Support Unpartitioned Analytic Invocation

[noahboerger]

UM	189	partitionClause

---

Issue Description

The partitionClause is defined in the reference manual 2.1 (p. 189) as partitionClause ::= partition by identifier { , identifier }* with the additional description of the default being If the clause is omitted, then the Data Set is partitioned by the Identifier Components that are not specified in the orderClause.

Therefore it is currently not possible to perform an analytic invocation without partitioning.

Proposed Solution

Support partitionClause without specifing any identifier or set the default when omitting the partitioningClause to this behaviour, to allow analytic invocations on unpartitioned datasets.

Example

When ds1 is

id_1	id_2	id_3	val
IDENTIFIER	IDENTIFIER	IDENTIFIER	MEASURE
INTEGER	INTEGER	INTEGER	INTEGER
1	1	1	1
1	1	2	1
1	2	1	1
2	1	1	1
2	2	1	1
2	2	2	1
3	1	1	1
3	2	1	1
3	2	2	1

The result of ds2 := ds1[calc rank_col := rank(over(order by id_1))]; or ds2 := ds1[calc rank_col := rank(over(partition by order by id_1))]; should be out of my point of view

id_1	id_2	id_3	val	rank_col
IDENTIFIER	IDENTIFIER	IDENTIFIER	MEASURE	MEASURE
INTEGER	INTEGER	INTEGER	INTEGER	INTEGER
1	1	1	1	1
1	1	2	1	1
1	2	1	1	1
2	1	1	1	4
2	2	1	1	4
2	2	2	1	4
3	1	1	1	7
3	2	1	1	7
3	2	2	1	7

but currently is (because of the default partitioning / not supported gramatic to have a partitionClause without specifying any identifier)

id_1	id_2	id_3	val	rank_col
IDENTIFIER	IDENTIFIER	IDENTIFIER	MEASURE	MEASURE
INTEGER	INTEGER	INTEGER	INTEGER	INTEGER
1	1	1	1	1
1	1	2	1	1
1	2	1	1	1
2	1	1	1	2
2	2	1	1	2
2	2	2	1	1
3	1	1	1	3
3	2	1	1	3
3	2	2	1	2

[antonio-olleros]

I agree, this seems like a limitation of the language, because it is not possible to get a result that may need to be calculated

[linardian]

Hi @noahboerger. First of all (sorry, but) I do not understand the results you are expecting, so if you can please elaborate a little bit more on this I would be very grateful. Secondly, the behaviour of the operator is exactly the one adopted by RDBMS (I made a test with ORACLE and DB2). Thank you so much.