jq
Interactive query using jq + fzf
cat /tmp/1.json | jqfzf
Assign to variable in a pipeline
The as keyword can be used to assign a value to a variable for later use in a
pipeline.
As a contrived example, assume a JSON document of the following schema:
[
{
"sender_id": 1,
"chat_participants": [
{"user_id": 1},
{"user_id": 2}
]
}
]
The order of chat_participants is not fixed, meaning to say that the sender
isn’t necessarily always the first element. We want to find the user IDs of all
the recipients, not the sender. We can do:
.[] | .sender_id as $sender_id | .chat_participants.user_id | select(. != $sender_id)
Construct a JSON object by referencing a nested field
Given the following source JSON:
[{
"timestamp": 1,
"labels": {
"method": "xxx"
},
"payload": "text"
}]
Constructing a JSON object would normally take the form of .[] | {payload}, which gives us:
{"payload": "text"}
If we wanted to select the .labels.method field, naturally we might expect jq
syntax to be {labels.method}. This does not work. Instead the syntax is .[] | {"method": .labels.method}
So with this contrived example, if we wanted to select timestamp and the
method, the jq filter would be .[] | {timestamp, "method": .labels.method}, which gives us:
{"timestamp": 1, "method": "xxx"}
Search for value at an arbitrarily deep nested key
.. | objects | select(.id == "x")