Github user ravipesala commented on the issue:
https://github.com/apache/carbondata/pull/2261
@sounakr Building schema should be common across spark-integration and sdk,
there should not be two paths. I feel it is better to refactor to use common
code.---
