SparkInterpreterdoesnotworkforSpark2versionofHDP2.5

What is this PR for?

Spark2 version of HDP 2.5 is "2.0.0.2.5.0.0-1245". Currently, Zeppelin parses this version to integer 2002500, which is higher than version "2.0.0" (i.e. 200 in integer). Therefore, Zeppelin thinks it's not supported version and fail.

I fixed some codes to consider major, minor, and patch version. Now, version parsed using only first 3 numbers (major, minor, patch) and parsed into a 5-digit integer. For example, 2.0.0 -> 20000, 1.6.2 -> 10602, 1.10.1 -> 11001. This makes HDP 2.5 compatible, and also make it more robust than before. Since maybe Spark's minor version can be a 2-digit number in the future

What type of PR is it?

[Bug Fix]

What is the Jira issue?

ZEPPELIN-1390

How should this be tested?

I added SparkVersion test codes.

该提问来源于开源项目：apache/zeppelin