[SPARK-5651][SQL] Update input46.q to be consistent with Hive

[OopsOutOfMemory]

This PR Update input46.q.

origin hive013 support `db.table` but it's a bug that fixed in hive0.14:
```
create table `table_in_database_creation.test2` as select * from src limit 1; in HiveContext.
```
in Hive0.14 we change it to `db`.`table`, and we should make it consistent with hive input46.q
```scala
create table `table_in_database_creation`.`test2` as select * from src limit 1; in HiveContext.
```

[AmplabJenkins]

Can one of the admins verify this patch?

[rxin]

Jenkins, test this please.

[SparkQA]

Test build #27150 has started for PR 4473 at commit 2db7db5.

• This patch merges cleanly.

[watermen]

@OopsOutOfMemory @rxin See the end of my PR(#4427), @yhuai say it is a bug in Hive and it has been fixed by https://issues.apache.org/jira/browse/HIVE-7673 (in Hive 0.14 and trunk).

[SparkQA]

Test build #27150 has finished for PR 4473 at commit 2db7db5.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds no public classes.

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/SparkPullRequestBuilder/27150/
Test PASSed.

[watermen]

@OopsOutOfMemory

create table `table_in_database_creation`.`test2` as select * from src limit 1;
create table `table_in_database_creation`.`test4` (a string);

are all supported in the latest version of the spark. You can try it.

[OopsOutOfMemory]

yes, I know it support because it's (db,tableName) case.