Not able to combine two column as Struct Type in hive

[#1]

Dear All.

i having some issue with combining two columns fields into single fields as STruct type.

below code i tried but still getting the same issue?

create table dummy_TBL2 (id int,fname string ,lname string) ;
insert into dummy_TBL2 (id,fname,lname) values(1,'bhau','anna') ; //dummy side
now below show the struct table:-
create table data_TBL2 (id int, name struct<fname:string,lname:string>) 
ROW FORMAT DELIMITED 
FIELDS TERMINATED BY ','  
STORED AS TEXTFILE;//in this table i took name as struct type
___inserting the data from dummy into struct type-----
insert into table data_TBL2 
select id,
      name(fname,lname) 
from dummy_TBL2 limit 1;

Thanks

HadoopHelp

[#2]

@HadoopHelp ,

You should use named_struct function to turn those strings into struct, please refer to Hive's UDF documentation:
Complex Type Constructors

So your query should be :

insert into table data_TBL2 
select id,
named_struct('fname', fname, 'lname', lname) 
from dummy_TBL2 limit 1;

Cheers

Eric

[#3]

Hi @EricL .

Thank you very much...

but the same use case is here:-

this is my output data structure from a JSON we required:-

{
    "name": [
        {
            "use": "official", //here "tab1.use" is column and value
            "family": "family",//here "tab1.family" is column and value
            "given": [    //this column we need to create and add value from "tab1.fn&ln"
                "first1",   //here "first1" is coming from tab1.fname
                "last1"      //here "last1"is coming from tab1.lname
            ]
        },
        {
            "use": "usual",  //here "tab2.use" is column and value
            "given": [ //here we need to create column with fn&ln
                "first1 last1"   //here "first1 last1" is coming from tab1.fname &tab1.lname
            ]
        }
    ]
}

here we want to create a column(name) from above columns :-
above data is JSON structure but i want in Hive with table columns.
then further we can convert the same into JSON in my use cases.

Note :- structure is matter here.


Thanks
HadoopHelp