[SPARK-5472][SQL] A JDBC data source for Spark SQL.

This pull request contains a Spark SQL data source that can pull data from, and can put data into, a JDBC database.

I have tested both read and write support with H2, MySQL, and Postgres.  It would surprise me if both read and write support worked flawlessly out-of-the-box for any other database; different databases have different names for different JDBC data types and different meanings for SQL types with the same name.  However, this code is designed (see `DriverQuirks.scala`) to make it *relatively* painless to add support for another database by augmenting the type mapping contained in this PR.

Author: Tor Myklebust <tmyklebu@gmail.com>

Closes #4261 from tmyklebu/master and squashes the following commits:

cf167ce [Tor Myklebust] Work around other Java tests ruining TestSQLContext.
67893bf [Tor Myklebust] Move the jdbcRDD methods into SQLContext itself.
585f95b [Tor Myklebust] Dependencies go into the project's pom.xml.
829d5ba [Tor Myklebust] Merge branch 'master' of https://github.com/apache/spark
41647ef [Tor Myklebust] Hide a couple things that don't need to be public.
7318aea [Tor Myklebust] Fix scalastyle warnings.
a09eeac [Tor Myklebust] JDBC data source for Spark SQL.
176bb98 [Tor Myklebust] Add test deps for JDBC support.