fetch the number of record from a JSON file using shell - shell

Question

I have a test.txt file in this format
{
"user": "sthapa",
"ticket": "LIN-5867_3",
"start_date": "2018-03-16",
"end_date": "2018-03-16",
"demo_nos": [692],
"service_names": [
"service1",
"service2",
"service3",
"service4",
"service5",
"service6",
"service7",
"service8",
"service9"
]
}
I need to look for a tag called demo_nos and provide the count of it.
For example in the above file "demo_nos": [692] which means only one demo nos...similarly if it had "demo_nos": [692,300] then the count would be 2
so what shell script can i write to fetch and print the count?
The output should say the demo nos = 1 or 2 depending on the values inside the tag [].
i.e I have a variable in my shell script called market_nos which should give me it's count

Answer 1

The gold standard for manipulating JSON data from the command line is jq:
$ jq '.demo_nos | length' test.txt
1
.demo_nos returns the value associated with the demo_nos key in the object, and that array is piped to the length function which does the obvious.

Answer 2

I'm assuming you have python and the file is JSON :)
$ cat some.json
{
"user": "sthapa",
"ticket": "LIN-5867_3",
"start_date": "2018-03-16",
"end_date": "2018-03-16",
"demo_nos": [692],
"service_names": [
"service1",
"service2",
"service3",
"service4",
"service5",
"service6",
"service7",
"service8",
"service9"
]
}
$ python -c 'import sys,json; print(len(json.load(sys.stdin)["demo_nos"]))' < some.json
1

Answer 3

Not the most elegant solution but this should do it
cat test.txt | grep -o -P 'demo_nos.{0,200}' | cut -d'[' -f2 | cut -d']' -f1 | awk -F',' '{ print NF }'
Please note that this is a quick and dirty solution treating input as raw text, and not taking into account JSON structure. In exceptional cases were "demo_nos" string would also appear elsewhere in the file, the output from the command above might be incorrect.