[SPARK-13493][SQL] Enable case sensitiveness in json schema inference.

[zuowang]

What changes were proposed in this pull request?

Pass "caseSensitiveAnalysis" to InferSchema. Change field name to lowercase when "caseSensitiveAnalysis" is false.

How was this patch tested?

Add new test case in DataFrameSuite.

[zuowang]

What would be the case like in CSV? I don't think anybody want to conbine "user" and "User" column when "caseSensitiveAnalysis" is false.
scala> val jsonDF = sqlContext.read.option("header", "true").csv("file.csv")
jsonDF: org.apache.spark.sql.DataFrame = [user: string, User: string ... 1 more field]

scala> jsonDF.show
+-------+-------+-----+
| user| User| hits|
+-------+-------+-----+
| om| scala| 120|
| daniel| spark| 80|
|3754978| spark| 1|
+-------+-------+-----+

[maropu]

@rxin Currently, do we have any policy to handle case-sensitivity inside spark? e.g., postgresql always holds names into lower case (https://www.postgresql.org/docs/9.5/static/sql-syntax-lexical.html) and h2 does into upper one. It seems there are many case-sensitivity issue in not only json, but also other other datasources.

[AmplabJenkins]

Can one of the admins verify this patch?

[gatorsmile]

@maropu Yes. We need to consider case sensitivity for the built-in data sources at least.

[gatorsmile]

@zuowang@maropu Are you still working on this? We should resolve this issue.

[maropu]

I'm not, so it's okay to do so.

[zuowang]

I'm not, could you work to resolve it?

[HyukjinKwon]

Let's close this if it is not in a progress.

[gatorsmile]

We are closing it due to inactivity. please do reopen if you want to push it forward. Thanks!