Hi,

So far only the query plan is present in the code. What are the missing part from the parser to the query plan and from the plan to the execution?

Is there a place where I can find that flow? Is it possible to refer me to another patch that can describe what components have to be modified and added?

Thanks,


On 2019/01/08 20:18:25, James Taylor wrote:
> The support for UNNEST was in the calcite branch and took advantage of the> > compilation support there. We unfortunately don't have the same support in> > the mainline branch. We'd welcome a contribution in this area, but it'd be>
> a fair amount of work.>
> Thanks,>
> James>
>
> On Tue, Jan 8, 2019 at 11:07 AM Thomas D'Silva >
> wrote:>
>
> > Looks like UNNEST array support was added in (>
> > https://issues.apache.org/jira/browse/PHOENIX-953).>
> > I'm not sure what remaining work PHOENIX-4311 requires. There is a>
> > test UnnestArrayIT that was disabled,>
> > perhaps as a first step you can try running the test and see the current>
> > state of things.>
> >>
> > On Tue, Jan 8, 2019 at 10:06 AM Josh Elser wrote:>
> >>
> > > Hi,>
> > >>
> > > Looking at PHOENIX-4311, I think any contribution you choose to make>
> > > would be appreciated. I don't have a good understanding of the>
> > > difficulties you would have in trying to implement it.>
> > >>
> > > In general, if there is no assignee and there isn't discussion on the> > > > Jira issue, you may feel free to request it be assigned to yourself to>
> > > work on.>
> > >>
> > > - Josh>
> > >>
> > > On 1/8/19 9:38 AM, tricolor wrote:>
> > > > Hi,>
> > > >>
> > > > I am currently working on a project where the function UNNEST would be> > > > > useful to rotate an array in order to perform an aggregation. According>
> > > to>
> > > > the Jira ticket 4311 there are pending tests to surface this feature.>
> > The>
> > > > ticket was untouched since October 2017. My questions are the>
> > following :>
> > > >>
> > > > 1. Is it a feature that will eventually be implemented?>
> > > >>
> > > > 2. Is there any blocker making its implementation impossible at the>
> > > moment?>
> > > >>
> > > > 3. If there is no blocker what are the missing components to implement>
> > > it?>
> > > >>
> > > > 4. Is anyone going to implement it eventually? If someone takes it what> > > > > would be an expected time of delivery? If no one implements it, would>
> > my>
> > > > contribution be appreciated?>
> > > >>
> > > >>
> > > > Best regards,>
> > > >>
> > >>
> >>
>

Reply via email to