awk（或sed / grep）获取子字符串的出现

Question

I have a json string in a bash variable, which is something like this:

{
    "items": [
      {
        "foo": null,
        "timestamp": 1553703000,
        "bar": 123
      },
      {
        "foo": null,
        "timestamp": 1553703200,
        "bar": 456
      },
      {
        "foo": null,
        "timestamp": 1553703400,
        "bar": 789
      }
    ]
}

I want to know how many of those timestamp s are after a given datetime, so if I have 1553703100 it'll return 2 .

(Bonus imaginary points if you can get me just that number!)

As a step towards that, I want to get just the matches of "timestamp": \\d+, in the string so that I can loop through them in a bash script.

I've used sed and grep a bit, but never used awk, and from my reading it seems like that might be the better match for the task.

Other info: - The json is already pretty-printed, as above, so the timestamps would always be on separate lines. - This is to run in Cygwin, so I have awk/gawk, sed, and grep/egrep, but probably not others. - Could be any number of timestamps in the json.

Answer 1

You didn't provide the expected output so it's a guess but is this what you're trying to do?

$ echo "$var" | jq '.items[].timestamp'
1553703000
1553703200
1553703400

or maybe:

$ echo "$var" | jq '.items[].timestamp | select(. > 1553703100)'
1553703200
1553703400

or:

$ echo "$var" | jq '[.items[].timestamp | select(. > 1553703100)] | length'
2

WARNING: I'm just learning jq so there may be better ways to do the above!

Answer 2

edit: The second approach listed below has serious problems that were very helpfully outlined by @EdMorton. I've elected to keep the old code for educational purposes.

Avoided substr() and caught null string i :

$ awk -v dt=1553703100 '
  /timestamp/ && $2+0>dt {i++}
  END {print i+0}
' <<< "$var"

2

---

WARNING: PROBLEMATIC CODE

Here I used substr(string, index, [characters]) to trim the comma off your second field. The /timestamp/ regex is not complex; it could be improved if your json became more intricate.

$ awk -v dt=1553703100 '
  /timestamp/ && substr($2, 0, length($2)) > dt {i++} 
  END {print i}
' <<< "$var"

2

Answer 3

You can also implement quickly a python solution:

input :

$ cat data.json 
{
    "items": [
      {
        "foo": null,
        "timestamp": 1553703000,
        "bar": 123
      },
      {
        "foo": null,
        "timestamp": 1553703200,
        "bar": 456
      },
      {
        "foo": null,
        "timestamp": 1553703400,
        "bar": 789
      }
    ]
}

code :

$ cat extract_value2.py 
import json

tLimit = 1553703100
with open('data.json') as f:
    data = json.load(f)
    print([t['timestamp'] for t in data["items"] if t['timestamp'] > tLimit])

output :

$ python extract_value2.py 
[1553703200, 1553703400]

count code:

$ cat extract_value2.py 
import json

tLimit = 1553703100
with open('data.json') as f:
    data = json.load(f)
    print(len([t['timestamp'] for t in data["items"] if t['timestamp'] > tLimit]))

output :

$ python extract_value2.py
2