Develop Reference

Best way to parse a JSON to store in SQL database (SQL stored procedure/Python)

I have a table of overly complex JSON files I'm trying to convert to tabular format to store in a SQL database. I'm pulling the JSONs from the quickbooks online API, and the format is messy to say the least.. (We're talking 7x nested JSONs for some bits of it..
The format resembles the code snippet down below. Currently I am using a bunch of OpenJSON's + Cross applys to dig down to the innermost ColData then work my way up but it looks like some of the ColData's get skipped over doing that.
Are there any better ways, using either Python (since I pull the JSON initially in Python before sending the JSON to a SQL database to parse) or SQL to convert it to tabular format besides manually trying to use OpenJSON with Cross applys?
The goal is to get all of the ColData's into a SQL table...
Thanks!
{
"Header": {
"ReportName": "BalanceSheet",
"Option": [
{
"Name": "AccountingStandard",
"Value": "GAAP"
},
{
"Name": "NoReportData",
"Value": "false"
}
],
"DateMacro": "this calendar year-to-date",
"ReportBasis": "Accrual",
"StartPeriod": "2016-01-01",
"Currency": "USD",
"EndPeriod": "2016-10-31",
"Time": "2016-10-31T09:42:21-07:00",
"SummarizeColumnsBy": "Total"
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"value": "ASSETS"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"value": "Current Assets"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"value": "Bank Accounts"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "35",
"value": "Checking"
},
{
"value": "1350.55"
}
],
"type": "Data"
},
{
"ColData": [
{
"id": "36",
"value": "Savings"
},
{
"value": "800.00"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "BankAccounts",
"Summary": {
"ColData": [
{
"value": "Total Bank Accounts"
},
{
"value": "2150.55"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Accounts Receivable"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "84",
"value": "Accounts Receivable (A/R)"
},
{
"value": "6383.12"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "AR",
"Summary": {
"ColData": [
{
"value": "Total Accounts Receivable"
},
{
"value": "6383.12"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Other current assets"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "81",
"value": "Inventory Asset"
},
{
"value": "596.25"
}
],
"type": "Data"
},
{
"ColData": [
{
"id": "4",
"value": "Undeposited Funds"
},
{
"value": "2117.52"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "OtherCurrentAssets",
"Summary": {
"ColData": [
{
"value": "Total Other current assets"
},
{
"value": "2713.77"
}
]
}
}
]
},
"type": "Section",
"group": "CurrentAssets",
"Summary": {
"ColData": [
{
"value": "Total Current Assets"
},
{
"value": "11247.44"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Fixed Assets"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"id": "37",
"value": "Truck"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "38",
"value": "Original Cost"
},
{
"value": "13495.00"
}
],
"type": "Data"
}
]
},
"type": "Section",
"Summary": {
"ColData": [
{
"value": "Total Truck"
},
{
"value": "13495.00"
}
]
}
}
]
},
"type": "Section",
"group": "FixedAssets",
"Summary": {
"ColData": [
{
"value": "Total Fixed Assets"
},
{
"value": "13495.00"
}
]
}
}
]
},
"type": "Section",
"group": "TotalAssets",
"Summary": {
"ColData": [
{
"value": "TOTAL ASSETS"
},
{
"value": "24742.44"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "LIABILITIES AND EQUITY"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"value": "Liabilities"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"value": "Current Liabilities"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"Header": {
"ColData": [
{
"value": "Accounts Payable"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "33",
"value": "Accounts Payable (A/P)"
},
{
"value": "1984.17"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "AP",
"Summary": {
"ColData": [
{
"value": "Total Accounts Payable"
},
{
"value": "1984.17"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Credit Cards"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "41",
"value": "Mastercard"
},
{
"value": "157.72"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "CreditCards",
"Summary": {
"ColData": [
{
"value": "Total Credit Cards"
},
{
"value": "157.72"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Other Current Liabilities"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "89",
"value": "Arizona Dept. of Revenue Payable"
},
{
"value": "4.55"
}
],
"type": "Data"
},
{
"ColData": [
{
"id": "90",
"value": "Board of Equalization Payable"
},
{
"value": "401.98"
}
],
"type": "Data"
},
{
"ColData": [
{
"id": "43",
"value": "Loan Payable"
},
{
"value": "4000.00"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "OtherCurrentLiabilities",
"Summary": {
"ColData": [
{
"value": "Total Other Current Liabilities"
},
{
"value": "4406.53"
}
]
}
}
]
},
"type": "Section",
"group": "CurrentLiabilities",
"Summary": {
"ColData": [
{
"value": "Total Current Liabilities"
},
{
"value": "6548.42"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Long-Term Liabilities"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "44",
"value": "Notes Payable"
},
{
"value": "25000.00"
}
],
"type": "Data"
}
]
},
"type": "Section",
"group": "LongTermLiabilities",
"Summary": {
"ColData": [
{
"value": "Total Long-Term Liabilities"
},
{
"value": "25000.00"
}
]
}
}
]
},
"type": "Section",
"group": "Liabilities",
"Summary": {
"ColData": [
{
"value": "Total Liabilities"
},
{
"value": "31548.42"
}
]
}
},
{
"Header": {
"ColData": [
{
"value": "Equity"
},
{
"value": ""
}
]
},
"Rows": {
"Row": [
{
"ColData": [
{
"id": "34",
"value": "Opening Balance Equity"
},
{
"value": "-9337.50"
}
],
"type": "Data"
},
{
"ColData": [
{
"id": "2",
"value": "Retained Earnings"
},
{
"value": "91.25"
}
],
"type": "Data"
},
{
"ColData": [
{
"value": "Net Income"
},
{
"value": "2440.27"
}
],
"type": "Data",
"group": "NetIncome"
}
]
},
"type": "Section",
"group": "Equity",
"Summary": {
"ColData": [
{
"value": "Total Equity"
},
{
"value": "-6805.98"
}
]
}
}
]
},
"type": "Section",
"group": "TotalLiabilitiesAndEquity",
"Summary": {
"ColData": [
{
"value": "TOTAL LIABILITIES AND EQUITY"
},
{
"value": "24742.44"
}
]
}
}
]
},
"Columns": {
"Column": [
{
"ColType": "Account",
"ColTitle": "",
"MetaData": [
{
"Name": "ColKey",
"Value": "account"
}
]
},
{
"ColType": "Money",
"ColTitle": "Total",
"MetaData": [
{
"Name": "ColKey",
"Value": "total"
}
]
}
]
}
}
Here is what I tried to get the ColData (unsuccessfully I might add), I think it might be a little too contrived to do in SQL but I'm not sure if I should continue trying this way or if there's a better way in Python:
declare #json nvarchar(max)
SELECT #json = json FROM QboApiRawJSONData WHERE ID = 2
--Outer layer of JSON breaks into 3 parts - header, columns, rows
SELECT * FROM OPENJSON(#json)
WITH
(
Rows nvarchar(max) AS JSON
) as MainLayer
CROSS APPLY OPENJSON (MainLayer.Rows)
WITH
(
Row nvarchar(max) AS JSON
) as SecondaryLayer
CROSS APPLY OPENJSON (SecondaryLayer.Row)
WITH
(
Rows nvarchar(max) AS JSON
) As ThirdLayer
CROSS APPLY OPENJSON (ThirdLayer.Rows)
WITH
(
Row nvarchar(max) AS JSON
) as FourthLayer
CROSS APPLY OPENJSON (FourthLayer.Row)
WITH
(
Rows nvarchar(max) AS JSON
) as FifthLayer
CROSS APPLY OPENJSON (FifthLayer.Rows)
WITH
(
Row nvarchar(max) AS JSON
) as SixthLayer
CROSS APPLY OPENJSON (SixthLayer.Row)
WITH
(
Rows nvarchar(max) AS JSON
) as SeventhLayer
CROSS APPLY OPENJSON (SeventhLayer.Rows)
WITH
(
Row nvarchar(max) AS JSON
) as EighthLayer
CROSS APPLY OPENJSON (EighthLayer.Row)
WITH
(
Rows nvarchar(max) AS JSON
) as LayerNine
---Things get funky here
CROSS APPLY OPENJSON (LayerNine.Rows)
WITH
(
Row nvarchar(max) AS JSON
) as LayerTen
CROSS APPLY OPENJSON (LayerTen.Row)
WITH
(
Rows nvarchar(max) AS JSON
) as LayerEleven
CROSS APPLY OPENJSON (LayerEleven.Rows)
WITH
(
Row nvarchar(max) AS JSON
) as LayerTwelve
--21 items in last col

There is JSON support for sql-server:
https://learn.microsoft.com/en-us/sql/relational-databases/json/json-data-sql-server?view=sql-server-ver15

There is a JSON storage method here: https://learn.microsoft.com/en-us/sql/relational-databases/json/store-json-documents-in-sql-tables?view=sql-server-ver15. Although it will be fairly complex here I recommend storing the JSON data as a Logs table and then following the tutorial above to see if that solves your issue.

Use Quickbooks API to get the JSON, then refer to this guide:
https://learn.microsoft.com/en-us/sql/relational-databases/json/json-data-sql-server?view=sql-server-ver15
and this guide:
https://learn.microsoft.com/en-us/sql/relational-databases/json/convert-json-data-to-rows-and-columns-with-openjson-sql-server?view=sql-server-ver15
You can also consider setting something like AWS Lambda or Google Cloud Functions if you need something more automated.

Convert csv data to nested json data

I have a csv data as:
dataf = pd.DataFrame{'title':['tit1', 'tit1', 'tit2' 'tit2','tit3'],
'context':['con1', 'con1', 'con2', 'con2', 'con3'],
'answers':['text1', 'text2', 'text3', 'text4', 'text5'],
'question':['que1', 'que2', 'que3', 'que4', 'que5'],
'id':['DDA', 'SAV', 'AFS', 'ML', 'MLI']}
I wanted to convert it to nested json format as below using python:
[
{
"title": "tit1",
"paragraph": [
{
"context": "con1",
"qas": [
{
"answers": "text1",
"question": "que1",
"id": "DDA"
},
{
"answers": "text2",
"question": "que2",
"id": "SAV"
}
]
}
]
},
{
"title": "tit2",
"paragraph": [
{
"context": "con2",
"qas": [
{
"answers": "text3",
"question": "que3",
"id": "AFS"
},
{
"answers": "text4",
"question": "que4",
"id": "ML"
}
],
"context": "con3",
"qas": [
{
"answers": "text5",
"question": "que5",
"id": "MLI"
}
]
}
]
}
]

Identify location of item in json file

Suppose I have the following json file. With data1["tenants"][1]['name'] I can select uniquename2. Is there a way to collect the '1' number by looping over the document?
{
"tenants": [{
"key": "identifier",
"name": "uniquename",
"image": "url",
"match": [
"identifier"
],
"tags": [
"tag1",
"tag2"
]
},
{
"key": "identifier",
"name": "uniquename2",
"image": "url",
"match": [
"identifier1",
"identifier2"
],
"tags": ["tag"]
}
]
}
in short: data1["tenants"][1]['name']= uniquename2 data1["tenants"][0]['name'] = uniquename
How can I find out which number has which name. So if I have uniquename2 what number/index corresponds with it?

you can iterate over the tenants to map the index to the name
data = {
"tenants": [{
"key": "identifier",
"name": "uniquename",
"image": "url",
"match": [
"identifier"
],
"tags": [
"tag1",
"tag2"
]
},
{
"key": "identifier",
"name": "uniquename2",
"image": "url",
"match": [
"identifier1",
"identifier2"
],
"tags": ["tag"]
}
]
}
for index, tenant in enumerate(data['tenants']):
print(index, tenant['name'])
OUTPUT
0 uniquename
1 uniquename2

Assuming, you have turned your json into a dictionary already, this is how you can get the index of the firs ocurrence of a name in your list (This relies on the names actually being unique):
data = {
"tenants": [{
"key": "identifier",
"name": "uniquename",
"image": "url",
"match": [
"identifier"
],
"tags": [
"tag1",
"tag2"
]
},
{
"key": "identifier",
"name": "uniquename2",
"image": "url",
"match": [
"identifier1",
"identifier2"
],
"tags": ["tag"]
}
]
}
def index_of(tenants, tenant_name):
try:
return tenants.index(
next(
tenant for tenant in tenants
if tenant["name"] == tenant_name
)
)
except StopIteration:
raise ValueError(
f"tenants list does not have tenant by name {tenant_name}."
)
index_of(data["tenants"], "uniquename") # 0

Fast way of adding fields to a nested dict

I need a help with improving my code.
I've got a nested dict with many levels:
{
"11": {
"FacLC": {
"immty": [
"in_mm",
"in_mm"
],
"moood": [
"in_oo",
"in_oo"
]
}
},
"22": {
"FacLC": {
"immty": [
"in_mm",
"in_mm",
"in_mm"
]
}
}
}
And I want to add additional fields on every level, so my output looks like this:
[
{
"id": "",
"name": "11",
"general": [
{
"id": "",
"name": "FacLC",
"specifics": [
{
"id": "",
"name": "immty",
"characteristics": [
{
"id": "",
"name": "in_mm"
},
{
"id": "",
"name": "in_mm"
}
]
},
{
"id": "",
"name": "moood",
"characteristics": [
{
"id": "",
"name": "in_oo"
},
{
"id": "",
"name": "in_oo"
}
]
}
]
}
]
},
{
"id": "",
"name": "22",
"general": [
{
"id": "",
"name": "FacLC",
"specifics": [
{
"id": "",
"name": "immty",
"characteristics": [
{
"id": "",
"name": "in_mm"
},
{
"id": "",
"name": "in_mm"
},
{
"id": "",
"name": "in_mm"
}
]
}
]
}
]
}
]
I managed to write a 4-times nested for loop, what I find inefficient and inelegant:
for main_name, general in my_dict.items():
generals = []
for general_name, specific in general.items():
specifics = []
for specific_name, characteristics in specific.items():
characteristics_dicts = []
for characteristic in characteristics:
characteristics_dicts.append({
"id": "",
"name": characteristic,
})
specifics.append({
"id": "",
"name": specific_name,
"characteristics": characteristics_dicts,
})
generals.append({
"id": "",
"name": general_name,
"specifics": specifics,
})
my_new_dict.append({
"id": "",
"name": main_name,
"general": generals,
})
I am wondering if there is more compact and efficient solution.

In the past I created a function to do it. Basically you call this function everytime that you need to add new fields to a nested dict, independently on how many levels this nested dict have. You only have to inform the 'full path' , that I called the 'key_map'.
Like ['node1','node1a','node1apart3']
def insert_value_using_map(_nodes_list_to_be_appended, _keys_map, _value_to_be_inserted):
for _key in _keys_map[:-1]:
_nodes_list_to_be_appended = _nodes_list_to_be_appended.setdefault(_key, {})
_nodes_list_to_be_appended[_keys_map[-1]] = _value_to_be_inserted

Convert deeply nested json from facebook to dataframe in python

I am trying to get user details of persons who has put likes, comments on Facebook posts. I am using python facebook-sdk package. Code is as follows.
import facebook as fi
import json
graph = fi.GraphAPI('Access Token')
data = json.dumps(graph.get_object('DSIfootcandy/posts'))
From the above, I am getting a highly nested json. Here I will put only a json string for one post in the fb.
{
"paging": {
"next": "https://graph.facebook.com/v2.0/425073257683630/posts?access_token=&limit=25&until=1449201121&__paging_token=enc_AdD0DL6sN3aDZCwfYY25rJLW9IZBZCLM1QfX0venal6rpjUNvAWZBOoxTjbOYZAaFiBImzMqiv149HPH5FBJFo0nSVOPqUy78S0YvwZDZD",
"previous": "https://graph.facebook.com/v2.0/425073257683630/posts?since=1450843741&access_token=&limit=25&__paging_token=enc_AdCYobFJpcNavx6STzfPFyFe6eQQxRhkObwl2EdulwL7mjbnIETve7sJZCPMwVm7lu7yZA5FoY5Q4sprlQezF4AlGfZCWALClAZDZD&__previous=1"
},
"data": [
{
"picture": "https://fbcdn-photos-e-a.akamaihd.net/hphotos-ak-xfa1/v/t1.0-0/p130x130/1285_5066979392443_n.png?oh=b37a42ee58654f08af5abbd4f52b1ace&oe=570898E7&__gda__=1461440649_aa94b9ec60f22004675c4a527e8893f",
"is_hidden": false,
"likes": {
"paging": {
"cursors": {
"after": "MTU3NzQxODMzNTg0NDcwNQ==",
"before": "MTU5Mzc1MjA3NDE4ODgwMA=="
}
},
"data": [
{
"id": "1593752074188800",
"name": "Maduri Priyadarshani"
},
{
"id": "427605680763414",
"name": "Darshi Mashika"
},
{
"id": "599793563453832",
"name": "Shakeer Nimeshani Shashikala"
},
{
"id": "1577418335844705",
"name": "Däzlling Jalali Muishu"
}
]
},
"from": {
"category": "Retail and Consumer Merchandise",
"name": "Footcandy",
"category_list": [
{
"id": "2239",
"name": "Retail and Consumer Merchandise"
}
],
"id": "425073257683630"
},
"name": "Timeline Photos",
"privacy": {
"allow": "",
"deny": "",
"friends": "",
"description": "",
"value": ""
},
"is_expired": false,
"comments": {
"paging": {
"cursors": {
"after": "WTI5dGJXVnVkRjlqZFhKemIzSUVXdNVFExTURRd09qRTBOVEE0TkRRNE5EVT0=",
"before": "WTI5dGJXVnVkRjlqZFhKemIzNE16Y3dNVFExTVRFNE9qRTBOVEE0TkRRME5UVT0="
}
},
"data": [
{
"from": {
"name": "NiFû Shafrà",
"id": "1025030640553"
},
"like_count": 0,
"can_remove": false,
"created_time": "2015-12-23T04:20:55+0000",
"message": "wow lovely one",
"id": "50018692683829_500458145118",
"user_likes": false
},
{
"from": {
"name": "Shamnaz Lukmanjee",
"id": "160625809961884"
},
"like_count": 0,
"can_remove": false,
"created_time": "2015-12-23T04:27:25+0000",
"message": "Nice",
"id": "500186926838929_500450145040",
"user_likes": false
}
]
},
"actions": [
{
"link": "https://www.facebook.com/425073257683630/posts/5001866838929",
"name": "Comment"
},
{
"link": "https://www.facebook.com/42507683630/posts/500186926838929",
"name": "Like"
}
],
"updated_time": "2015-12-23T04:27:25+0000",
"link": "https://www.facebook.com/DSIFootcandy/photos/a.438926536298302.1073741827.4250732576630/50086926838929/?type=3",
"object_id": "50018692838929",
"shares": {
"count": 3
},
"created_time": "2015-12-23T04:09:01+0000",
"message": "Reach new heights in the cute and extremely comfortable \"Silviar\" www.focandy.lk",
"type": "photo",
"id": "425077683630_50018926838929",
"status_type": "added_photos",
"icon": "https://www.facebook.com/images/icons/photo1.gif"
}
]
}
Now I need to get this data into a dataframe as follows(no need to get all).
item | Like_id |Like_username | comments_userid |comments_username|comment(msg)|
-----+---------+--------------+-----------------+-----------------+------------+
Bag | 45546 | noel | 641 | James | nice work |
-----+---------+--------------+-----------------+-----------------+------------+
Any Help will be Highly Appreciated.

Not exactly like your intended format, but here is the making of a solution :
import pandas
DictionaryObject_as_List = str(mydict).replace("{","").replace("}","").replace("[","").replace("]","").split(",")
newlist = []
for row in DictionaryObject_as_List :
row = row.replace('https://',' ').split(":")
exec('newlist.append ( ' + "[" + " , ".join(row)+"]" + ')')
DataFrame_Object = pandas.DataFrame(newlist)
print DataFrame_Object