[SPARK-17068][SQL] Make view-usage visible during analysis

[hvanhovell]

What changes were proposed in this pull request?

This PR adds a field to subquery alias in order to make the usage of views in a resolved LogicalPlan more visible (and more understandable).

For example, the following view and query:

create view constants as select 1 as id union all select 1 union all select 42
select * from constants;

...now yields the following analyzed plan:

Project [id#39]
+- SubqueryAlias c, `default`.`constants`
   +- Project [gen_attr_0#36 AS id#39]
      +- SubqueryAlias gen_subquery_0
         +- Union
            :- Union
            :  :- Project [1 AS gen_attr_0#36]
            :  :  +- OneRowRelation$
            :  +- Project [1 AS gen_attr_1#37]
            :     +- OneRowRelation$
            +- Project [42 AS gen_attr_2#38]
               +- OneRowRelation$

How was this patch tested?

Added tests for the two code paths in SessionCatalogSuite (sql/core) and HiveMetastoreCatalogSuite (sql/hive)

[SparkQA]

Test build #63811 has finished for PR 14657 at commit 6576496.

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

[hvanhovell]

cc @rxin @sameeragarwal

[SparkQA]

Test build #63841 has finished for PR 14657 at commit 7b39d2b.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • class HiveMetastoreCatalogSuite extends TestHiveSingleton with SQLTestUtils

[rxin]

not that big of a deal, but wouldn't it be easier and more robust if we just do

case s: SubqueryAlias => cleanQuery(s.child)
case p: Project => cleanQuery(p.child)
case child => child

[rxin]

Can you update the documentation of relevant classes to make sure we document this new contract clearly?

[SparkQA]

Test build #63868 has finished for PR 14657 at commit ddd4ea0.

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

[rxin]

LGTM - merging in master.

[rxin]

(I didn't merge this in 2.0)

[cloud-fan]

we are too conservative here, CatalogTable.tableType should never be null, and we use this assumption in a lot of places.