Github user xuchuanyin commented on the issue:
https://github.com/apache/carbondata/pull/2252
@kumarvishal09 please review the latest update.
1. still use 'long_string_columns' instead of `varchar` datatype to make it
consistent it spark/hive
2. internally add a new datatype called `text` to represent the long string
column
3. add a new encoding called DIRECT_COMPRESS_TEXT to the text column page
meta
4. use an integer (previously short) to store the length of bytes content.
5. A test was added to test query/select on the text columns
---
