expose createConverterToSQL for spark-bigquery

[nevillelyh]

No description provided.

[nevillelyh]

We are using this for RDD[GenericRecord] -> DataFrame conversion in spark-bigquery:
https://github.com/spotify/spark-bigquery/blob/master/src/main/scala/com/spotify/spark/bigquery/package.scala#L102

[codecov-io]

Current coverage is 94.91%

Merging #132 into master will not change coverage

@@             master       #132   diff @@
==========================================
  Files             6          6          
  Lines           275        275          
  Methods         252        252          
  Messages          0          0          
  Branches         23         23          
==========================================
  Hits            261        261          
  Misses           14         14          
  Partials          0          0          

Powered by Codecov. Last updated by 5a40e16...913dbb6

[JoshRosen]

Hey, so I think that adding an RDD[AvroRecord] to DataFrame conversion would make a bunch of sense, but I'm not sure if this is the best / cleanest way to do this, especially given the lack of user-facing documentation and the awkward interfaces / names exposed here. Given this, I'm going to have to close this PR, but I'll try to address this use-case in a more principled way by exposing more conversions with a cleaner API.