Hacker News Reader

Comment by vitorbaptistaa | Hacker News Reader

Using the excellent jq [1] command

  wget https://gist.githubusercontent.com/jorin-vogel/7f19ce95a9a842956358/raw/e319340c2f6691f9cc8d8cc57ed532b5093e3619/data.json
  echo 'Name,Credit Card' > 20150425.csv
  jq -r '.[] | [.name, .creditcard] | @csv' data.json >> 20150425.csv

This consider that all timestamps are from 2015/04/25 (which is the case with the example dataset, but might not be).

[1] https://stedolan.github.io/jq/

View on HN

You forgot the select(.creditcard) - see my comment on the gist - as this one does not filter out null'ed creditcards:

  name=`date +'%Y%m%d'`; 
  echo "name,creditcard" > $name; 
  jq -r '.[] | select(.creditcard) | [.name, .creditcard] | @csv' data.json >> $name

You're right! I missed that requirement.

If we're going for succinctness, you could combine these all together:

(echo '"Name","Credit Card"'; jq -r '.[] | select(.creditcard) | [.name, .creditcard] | @csv' data.json) >> `date +%Y%m%d`.csv

Nice trick to write the headers, thanks!

I used jq as well, though instead of doing a seperate echo I did this:

    ([{name: "name", creditcard: "creditcard"}] + .)[]

instead of .[]

Another less hacky way would be:

    [["name", "creditcard"]] + map(select(.creditcard) | [.name, .creditcard]) | .[] | @csv

Reply on news.ycombinator.com