Bugs item #1494263, was opened at 2006-05-24 14:47
Message generated for change (Comment added) made by sjoerd
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=1494263&group_id=56967

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Core
Group: MonetDB4 CVS Head
Status: Open
Resolution: Works For Me
Priority: 5
Private: No
Submitted By: Torsten Grust (teggy)
>Assigned to: Torsten Grust (teggy)
Summary: MIL: sintersect() and non-uniform BAT arguments

Initial Comment:
Hi guys,

  is the following supposed to work (the question
relates to any non-uniform mix of oid/void types in the
argument BATs)?

var t1;
var t2;

t1 := new (oid, oid);
t2 := new (oid, void);

t1.sintersect (t2);

It currently leads to:

!ERROR: interpret: no matching MIL operator to
'sintersect(BAT[oid,oid], BAT[oid,void])'.
!MAYBE YOU MEAN:
!       sintersect(BAT[any::1,any::2],
BAT[any::1,any::2]) : BAT[any::1,any::2]


In an ideal world, we would like to not materialize the
BATs before the sintersect().

Cheers,
   --Teggy

----------------------------------------------------------------------

>Comment By: Sjoerd Mullender (sjoerd)
Date: 2007-02-19 16:31

Message:
Logged In: YES 
user_id=43607
Originator: NO

Teggy, can this bug be closed?

----------------------------------------------------------------------

Comment By: Stefan Manegold (stmane)
Date: 2006-11-05 16:33

Message:
Logged In: YES 
user_id=572415

extended test to also check, whether sintersect set the
seqbase without property checking.
it does.

can this bug report be closed, now?


----------------------------------------------------------------------

Comment By: Niels Nes (nielsnes)
Date: 2006-05-28 11:19

Message:
Logged In: YES 
user_id=43556

added test in tests/BugTracker/Tests/sintersect_on_void_bats
The test however doesn't fail..

----------------------------------------------------------------------

Comment By: Sjoerd Mullender (sjoerd)
Date: 2006-05-24 17:02

Message:
Logged In: YES 
user_id=43607

sintersect and friends (kintersect, sdiff, kdiff) should set
the seqbase.

----------------------------------------------------------------------

Comment By: Torsten Grust (teggy)
Date: 2006-05-24 16:40

Message:
Logged In: YES 
user_id=642632

Hi Fabina (and Niels),

  thanks for your immediate reply.  I guess I have to be
more specific here -- my first posting was too vague.  Sorry
for that.  

Consider:
var a0, a1, a2;

a0 := new (void,void).seqbase ([EMAIL PROTECTED]).reverse ().seqbase ([EMAIL 
PROTECTED]);
a1 := new (void,void).seqbase ([EMAIL PROTECTED]).reverse ().seqbase ([EMAIL 
PROTECTED]);

a2 := a0.sintersect(a1);


That is, we're working with void|void BATs here, all of them
seqbased.  The first sintersect() works fine, yielding a BAT
of type oid|void which is then bound to variable a2.  Note
that the void column of a2 has *no* seqbase now.  (By the
way, it is unclear to me why one column comes out oid, the
other void.)

Is this considered the correct behavior?

Greetings,
   --Teggy



----------------------------------------------------------------------

Comment By: Fabian (mr-meltdown)
Date: 2006-05-24 15:54

Message:
Logged In: YES 
user_id=963970

(According to Niels, playing proxy here)

This is a valid error.  oid != void.  You have to set the
seqbase/mark on the void column before you can intersect them.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=1494263&group_id=56967

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Monetdb-bugs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/monetdb-bugs

Reply via email to