[SPARK-12820][SQL]Resolve db.table.column

[zhichao-li]

Currently spark only support to specify col name like: table.col, or col in projection, but it's very common that user use db.table.col especially when join table across database.
Hive doesn't support this for now but it has been used in lot of other traditional db like mysql.

[rxin]

When we have a database named "a", and two tables:

1. a table also named "a", and a struct column named "b", and a field in the struct named "c".
2. a table named "b", and a column in it named "c"

What happens if the user specifies "a.b.c"?

[zhichao-li]

It would return attribute c in table b up on this patch. Any suggestion? throw ambiguous exception for such case?　

[rxin]

What do we do if the user specifies "a.b" right now? I'd say we should follow that, and make sure we have a test case for it.

[zhichao-li]

Would return attribute b in table a. resolved within the logic of table.col

[SparkQA]

Test build #49381 has finished for PR 10753 at commit 1cafed7.

• This patch passes all tests.
• This patch merges cleanly.
• This patch adds the following public classes (experimental):
  • case class Subquery(alias: String, child: LogicalPlan, databaseName: Option[String] = None)

[chenghao-intel]

I would say we'd better keep the same checking logic with mysql/hive for the ambiguous case, @zhichao-li can you please check that with mysql/hive?

[zhichao-li]

@chenghao-intel , mysql(5.5) doesn't support nested data type for now, and hive(1.2.1) doesn't support the usage of "db.table.field" in projection list at the moment. I guess most of the user work around this by using alias.

[SparkQA]

Test build #51930 has finished for PR 10753 at commit 4155ffe.

• This patch fails Spark unit tests.
• This patch does not merge cleanly.
• This patch adds no public classes.

[SparkQA]

Test build #52031 has finished for PR 10753 at commit a3c4f20.

• This patch fails Spark unit tests.
• This patch merges cleanly.
• This patch adds no public classes.

[rxin]

Thanks for the pull request. I'm going through a list of pull requests to cut them down since the sheer number is breaking some of the tooling we have. Due to lack of activity on this pull request, I'm going to push a commit to close it. Feel free to reopen it or create a new one. We can also continue the discussion on the JIRA ticket.