Github user jackylk commented on a diff in the pull request:
https://github.com/apache/carbondata/pull/2768#discussion_r221411345
--- Diff:
core/src/main/java/org/apache/carbondata/core/metadata/schema/table/TableSchemaBuilder.java
---
@@ -224,7 +224,7 @@ private ColumnSchema addColumn(StructField field,
String parentName, AtomicInteg
}
}
}
- if (newColumn.isDimensionColumn()) {
+ if (newColumn.isDimensionColumn() && newColumn.isSortColumn()) {
--- End diff --
When I change CarbonCliTest to write carbon files with no sort,
```
TestUtil.writeFilesAndVerify(5000000, new Schema(fields), path, new
String[0], 3, 8);
```
The output of the CarbonCli is
```
## Schema
schema in part-0-31810056985663_batchno0-0-null-31809010599854.carbondata
version: V3
timestamp: 2018-09-29 09:41:19.839
Column Name Data Type Column Type SortColumn Encoding Ordinal Id
name STRING dimension false [] 0 NA
age INT measure false [] 1 NA
```
So it indicate that the sort columns property in each ColumnSchema is
false, right?
---
