LOOKUP shouldn't duplicate the output if the same field was already present in the input

[astefan]

Description

This is a follow up to LOOKUP work where, if there is a field with a name identical to the one LOOKUP introduces, both of them appear in the results. We need to be consistent here and:

• only one of the fields should be in the results
• that field should be the one introduced by LOOKUP (same approach is being used by the ENRICH command)

[elasticsearchmachine]

Pinging @elastic/es-analytical-engine (Team:Analytics)

[nik9000]

I don't believe both appear in the results:

// Makes sure the LOOKUP squashes previous names 
doesNotDuplicateNames
required_capability: lookup
FROM employees
| SORT emp_no
| LIMIT 4
| RENAME languages.long AS long
| EVAL name = CONCAT(first_name, " ", last_name)
| LOOKUP long_number_names ON long
| RENAME long AS languages
| KEEP emp_no, languages, name
;

emp_no:integer | languages:long | name:keyword
         10001 |              2 | two
         10002 |              5 | five
         10003 |              4 | four
         10004 |              5 | five
;

At least, not in the code as it stands as of yesterday. The LOOKUP result wins. Now, the column pruning happens too late which is a problem, but the output presently looks ok.

Also, I might be doing it in a weird way. The QL rules are indeed complex.

[astefan]

@nik9000 the query I was looking at few hours ago is much simpler:

{
    "query": "ROW int = 5, name = 123 | LOOKUP int_number_names ON int",
    "tables": {"int_number_names": {"int:integer": [0,1,2,3,4,5,6,7,8,9,10], "name:keyword": ["zero","one","two","three","four","five","six","seven","eight","nine","ten"]}}
}

The result I got:

      int      |     name      |     name      
---------------+---------------+---------------
5              |123            |five           

[nik9000]

neat!