I have two jsons as follows: json1:This is the payload I am submitting to an API

    {
        "id": 0,
        "object_type": "SCHEMA",
        "object_meta": {
            "source_id": 1002,
            "database": "raw_exadata",
            "schema": "clv"
                       },
        "business_name": "",
        "business_logic": "",
        "verified_use_cases": ["p"],
        "classifications": ["bb464f04-f879-361c-95e5-42bcca6a9535"]
    }
,
    {
        "id": 1,
        "object_type": "TABLE",
        "object_meta": {
            "source_id": 1002,
            "database": "raw",
            "schema": "clv",
            "table": "clv_cust"
                       },
       
        "business_name": "",
        "business_logic": "",
        "verified_use_cases": [],
        "classifications": ["p"],
       
    }
]

json2: this is the response of an API call which tells me the error in the data that was posted

{
   "errors": {
      "0": {
         "verified_use_cases": {
            "0": [
               "Not a valid UUID."
            ]
         }
      },
      "1": {
         "classifications": {
            "0": [
               "Not a valid UUID."
            ]
         }
      }
   }
}

I want to produce a very user friendly response to the posted payload stating such and such an object has an invalid entry hence it has failed.

For the above, I need to map (or lookup) on the "id" from json2 to json1 and fetch the object_type and also that the entry in the verified_use_cases, classifications is wrong with corresponding id.

What I have tried so far:

from collections import defaultdict

p=json.loads(response.text)
result=defaultdict(list)
for i in p['errors']:
    result['obj_id'].append(int(i))
    result['error'].append(p['errors'][i])

Now this gives me the object_id, but I am not sure how I can map it further with json1.

Kindly help.

PS: the fields, verified use cases, classifications are all multi valued with comma separated actually, hence you can see the "0" int the error json

Thanks in advance

CodePudding user response：

You should consider a couple of things in mind:

1. When you are working with APIs, make sure that you use some kind of validation or preprocesses, I suggest you to use Pydantic with it's validators and easiest ways of converting any json to necessary class.
2. Example of using pydantic with your code, example will be below.
3. Taking look at docs, here it is: https://pydantic-docs.helpmanual.io/

And as a result when you'd like to create any of these defined classes you can do:

parse_obj_as(ApiObject, response.json)

Example:

class ApiObject(BaseModel):
    id: int = Field(alias="id")
    object_type: str = Field(alias="object_type")
    object_meta: ObjectMeta = Field(alias="object_meta")  # Same class with it's defined fields
    business_name: str = Field(alias="business_name")
    business_logic: str = Field(alias="business_logic")
    verified_use_cases: List[str] = Field(alias="verified_use_cases")
    classifications: List[str] = Field(alias="classification")

CodePudding user response：

json1 = [{
        "id": 0,
        "object_type": "SCHEMA",
        "object_meta": {
            "source_id": 1002,
            "database": "raw_exadata",
            "schema": "clv"
                       },
        "business_name": "",
        "business_logic": "",
        "verified_use_cases": ["p"],
        "classifications": ["bb464f04-f879-361c-95e5-42bcca6a9535"]
    }
,
    {
        "id": 1,
        "object_type": "TABLE",
        "object_meta": {
            "source_id": 1002,
            "database": "raw",
            "schema": "clv",
            "table": "clv_cust"
                       },
       
        "business_name": "",
        "business_logic": "",
        "verified_use_cases": [],
        "classifications": ["p"],
       
    }
]

json2 = {
   "errors": {
      "0": {
         "verified_use_cases": {
            "0": [
               "Not a valid UUID."
            ]
         }
      },
      "1": {
         "classifications": {
            "0": [
               "Not a valid UUID."
            ]
         }
      }
   }
}

resp = {}

for record in json1:
    key = str(record["id"])
    if key in json2["errors"]:
        record['errors'] = json2["errors"][key]#adding json2
        resp[key] = record
    

Response

print(json.dumps(resp))
    {
        "0": {
            "id": 0,
            "object_type": "SCHEMA",
            "object_meta": {
                "source_id": 1002,
                "database": "raw_exadata",
                "schema": "clv"
            },
            "business_name": "",
            "business_logic": "",
            "verified_use_cases ": ["p "],
            "classifications ": ["bb464f04 - f879 - 361 c - 95e5 - 42 bcca6a9535 "],
            "errors ": {
                "verified_use_cases ": {
                    "0 ": ["Not a valid UUID."]
                }
            }
        },
        "1 ": {
            "id ": 1,
            "object_type ": "TABLE ",
            "object_meta ": {
                "source_id ": 1002,
                "database ": "raw ",
                "schema ": "clv ",
                "table ": "clv_cust "
            },
            "business_name ": "",
            "business_logic ": "",
            "verified_use_cases ": [],
            "classifications": ["p"],
            "errors": {
                "classifications": {
                    "0": ["Not a valid UUID."]
                }
            }
        }
    }