id	summary	reporter	owner	description	type	status	component	version	severity	resolution	keywords	cc	stage	has_patch	needs_docs	needs_tests	needs_better_patch	easy	ui_ux
17273	Wrong slice on QuerySet when using Oracle database	rodolfo.3+django@…	nobody	"I have a model like this:
{{{
class MyTest(models.Model):
    afield = models.CharField(max_length=20, null=True, blank=True)
}}}

and populate it with this code:
{{{
>>> for i in xrange(10):
...     MyTest(afield=None).save()
...     MyTest(afield=str(i)).save()
}}}

And I use the field to query, sorting by the field:
{{{
>>> [i.pk for i in MyTest.objects.all().order_by('afield')]
[2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 13, 15, 17, 19, 11, 9, 1, 5, 3, 7]
}}}

But, if I slice in the queryset, the results will be wrong:

{{{
>>> [i.pk for i in MyTest.objects.all().order_by('afield')[10:15]]
[1, 15, 13, 11, 9]
>>> [i.pk for i in MyTest.objects.all().order_by('afield')[15:20]]
[11, 9, 7, 5, 3]
}}}

Please, note the objects with pk = 11 and 9 are repeated into the 2 different queries. This happen because Oracle first do a ""where"" and then sort the items. I solve it changing 2 lines into the backend (I'll attach a diff file). With this patch applied, the results are ok:
{{{
>>> [i.pk for i in MyTest.objects.all().order_by('afield')]
[2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 13, 15, 17, 19, 11, 9, 1, 5, 3, 7]
>>> [i.pk for i in MyTest.objects.all().order_by('afield')[10:15]]
[13, 15, 17, 19, 11]
>>> [i.pk for i in MyTest.objects.all().order_by('afield')[15:20]]
[9, 1, 5, 3, 7]
}}}"	Bug	closed	Database layer (models, ORM)	dev	Normal	invalid			Unreviewed	0	0	0	0	0	0