Victoria Markman created DRILL-3409:
---------------------------------------
Summary: Specifying default frame explicitly results in an error
Key: DRILL-3409
URL: https://issues.apache.org/jira/browse/DRILL-3409
Project: Apache Drill
Issue Type: Bug
Components: Query Planning & Optimization
Affects Versions: 1.1.0
Reporter: Victoria Markman
Assignee: Jinfeng Ni
If I spell out default frame, I get an error:
{code}
0: jdbc:drill:schema=dfs> select c_bigint, min(c_double) over(partition by
c_bigint order by c_date, c_time nulls first range between unbounded preceding
and current row) from j9;
Error: PARSE ERROR: From line 1, column 95 to line 1, column 99: RANGE clause
cannot be used with compound ORDER BY clause
[Error Id: fe955fc0-bc0f-4588-bdc2-24defdc9390c on atsqa4-133.qa.lab:31010]
(state=,code=0)
{code}
If I don't specify explicitly "default" frame as in the example above: query
parses and returns the same result as Postgres:
{code}
0: jdbc:drill:schema=dfs> explain plan for select c_bigint, min(c_double)
over(partition by c_bigint order by c_date, c_time nulls first) from j9;
00-00 Screen
00-01 ProjectAllowDup(c_bigint=[$0], EXPR$1=[$1])
00-02 Project(c_bigint=[$1], w0$o0=[$5])
00-03 Window(window#0=[window(partition {1} order by [3, 4
ASC-nulls-first] range between UNBOUNDED PRECEDING and CURRENT ROW aggs
[MIN($2)])])
00-04 SelectionVectorRemover
00-05 Sort(sort0=[$1], sort1=[$3], sort2=[$4], dir0=[ASC],
dir1=[ASC], dir2=[ASC-nulls-first])
00-06 Project(T32¦¦*=[$0], c_bigint=[$1], c_double=[$2],
c_date=[$3], c_time=[$4])
00-07 Scan(groupscan=[ParquetGroupScan
[entries=[ReadEntryWithPath [path=maprfs:///drill/testdata/subqueries/j9]],
selectionRoot=/drill/testdata/subqueries/j9, numFiles=1, columns=[`*`]]])
{code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
