I think what you are saying is that basically you need to qualify columns 
if you think Dataset#qualify will happen later in the chain and you don't 
want it to change the current columns...

Sorry for the bad example, by "computed" I mean simply any computed column 
like COUNT(*) as "count", doesn't really matter.

On Wednesday, June 24, 2020 at 9:31:10 PM UTC+3, Jeremy Evans wrote:
>
> On Wednesday, June 24, 2020 at 10:55:18 AM UTC-7, Aryk Grosz wrote:
>>
>> I have a sequel pagination library, that runs "qualify" and changes 
>> previous conditions earlier up in the chain causing the sql too fail.
>>
>> What is the best practice here?
>>
>> Is there a way for me to do:
>>
>> User.select(<computed>.as(:is_nearby)).order(:is_nearby)
>>
>> and *not* be worried about "qualify" being run later in the chain and 
>> making the sql no longer valid.
>>
>> I've tried:
>>
>> User.select(<computed>.as(:is_nearby)).order(Sequel[:is_nearby]).qualify
>>
>> User.select(<computed>.as(:is_nearby)).order(Sequel[:is_nearby].as(:is_nearby)).qualify
>>
>> #qualify is still applying a condition on it.
>>
>
> I'm not sure what exact problem you are running into, as you don't specify 
> what <computed> is.  Making sure your examples are self contained before 
> posting them makes it easier to provide support.
>
> If you don't want #qualify to qualify an identifier, the identifier needs 
> to be already qualified, or you can use literal SQL instead of an 
> identifier.
>
> An alternative approach would be not using #qualify, and manually 
> qualifying identifiers that would be ambiguous.
>
> Thanks,
> Jeremy
>

-- 
You received this message because you are subscribed to the Google Groups 
"sequel-talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/sequel-talk/ba423d4b-4f6c-42c9-89c5-2927af24aa14o%40googlegroups.com.

Reply via email to