How to nicely remove empty array in jq

Question

I am reading O'Reilly's new "Data Science at the Command Line" and running into trouble using jq. I have some JSON (returned from the NYTimes Articles API) I am parsing with jq like so:

jq -c \ '[.response.docs[] | {date: .pub_date, type: .document_type, title: .headline.main }]' \ < myjsonfile.json 

So, I am looking for "response":"docs" (which is an array) and then matching every item in that array with "pub_type" etc., renaming it, and so on. This works great, but it adds an empty array at the end:

[{"date":"2009-01-02T00:00:00Z","type":"article","title":"SPARE TIMES: AROUND TOWN"},   {"date":"2009-01-02T00:00:00Z","type":"article","title":"Catskill Home Prices: How Low Will They Go?"}, {"date":"2009-01-01T00:00:00Z","type":"article","title":"Ominous Cutbacks At Chanel"}] []  

How do I get rid of the empty array? My solution right now is to pipe the output back into jq, but that feels really suboptimal. So this works:

jq -c \ '[.response.docs[] | {date: .pub_date, type: .document_type, title: .headline.main }]' | \ < myjsonfile.json |  jq 'if length > 0 then . else empty end' 

But that feels ugly. Is there a nicer way of doing this?

Answer 1

Use the select filter where length > 0.

select(length > 0)