I think you are using an old build and running into https://issues.apache.org/jira/browse/DRILL-1195
Could you please retry with the Drill rc1 build [1]. I used your sample data and query against it and seems to work. 1. http://people.apache.org/~jacques/apache-drill-0.4.0.rc1/ On Thu, Jul 31, 2014 at 11:45 AM, Andy Pernsteiner < [email protected]> wrote: > Hello, I'm currently attempting to perform queries against a JSON file > which has some nested elements, some of which are arrays. Here are 2 > records worth of the file: > > > > {"trans_id":16,"date":"06/08/2012","time":"07:09:25","user_info":{"cust_id":6,"device":"IOS5","state":"tx"},"marketing_info":{"camp_id":17,"keywords":[]},"trans_info":{"prod_id":[65,1233,182,70,90,12,1520,1563,2144],"purch_flag":"false"}} > > > {"trans_id":17,"date":"07/31/2013","time":"16:22:26","user_info":{"cust_id":9,"device":"IOS5","state":"oh"},"marketing_info":{"camp_id":1,"keywords":["careful","us","around","you","whatever","i","how"]},"trans_info":{"prod_id":[10,2597,2658,2202,999],"purch_flag":"false"}} > > > > Note that the "prod_id" and "keywords" fields contain arrays. Here's an > example query without trying to specify the array index: > > 0: jdbc:drill:> select t.trans_info.prod_id from > dfs.`/myfiles/JSON/mobile.json` t limit 3; > > +------------+ > | EXPR$0 | > +------------+ > | [1435,673] | > | [785] | > | [26] | > +------------+ > 3 rows selected (0.507 seconds) > > > So, I can see the array elements just fine, but when I try to access a > specific one I come up w/ nulls: > > > 0: jdbc:drill:> select t.trans_info.prod_id[0] from > `dfs`.`/myfiles/JSON/mobile.json` t limit 10; > +------------+ > | EXPR$0 | > +------------+ > | null | > | null | > | null | > | null | > | null | > | null | > | null | > | null | > | null | > | null | > +------------+ > 10 rows selected (0.303 seconds) > > > Am I using the correct notation/subscript syntax here? > > Thanks! > > > -- > Andy Pernsteiner > Manager, Field Enablement > ph: 206.228.0737 > > www.mapr.com >
