[ monetdb-Bugs-1637867 ] XQ: seqcast int? to numeric unsupported

[SourceForge.net]

Bugs item #1637867, was opened at 2007-01-17 17:41
Message generated for change (Settings changed) made by stmane
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=1637867&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: PF/tests
Group: Pathfinder CVS Head
>Status: Closed
>Resolution: Out of Date
Priority: 8
Private: No
Submitted By: Stefan Manegold (stmane)
Assigned to: Jan Rittinger (tsheyar)
Summary: XQ: seqcast int? to numeric unsupported

Initial Comment:
Since somtime in November or December 2006 (sorry, I'm finally fed up with doing all the back-tracking),
tests
complex-XQuery-batbat_lng_add_inplace-error-OK.SF-1562868.xq
complex-XQuery-batbat_lng_add_inplace-error-KO.SF-1562868.xq
complex-XQuery-batbat_lng_add_inplace-error-KO2.SF-1562868.xq
in
tests/BugTracker/Tests/
fail with

ERROR = !type error: can't cast type 'xs:integer?' to type 'numeric'

http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.mTests103/GNU.64.64.d-Fedora4/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-OK.SF-1562868.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.mTests103/GNU.64.64.d-Fedora4/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-KO.SF-1562868.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.mTests103/GNU.64.64.d-Fedora4/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-KO2.SF-1562868.err.00.html


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

>Comment By: Stefan Manegold (stmane)
Date: 2009-11-07 14:02

Message:
closing as all related test seem to work fine now/again:

http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTests103/GNU.64.64.d.0-Fedora10/tests_BugTracker/seqcast_opt_int_to_numeric.SF-1637867.out.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTests103/GNU.64.64.d.0-Fedora10/tests_BugTracker/seqcast_opt_int_to_numeric.SF-1637867.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/seqcast_opt_int_to_numeric.SF-1637867.out.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/seqcast_opt_int_to_numeric.SF-1637867.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-OK.SF-1562868.out.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-OK.SF-1562868.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-KO.SF-1562868.out.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-KO.SF-1562868.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-KO2.SF-1562868.out.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTestsg103/GNU.64.64.d.0-Fedora10/tests_BugTracker/complex-XQuery-batbat_lng_add_inplace-error-KO2.SF-1562868.err.00.html



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

Comment By: Jan Rittinger (tsheyar)
Date: 2009-02-16 11:13

Message:
This one fails due to incorrect type disjointness test:
http://sourceforge.net/tracker2/?func=detail&aid=2211565&group_id=56967&atid=482468

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

Comment By: Stefan Manegold (stmane)
Date: 2009-02-16 00:43

Message:
re-opened as the test output (still?) differs from the stable one:
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/pathfinder/.mTests103/GNU.64.64.d.1-Fedora8/tests_BugTracker/seqcast_opt_int_to_numeric.SF-1637867.out.00.html


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

Comment By: Jan Rittinger (tsheyar)
Date: 2007-02-22 09:26

Message:
Logged In: YES
user_id=993208
Originator: NO

Testing did not produce more errors -- Thus I close the bug report.

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

Comment By: Jan Rittinger (tsheyar)
Date: 2007-02-21 14:30

Message:
Logged In: YES
user_id=993208
Originator: NO

I checked in a fix that removes seqcast operators that cast to numeric (--
these seqcasts were only generated for predicates). In addition I
introduced a core rewrite that replaces the typeswitches that test for the
type numeric.

The problematic query now works for me. I'm however not 100% sure whether
MPS does break now at a different place or not. Testing will hopefully tell
us tomorrow.

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

Comment By: Sjoerd Mullender (sjoerd)
Date: 2007-02-19 14:58

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

Who is responsible for getting this fixed?  Please assign somebody.

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

Comment By: Jan Rittinger (tsheyar)
Date: 2007-02-19 13:53

Message:
Logged In: YES
user_id=993208
Originator: NO

The mps version still cannot cope with the predicate that initially
triggered the bug. Thus the bug should remain open for now. See also:
http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.mTests103/GNU.64.64.d-Fedora4/tests_BugTracker/seqcast_opt_int_to_numeric.SF-1637867.err.00.html

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

Comment By: Sjoerd Mullender (sjoerd)
Date: 2007-02-19 13:38

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

What is the status of this bug?  Can it be closed?  If so, please do so.

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

Comment By: Stefan Manegold (stmane)
Date: 2007-01-18 20:08

Message:
Logged In: YES
user_id=572415
Originator: YES

Jan,

Thank you very much for clearing this up fixing the old test and adding
the new one!!

Stefan


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

Comment By: Jan Rittinger (tsheyar)
Date: 2007-01-18 18:10

Message:
Logged In: YES
user_id=993208
Originator: NO

changed summary (for explanation read my first comment)

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

Comment By: Jan Rittinger (tsheyar)
Date: 2007-01-18 17:19

Message:
Logged In: YES
user_id=993208
Originator: NO

Fixed complex... queries and introduced new query
seqcast_opt_int_to_numeric.SF-1637867 that tests the observed problem.

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

Comment By: Stefan Manegold (stmane)
Date: 2007-01-18 16:48

Message:
Logged In: YES
user_id=572415
Originator: YES

Maurice,

it's "your bug", so feel free to fix the tests ;-)

Thank you very much in advance!

Stefan


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

Comment By: Maurice van Keulen (mvankeulen)
Date: 2007-01-18 16:39

Message:
Logged In: YES
user_id=654674
Originator: NO

The problematic expressions are like "$seq[path cast as xs:integer]" which
is not value anymore. What I have found to be accepted is
"$seq[string(path) cast as xs:integer]" or "$seq[string(exactly-one(path))
cast as xs:integer]". Apparently, the requested type is numeric, so perhaps
'cast as numeric' will work as well.
I believe we should adapt these testcases, because otherwise they do not
test the original problem anymore.

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

Comment By: Stefan Manegold (stmane)
Date: 2007-01-18 15:26

Message:
Logged In: YES
user_id=572415
Originator: YES

If the error message is the correct/expected behaviour after the static
typing changes for this test, then the error message should be approved as
such by who ever changed the static typing and hence triggered the new
output --- the respective checkin message should include a hint that this
actually renders the test more or less useless wrt. the original bug; even
adding a comment to the original bug report about this should be
considered.


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

Comment By: Jan Rittinger (tsheyar)
Date: 2007-01-18 15:15

Message:
Logged In: YES
user_id=993208
Originator: NO

This bug occurs because the static typing has been improved in the current
branch while mps cannot cope with it.
(smaller example: doc("foo")[./@o cast as xs:integer?])

The reason is that a seqcast is treated like a normal cast in mps. In
comparison to a normal cast a seqcast should be used only during typing to
'assert' a certain type. The algebra thus ignores all seqcasts. Mps in
comparison however needs to interpret them as cast because it relies on the
typing information to refer to the correct value containers.

The only hack to 'avoid' the error message is to introduce a special case
for numeric where seqcasts are ignored. The introduced typeswitch up front
of already ensures that the type is integer instead of integer?. This
information however cannot be inferred as a generic type intersection is
impossible.

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

Comment By: Stefan Manegold (stmane)
Date: 2007-01-17 18:23

Message:
Logged In: YES
user_id=572415
Originator: YES

Jan,

since you changed the respective tests on Dec 07 2006 (only in the
development trunk; see below), you might want to have a look at this and
comment.
Feel free to assign to someone else, in case you think this is not your
bug.

Thanks!

===================================================================
2006/12/07 - tsheyar:
       
tests/BugTracker/Tests/complex-XQuery-batbat_lng_add_inplace-error-KO2.SF-1562868.xq,1.2
       
tests/BugTracker/Tests/complex-XQuery-batbat_lng_add_inplace-error-KO.SF-1562868.xq,1.2
       
tests/BugTracker/Tests/complex-XQuery-batbat_lng_add_inplace-error-OK.SF-1562868.xq,1.2
-- Fixed output/queries to cope with the last week introduced more
restrictive
   type tests in the cast operator
===================================================================


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

Comment By: Stefan Manegold (stmane)
Date: 2007-01-17 17:51

Message:
Logged In: YES
user_id=572415
Originator: YES

Since these tests do work in the XQuery_0-14 release branch, i.e., they
have (recently) been broken only in the development trunk, I consider this
bug report high priority.


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

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=482468&aid=1637867&group_id=56967

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Monetdb-bugs mailing list
Monetdb-bugs@...
https://lists.sourceforge.net/lists/listinfo/monetdb-bugs