FilterPushDownRule::visitSubQuery does not consider aliased columns.

I found a bug for FilterPushDownRule as follows:

• Schema

  default> \d table1;
  
  table name: default.table1
  table path: hdfs://localhost:9010/tajo/warehouse/table1
  store type: CSV
  number of rows: 0
  volume: 60 B
  Options: 
  	'csvfile.delimiter'='|'
  
  schema: 
  id	INT4
  name	TEXT
  score	FLOAT4
  type	TEXT
  
  
  default> \d table2;
  
  table name: default.table2
  table path: hdfs://localhost:9010/tajo/warehouse/table2
  store type: CSV
  number of rows: 0
  volume: 61 B
  Options: 
  	'csvfile.delimiter'='|'
  
  schema: 
  id	INT4
  name	TEXT
  score	FLOAT4
  type	TEXT
  

• SQL

  select t1.id, t1.name, t1.cnt
  from (
    select a.id, a.name, count(*) as cnt
    from table1 a
    group by a.id, a.name
  ) t1
  where t1.cnt > 0
  

• Error Message

  java.lang.NullPointerException
  	at org.apache.tajo.catalog.Column.equals(Column.java:115)
  	at java.util.ArrayList.indexOf(ArrayList.java:216)
  	at java.util.ArrayList.contains(ArrayList.java:199)
  	at java.util.AbstractCollection.containsAll(AbstractCollection.java:278)
  	at org.apache.tajo.catalog.Schema.containsAll(Schema.java:269)
  	at org.apache.tajo.engine.planner.LogicalPlanner.checkIfBeEvaluatedAtRelation(LogicalPlanner.java:1575)
  	at org.apache.tajo.engine.planner.rewrite.FilterPushDownRule.visitScan(FilterPushDownRule.java:315)
  	at org.apache.tajo.engine.planner.rewrite.FilterPushDownRule.visitScan(FilterPushDownRule.java:35)
  
  

  If I remove where clause, it will run successfully.