使用“jq”从另一个 json 文件附加到 json 文件中的数组

Question

i have a file "stats.json" with the following initial structure"

[{
"price": 66.55,
"created": "2018-02-24T14:32:57"
}]

With a bash/curl script i take every 10 minutes data from an api and save it to "temp.json"

{
"price": 88.55,
"created": "2018-02-24T15:32:57"
}

I would like to merge the temp.json (which is updated every 10min) and to fill the stats.json file. I tried to do it with "jq" but without success.

Answer 1

One way to add the item in temp.obj to the array in stats.json:

jq —-slurpfile temp temp.obj '. + $temp' stats.json

Another way:

jq —s '.[0] + [.[1]]' stats.json temp.json

(If you want to overwrite stats.json, then I'd suggest using sponge as illustrated below.)

However, if at all possible, it would almost certainly be better to maintain stats.json as a log file in JSONL format: http://jsonlines.org/

That is, each object in temp.json would be appended as a single line to stats.json:

jq -c temp.json >> stats.json

This of course assumes that stats.json is initially empty or already in the JSONL format. To convert your existing stats.json to JSONL, you could do something like this:

cp -p stats.json stats.json.bak

jq -c .[] stats.json | sponge stats.json