使用jq从多个输入对象生成单个JSON数组

Question

I do have several similar JSON objects (eg in files):

file_1.json: {"myArray":[{a},{b}]}
...
file_n.json: {"myArray":[{n},{m}]}

I would like to transform all these files into one output with the following format, using jq and Linux or Windows command line tools:

result_file.json: [{a},{b},...,{n},{m}]

So, I just want one array having all the objects that can be found in the "myArray" arrays from the different inputs.

Currently I use this command, which separates the objects, but doesn't create an array that includes them:

type file_1.json file_n.json | jq ".result[]" > result_file.json

(replace type with cat on Linux)

How can I get the format right?

Answer 1

You can "slurp" your inputs:

-s read (slurp) all inputs into an array; apply filter to it;

(man jq)

Like that:

1.json

{"myArray":[1,2,3]}

2.json

{"myArray":[3,4,5]}

%jq -s '[.[].myArray[]]' 1.json 2.json

[
  1,
  2,
  3,
  3,
  4,
  5
]