David G Johnston wrote
>
> David G Johnston wrote
>>
>> mozsar.laszlo wrote
>>> The following bug has been logged on the website:
>>>
>>> Bug reference: 11130
>>> Logged by: Laszlo Mozsar
>>> Email address:
>>> mozsar.laszlo@
>>> PostgreSQL version: 9.3.4
>>> Operating system: Ubuntu 14.04 LTS
>>> Description:
>>>
>>> The "is null" contition on a date field returns wrong result. I can't
>>> reproduce this with a test table. My tables contains personal and
>>> business
>>> informations, so I can't send a full log with datas.
>>> (I used reindex and vacuum before submitting this report.)
>> Cannot test myself at the moment but is it that your attempt at a test
>> case didn't fail or that you didn't attempt a self-contained test case?
>>
>> If, on the same database, you cannot reproduce the error with a minimal
>> test case then something about the specific tables and queries is wrong
>> and you (since you cannot share) need to figure out what that is. Even
>> if you cannot share the data sharing the schema will let others try to
>> see whatever it is you may be missing.
>>
>> David J.
> And yes I see you shared a schema but nothing about it looks wrong so with
> additional info is needed or you are showing what you think is the schema
> while what is actually in production is different. There are only a
> couple of relevant fields so if you can even so a subset of the data that
> would probably help.
>
> David J.
Actually, upon further review...
I suspect your problem is with "lajarat_tipus"
This is a char(1) field that you likely are interpreting as null what is
actually a blank/space and so the first when is false as is the second (not
L) and so the else matches.
I do not use char so am not that knowledgable as to its nuances with respect
to padding spaces.
David J.
--
View this message in context:
http://postgresql.1045698.n5.nabble.com/BUG-11130-Case-condition-error-tp5814246p5814308.html
Sent from the PostgreSQL - bugs mailing list archive at Nabble.com.