对大查询表的API请求

Question

I'm facing an issue, regarding a nice way to modify JSON file by using python

JSON REQUEST:

{
    "reports": [{
            "data": {
                "rows": [{
                        "metrics": [{
                                "values": ["27.8", "4", "4", "6.95", "1.0", "0.0", "3.8834951456310676"]
                            }
                        ],
                        "dimensions": ["TEST1", "20180725"]
                    }, {
                        "metrics": [{
                                "values": ["75.0", "12", "12", "6.25", "1.0", "0.0", "3.4782608695652173"]
                            }
                        ],
                        "dimensions": ["TEST2", "20180725"]
                    }
                ],
                "maximums": [{
                        "values": ["1665.0", "140", "126", "65.0", "3.0", "0.0", "50.0"]
                    }
                ],
                "minimums": [{
                        "values": ["0.0", "0", "0", "0.0", "0.0", "0.0", "0.0"]
                    }
                ],
                "isDataGolden": true,
                "totals": [{
                        "values": ["27045.99", "3274", "2831", "8.260839951130116", "1.1564818085482163", "0.0", "4.949387227049424"]
                    }
                ],
                "rowCount": 358
            },
            "columnHeader": {
                "dimensions": ["ga:productName", "ga:date"],
                "metricHeader": {
                    "metricHeaderEntries": [{
                            "type": "CURRENCY",
                            "name": "ga:itemRevenue"
                        }, {
                            "type": "INTEGER",
                            "name": "ga:itemQuantity"
                        }, {
                            "type": "INTEGER",
                            "name": "ga:uniquePurchases"
                        }, {
                            "type": "CURRENCY",
                            "name": "ga:revenuePerItem"
                        }, {
                            "type": "FLOAT",
                            "name": "ga:itemsPerPurchase"
                        }, {
                            "type": "CURRENCY",
                            "name": "ga:productRefundAmount"
                        }, {
                            "type": "PERCENT",
                            "name": "ga:buyToDetailRate"
                        }
                    ]
                }
            }
        }
    ]
}

LOOKING FOR:

Values in matrics based on "dimensions" and "metricHeaderEntries"

What is the clean way to modify report (or recreate it) so that I will have

LINE1 - {"ga:productName": "NAME","ga:date": "NAME","ga:itemRevenue": "value1", "ga:itemQuantity": "value2", ... }
LINE2 - {"ga:productName": "NAME","ga:date": "NAME","ga:itemRevenue": "value1", "ga:itemQuantity": "value2", ... }

EDIT1:

{
"ga:productName": "NAME", #from dimension 
"ga:date": "NAME", #from dimension 
"ga:itemRevenue": "value1", #from metricHeaderEntries 
"ga:itemQuantity": "value2", #from metricHeaderEntries 
... 
}
{
"ga:productName": "NAME2", #from dimension 
"ga:date": "NAME2", #from dimension 
"ga:itemRevenue": "value3", #from metricHeaderEntries 
"ga:itemQuantity": "value4", #from metricHeaderEntries 
... 
}

Value working this way:

"metrics": [{"values": ["27.8", "4", "4", "6.95", "1.0", "0.0","3.8834951456310676"] #headers in metricHeaderEntries 
"dimensions": ["TEST1", "20180725"] #header in dimension 

Or similar (I'm not interested in totals and so on)

Looking for solution/sample/explanation how to do it, with the way that BQ will accept it.

EXTRA:

I understand the way of getting data from JSON request like:

responce[][][]

But this situation is too tricky for me (

SAMPLE:

This is Ideally how the table should look like THIS IS HOW IDEALLY TABLE SHOULD LOOK LIKE

This is what google offer in the way of printing this data ( but need to convert it to the format which I explain above

def print_response(response):
  for report in response.get('reports', []):
    columnHeader = report.get('columnHeader', {})
    dimensionHeaders = columnHeader.get('dimensions', [])
    metricHeaders = columnHeader.get('metricHeader', {}).get('metricHeaderEntries', [])

    for row in report.get('data', {}).get('rows', []):
      dimensions = row.get('dimensions', [])
      dateRangeValues = row.get('metrics', [])

      for header, dimension in zip(dimensionHeaders, dimensions):
        print header + ': ' + dimension

      for i, values in enumerate(dateRangeValues):
        print 'Date range: ' + str(i)
        for metricHeader, value in zip(metricHeaders, values.get('values')):
          print metricHeader.get('name') + ': ' + value

Answer 1

So your edit still don't match your JSON; in dimension you have a list of values not a key:value

"dimensions": [
          "ga:productName",
          "ga:date"
        ],

this mean that you don't have any value to take so your example is not correct. In "metricHeaderEntries" you have the following:

"metricHeader": {
          "metricHeaderEntries": [
            {
              "type": "CURRENCY",
              "name": "ga:itemRevenue"
            },
            {
              "type": "INTEGER",
              "name": "ga:itemQuantity"
            },
            {
              "type": "INTEGER",
              "name": "ga:uniquePurchases"
            },
            {
              "type": "CURRENCY",
              "name": "ga:revenuePerItem"
            },
            {
              "type": "FLOAT",
              "name": "ga:itemsPerPurchase"
            },
            {
              "type": "CURRENCY",
              "name": "ga:productRefundAmount"
            },
            {
              "type": "PERCENT",
              "name": "ga:buyToDetailRate"
            }
          ]
        }

So even this case doesn't match your example because under "metricHeaderEntries" you don't have any values of "ga:itemRevenue" or "ga:itemQuantity" that you show in your example.

In any case you can go true the JSON in the same way of a python dictionary so you can select the elements by key in case of dictionary and by index in case of list.

If I get some time I'll try to solve your problem taking the values from your example even if the nodes gave by you are not the correct one.

ANSWER:

I solved your problem even if I hard-coded the key values instead than take from the original JSON just for let you understand how it works; please let me know if it's what you are expecting:

new_list=[]

l=a["reports"][0]["data"]["rows"]#get to "rows" key from a, where a is your JSON readed as dictionary
for i in l:#iterate rows key for search the needed values for each lines
    dict_line={}#create a dictionary for each line
    dict_line["ga:productName"]=i["dimensions"][0]#add to the dictionary dict_lineth key ga:productName and the product name as value
    dict_line["ga:date"]=i["dimensions"][1]#add to the dictionary dict_lineth key ga:date and the product date as value
    j= (i["metrics"][0]['values'])#for each product line I create a key node and value
    dict_line["ga:itemRevenue"]=j[0]
    dict_line["ga:itemQuantity"]=j[1]
    dict_line["ga:uniquePurchases"]=j[2]
    dict_line["ga:revenuePerIte"]=j[3]
    dict_line["ga:itemsPerPurchase"]=j[4]
    dict_line["ga:productRefundAmount"]=j[5]
    dict_line["ga:buyToDetailRate"]=j[6]
    new_list.append(dict_line)

print (new_list)

this is the result:

[
  {
    "ga:productName": "TEST1",
    "ga:itemRevenue": "27.8",
    "ga:uniquePurchases": "4",
    "ga:date": "20180725",
    "ga:revenuePerIte": "6.95",
    "ga:productRefundAmount": "0.0",
    "ga:itemQuantity": "4",
    "ga:itemsPerPurchase": "1.0",
    "ga:buyToDetailRate": "3.8834951456310676"
  },
  {
    "ga:productName": "TEST2",
    "ga:itemRevenue": "75.0",
    "ga:uniquePurchases": "12",
    "ga:date": "20180725",
    "ga:revenuePerIte": "6.25",
    "ga:productRefundAmount": "0.0",
    "ga:itemQuantity": "12",
    "ga:itemsPerPurchase": "1.0",
    "ga:buyToDetailRate": "3.4782608695652173"
  }
]