[SPARK-11197] [SQL] run SQL on files directly

[davies]

This PR introduce a new feature to run SQL directly on files without create a table, for example:

select id from json.`path/to/json/files` as j

[SparkQA]

Test build #43953 has finished for PR 9173 at commit d1b0463.

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

[rxin]

The syntax looks good to me. Couple comments:

1. Let's add a config variable so users can disable this.
2. Make sure the error messages make sense. e.g. if there is a database named "json", the error message should say "there is no table or file named xxx"

[rxin]

Table not found

[SparkQA]

Test build #44007 has finished for PR 9173 at commit 4e7a995.

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

[SparkQA]

Test build #44008 has finished for PR 9173 at commit 827bdab.

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

[SparkQA]

Test build #44010 has finished for PR 9173 at commit 683921f.

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

[marmbrus]

Thanks! Merging to master.