User Defined Function doesn't exclude the possibility to include the function in Drill's code base, if the user chooses to contribute the function back as a patch of a pull request.
We generally suggest the creation of an UDF because it can be much faster than opening a JIRA and waiting for a community member to implement the function. You are right though as not all users will have the capabilities nor the time to do it. On Wed, Mar 9, 2016 at 2:11 PM, John Omernik <[email protected]> wrote: > Abdel - > > Is there a process or guideline for determining what functions should be in > Drill itself vs. developed by users? I.e. how is it determined if a > function should exist in the main Drill project vs. determining it should > be user developed? Does the Apache foundation have a process for "add-on" > projects? I think from a community perspective, the challenge with the "you > can always develop..." response (and please, I am not calling you out > directly, I've seen this response quite a bit) is that: > > A. It can be discouraging for non developer users of Drill. (I'd argue that > the target user base of Drill is going to be less than 25% Java Developers) > to hear this response. Leading some to halt their PoC or move on to other > projects that may support what they are looking for. > B. It will cause a lot of duplicate work (Company A develops string > aggregate function, Company B develops string aggregate function, User C > develops String Aggregate functions etc) to be done outside of the > community, wasting resources as companies need to develop these functions > over and over again, and also keeping some users of Drill in the dark about > potential functions they could use. (I for one expand my knowledge of data > by reviewing functions and what they do) > > So to my question about add on projects. Could a sub-project under the > Drill project be run where functions can be talked about at a community > level? Releasing code that can be added to drill, high quality functions, > that operate well, that "put back" into community? From an Apache > standpoint would this be a separate project? I see so many wonderful ways > that this could be a huge user base driver, excitement driver, and > functionality driver for the Drill project. Wish lists, challenges, new > functionality can be developed openly, and maybe (depending on the answer > to my first question) some of these could even make it into the Drill > project itself. Almost like a developer preview of self contained > functions. > > What are the community thoughts on something like this? > > John > > On Tue, Mar 8, 2016 at 11:43 PM, Abdel Hakim Deneche < > [email protected]> > wrote: > > > You can always develop a User Defined Aggregate Function: > > > > http://drill.apache.org/docs/develop-custom-functions/ > > > > Thanks > > > > On Wed, Mar 9, 2016 at 12:29 AM, Bosung Seo <[email protected]> > > wrote: > > > > > Hello, > > > > > > I found that Drill doesn't support string_agg function yet. > > > Is there another way to query as the string_agg function? > > > > > > If I have a table, > > > user | id > > > user1 | 1 > > > user1 | 2 > > > user1 | 3 > > > user2 | 1 > > > > > > I want to make like this. > > > user | ids > > > user1 | 1,2,3 > > > user2 | 1 > > > > > > Any help would be appreciated. > > > > > > Thanks, > > > Bo > > > > > > > > > > > -- > > > > Abdelhakim Deneche > > > > Software Engineer > > > > <http://www.mapr.com/> > > > > > > Now Available - Free Hadoop On-Demand Training > > < > > > http://www.mapr.com/training?utm_source=Email&utm_medium=Signature&utm_campaign=Free%20available > > > > > > -- Abdelhakim Deneche Software Engineer <http://www.mapr.com/> Now Available - Free Hadoop On-Demand Training <http://www.mapr.com/training?utm_source=Email&utm_medium=Signature&utm_campaign=Free%20available>
