) > 0)
 
@@ -566,7 +566,7 @@ class ColumnExpressionSuite extends QueryTest with SharedSQLContext {
       val q = df.select(
         input_file_name(), expr("input_file_block_start()"), expr("input_file_block_length()"))
       val firstRow = q.head()
-      assert(firstRow.getString(0).contains(dir.getCanonicalPath))
+      assert(firstRow.getString(0).contains(dir.toURI.getPath))
       assert(firstRow.getLong(1) == 0)
       assert(firstRow.getLong(2) > 0)
 
@@ -595,7 +595,7 @@ class ColumnExpressionSuite extends QueryTest with SharedSQLContext {
       val q = df.select(
         input_file_name(), expr("input_file_block_start()"), expr("input_file_block_length()"))
       val firstRow = q.head()
-      assert(firstRow.getString(0).contains(dir.getCanonicalPath))
+      assert(firstRow.getString(0).contains(dir.toURI.getPath))
       assert(firstRow.getLong(1) == 0)
       assert(firstRow.getLong(2) > 0)
 
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/execution/GlobalTempViewSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/execution/GlobalTempViewSuite.scala
index 391bcb8b35..5c63c6a414 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/execution/GlobalTempViewSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/execution/GlobalTempViewSuite.scala
@@ -93,7 +93,7 @@ class GlobalTempViewSuite extends QueryTest with SharedSQLContext {
     withTempPath { path =>
       try {
         Seq(1 -> "a").toDF("i", "j").write.parquet(path.getAbsolutePath)
-        sql(s"CREATE GLOBAL TEMP VIEW src USING parquet OPTIONS (PATH '${path.getAbsolutePath}')")
+        sql(s"CREATE GLOBAL TEMP VIEW src USING parquet OPTIONS (PATH '${path.toURI}')")
         checkAnswer(spark.table(s"$globalTempDB.src"), Row(1, "a"))
         sql(s"INSERT INTO $globalTempDB.src SELECT 2, 'b'")
         checkAnswer(spark.table(s"$globalTempDB.src"), Row(1, "a") :: Row(2, "b") :: Nil)
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/sources/CreateTableAsSelectSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/sources/CreateTableAsSelectSuite.scala
index 61939fe5ef..99da1969fc 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/sources/CreateTableAsSelectSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/sources/CreateTableAsSelectSuite.scala
@@ -70,7 +70,7 @@ class CreateTableAsSelectSuite
            |CREATE TABLE jsonTable
            |USING json
            |OPTIONS (
-           |  path '${path.toString}'
+           |  path '${path.toURI}'
            |) AS
            |SELECT a, b FROM jt
          """.stripMargin)
@@ -94,7 +94,7 @@ class CreateTableAsSelectSuite
            |CREATE TABLE jsonTable
            |USING json
            |OPTIONS (
-           |  path '${childPath.toString}'
+           |  path '${childPath.toURI}'
            |) AS
            |SELECT a, b FROM jt
          """.stripMargin)
@@ -112,7 +112,7 @@ class CreateTableAsSelectSuite
            |CREATE TABLE jsonTable
            |USING json
            |OPTIONS (
-           |  path '${path.toString}'
+           |  path '${path.toURI}'
            |) AS
            |SELECT a, b FROM jt
          """.stripMargin)
@@ -127,7 +127,7 @@ class CreateTableAsSelectSuite
            |CREATE TABLE IF NOT EXISTS jsonTable
            |USING json
            |OPTIONS (
-           |  path '${path.toString}'
+           |  path '${path.toURI}'
            |) AS
            |SELECT a * 4 FROM jt
          """.stripMargin)
@@ -145,7 +145,7 @@ class CreateTableAsSelectSuite
            |CREATE TABLE jsonTable
            |USING json
            |OPTIONS (
-           |  path '${path.toString}'
+           |  path '${path.toURI}'
            |) AS
            |SELECT b FROM jt
          """.stripMargin)
@@ -162,7 +162,7 @@ class CreateTableAsSelectSuite
         sql(
           s"""
              |CREATE TEMPORARY TABLE t USING PARQUET
-             |OPTIONS (PATH '${path.toString}')
+             |OPTIONS (PATH '${path.toURI}')
              |PARTITIONED BY (a)
              |AS SELECT 1 AS a, 2 AS b
            """.stripMargin
@@ -179,7 +179,7 @@ class CreateTableAsSelectSuite
         sql(
           s"""
              |CREATE EXTERNAL TABLE t USING PARQUET
-             |OPTIONS (PATH '${path.toString}')
+             |OPTIONS (PATH '${path.toURI}')
              |AS SELECT 1 AS a, 2 AS b
            """.stripMargin
         )
@@ -196,7 +196,7 @@ class CreateTableAsSelectSuite
       sql(
         s"""
            |CREATE TABLE t USING PARQUET
-           |OPTIONS (PATH '${path.toString}')
+           |OPTIONS (PATH '${path.toURI}')
            |PARTITIONED BY (a)
            |AS SELECT 1 AS a, 2 AS b
          """.stripMargin
@@ -212,7 +212,7 @@ class CreateTableAsSelectSuite
       sql(
         s"""
            |CREATE TABLE t USING PARQUET
-           |OPTIONS (PATH '${path.toString}')
+           |OPTIONS (PATH '${path.toURI}')
            |CLUSTERED BY (a) SORTED BY (b) INTO 5 BUCKETS
            |AS SELECT 1 AS a, 2 AS b
          """.stripMargin
@@ -228,7 +228,7 @@ class CreateTableAsSelectSuite
         sql(
           s"""
              |CREATE TABLE t USING PARQUET
-             |OPTIONS (PATH '${path.toString}')
+             |OPTIONS (PATH '${path.toURI}')
              |CLUSTERED BY (a) SORTED BY (b) INTO 0 BUCKETS
              |AS SELECT 1 AS a, 2 AS b
            """.stripMargin
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/streaming/test/DataStreamReaderWriterSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/streaming/test/DataStreamReaderWriterSuite.scala
index 097dd6e367..30a957ef81 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/streaming/test/DataStreamReaderWriterSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/streaming/test/DataStreamReaderWriterSuite.scala
@@ -22,6 +22,7 @@ import java.util.concurrent.TimeUnit
 
 import scala.concurrent.duration._
 
+import org.apache.hadoop.fs.Path
 import org.mockito.Mockito._
 import org.scalatest.{BeforeAndAfter, PrivateMethodTester}
 import org.scalatest.PrivateMethodTester.PrivateMethod
@@ -355,7 +356,7 @@ class DataStreamReaderWriterSuite extends StreamTest with BeforeAndAfter with Pr
   test("source metadataPath") {
     LastOptions.clear()
 
-    val checkpointLocation = newMetadataDir
+    val checkpointLocationURI = new Path(newMetadataDir).toUri
 
     val df1 = spark.readStream
       .format("org.apache.spark.sql.streaming.test")
@@ -367,21 +368,21 @@ class DataStreamReaderWriterSuite extends StreamTest with BeforeAndAfter with Pr
 
     val q = df1.union(df2).writeStream
       .format("org.apache.spark.sql.streaming.test")
-      .option("checkpointLocation", checkpointLocation)
+      .option("checkpointLocation", checkpointLocationURI.toString)
       .trigger(ProcessingTime(10.seconds))
       .start()
     q.stop()
 
     verify(LastOptions.mockStreamSourceProvider).createSource(
       spark.sqlContext,
-      checkpointLocation + "/sources/0",
+      s"$checkpointLocationURI/sources/0",
       None,
       "org.apache.spark.sql.streaming.test",
       Map.empty)
 
     verify(LastOptions.mockStreamSourceProvider).createSource(
       spark.sqlContext,
-      checkpointLocation + "/sources/1",
+      s"$checkpointLocationURI/sources/1",
       None,
       "org.apache.spark.sql.streaming.test",
       Map.empty)