Parsing Python JSON with multiple same strings with different values

Question

I am stuck on an issue where I am trying to parse for the id string in JSON that exists more than 1 time. I am using the requests library to pull json from an API. I am trying to retrieve all of the values of "id" but have only been able to successfully pull the one that I define. Example json:

{
"apps": [{
    "id": "app1",
    "id": "app2",
    "id": "new-app"
}]
}

So what I have done so far is turn the json response into dictionary so that I am actually parse the first iteration of "id". I have tried to create for loops but have been getting KeyError when trying to find string id or TypeError: list indices must be integers or slices, not str. The only thing that I have been able to do successfully is define which id locations to output.

(data['apps'][N]['id']) -> where N = 0, 1 or 2

This would work if there was only going to be 1 string of id at a time but will always be multiple and the location will change from time to time.

So how do return the values of all strings for "id" from this single json output? Full code below:

import requests
url = "http://x.x.x.x:8080/v2/apps/"
response = requests.get(url)

#Error if not 200 and exit
ifresponse.status_code!=200:
print("Status:", response.status_code, "CheckURL.Exiting")
exit()

#Turn response into a dict and parse for ids 
data = response.json()
for n in data:
    print(data['apps'][0]['id'])

OUTPUT:
app1

UPDATE: Was able to get resolution thanks to Robᵩ. Here is what I ended up using:

def list_hook(pairs):
result = {}
for name, value in pairs:
    if name == 'id':
        result.setdefault(name, []).append(value)
    print(value)

data = response.json(object_pairs_hook = list_hook)

Also The API that I posted as example is not a real API. It was just supposed to be a visual representation of what I was trying to achieve. I am actually using Mesosphere's Marathon API . Trying to build a python listener for port mapping containers.

Answer 1

Your best choice is to contact the author of the API and let him know that his data format is silly.

Your next-best choice is to modify the behavior of the the JSON parser by passing in a hook function. Something like this should work:

def list_hook(pairs):
    result = {}
    for name, value in pairs:
        if name == 'id':
            result.setdefault(name, []).append(value)
        else:
            result[name] = value
    return result

data = response.json(object_pairs_hook = list_hook)

for i in range(3):
    print(i, data['apps'][0]['id'][i])