Convert complex Json to CSV - python

The file is from a slack server export file, so the structure varies every time (if people responded to a thread with text or reactions).
I have tried several SO questions, with similar problems. But I guarantee my question is different. This one, This one too,This one as well
Sample JSON file:
"client_msg_id": "f347abdc-9e2a-4cad-a37d-8daaecc5ad51",
"type": "message",
"text": "I came here just to check <#U3QSFG5A4> This is a sample :slightly_smiling_face:",
"user": "U51N464MN",
"ts": "1550511445.321100",
"team": "T1559JB9V",
"user_team": "T1559JB9V",
"source_team": "T1559JB9V",
"user_profile": {
"avatar_hash": "gcc8ae3d55bb",
"image_72": "https:\/\/secure.gravatar.com\/avatar\/fcc8ae3d55bb91cb750438657694f8a0.jpg?s=72&d=https%3A%2F%2Fa.slack-edge.com%2Fdf10d%2Fimg%2Favatars%2Fava_0026-72.png",
"first_name": "A",
"real_name": "a name",
"display_name": "user",
"team": "T1559JB9V",
"name": "name",
"is_restricted": false,
"is_ultra_restricted": false
},
"thread_ts": "1550511445.321100",
"reply_count": 3,
"reply_users_count": 3,
"latest_reply": "1550515952.338000",
"reply_users": [
"U51N464MN",
"U8DUH4U2V",
"U3QSFG5A4"
],
"replies": [
{
"user": "U51N464MN",
"ts": "1550511485.321200"
},
{
"user": "U8DUH4U2V",
"ts": "1550515191.337300"
},
{
"user": "U3QSFG5A4",
"ts": "1550515952.338000"
}
],
"subscribed": false,
"reactions": [
{
"name": "trolldance",
"users": [
"U51N464MN",
"U4B30MHQE",
"U68E6A0JF"
],
"count": 3
},
{
"name": "trollface",
"users": [
"U8DUH4U2V"
],
"count": 1
}
]
},
The issue is that there are several keys that vary, so the structure changes within the same json file between messages depending on how other users interact to a given message.

with open("file.json") as file:
d = json.load(file)
df = pd.io.json.json_normalize(d)
df.columns = df.columns.map(lambda x: x.split(".")[-1])

Related

Using pandas to convert csv into nested json with dynamic strucutre

I am new to python and now want to convert a csv file into json file. Basically the json file is nested with dynamic structure, the structure will be defined using the csv header.
From csv input:
ID, Name, person_id/id_type, person_id/id_value,person_id_expiry_date,additional_info/0/name,additional_info/0/value,additional_info/1/name,additional_info/1/value,salary_info/details/0/grade,salary_info/details/0/payment,salary_info/details/0/amount,salary_info/details/1/next_promotion
1,Peter,PASSPORT,A452817,1-01-2055,Age,19,Gender,M,Manager,Monthly,8956.23,unknown
2,Jane,PASSPORT,B859804,2-01-2035,Age,38,Gender,F,Worker, Monthly,125980.1,unknown
To json output:
[
{
"ID": 1,
"Name": "Peter",
"person_id": {
"id_type": "PASSPORT",
"id_value": "A452817"
},
"person_id_expiry_date": "1-01-2055",
"additional_info": [
{
"name": "Age",
"value": 19
},
{
"name": "Gender",
"value": "M"
}
],
"salary_info": {
"details": [
{
"grade": "Manager",
"payment": "Monthly",
"amount": 8956.23
},
{
"next_promotion": "unknown"
}
]
}
},
{
"ID": 2,
"Name": "Jane",
"person_id": {
"id_type": "PASSPORT",
"id_value": "B859804"
},
"person_id_expiry_date": "2-01-2035",
"additional_info": [
{
"name": "Age",
"value": 38
},
{
"name": "Gender",
"value": "F"
}
],
"salary_info": {
"details": [
{
"grade": "Worker",
"payment": " Monthly",
"amount": 125980.1
},
{
"next_promotion": "unknown"
}
]
}
}
]
Is this something can be done by the existing pandas API or I have to write lots of complex codes to dynamically construct the json object? Thanks.

Ignore specific JSON keys when extracting data in Python

I'm extracting certain keys in several JSON files and then converting it to a CSV in Python. I'm able to define a key list when I run my code and get the information I need.
However, there are certain sub-keys that I want to ignore from the JSON file. For example, if we look at the following snippet:
JSON Sample
[
{
"callId": "abc123",
"errorCode": 0,
"apiVersion": 2,
"statusCode": 200,
"statusReason": "OK",
"time": "2020-12-14T12:00:32.744Z",
"registeredTimestamp": 1417731582000,
"UID": "_guid_abc123==",
"created": "2014-12-04T22:19:42.894Z",
"createdTimestamp": 1417731582000,
"data": {},
"preferences": {},
"emails": {
"verified": [],
"unverified": []
},
"identities": [
{
"provider": "facebook",
"providerUID": "123",
"allowsLogin": true,
"isLoginIdentity": true,
"isExpiredSession": true,
"lastUpdated": "2014-12-04T22:26:37.002Z",
"lastUpdatedTimestamp": 1417731997002,
"oldestDataUpdated": "2014-12-04T22:26:37.002Z",
"oldestDataUpdatedTimestamp": 1417731997002,
"firstName": "John",
"lastName": "Doe",
"nickname": "John Doe",
"profileURL": "https://www.facebook.com/John.Doe",
"age": 50,
"birthDay": 31,
"birthMonth": 12,
"birthYear": 1969,
"city": "City, State",
"education": [
{
"school": "High School Name",
"schoolType": "High School",
"degree": null,
"startYear": 0,
"fieldOfStudy": null,
"endYear": 0
}
],
"educationLevel": "High School",
"favorites": {
"music": [
{
"name": "Music 1",
"id": "123",
"category": "Musician/band"
},
{
"name": "Music 2",
"id": "123",
"category": "Musician/band"
}
],
"movies": [
{
"name": "Movie 1",
"id": "123",
"category": "Movie"
},
{
"name": "Movie 2",
"id": "123",
"category": "Movie"
}
],
"television": [
{
"name": "TV 1",
"id": "123",
"category": "Tv show"
}
]
},
"followersCount": 0,
"gender": "m",
"hometown": "City, State",
"languages": "English",
"likes": [
{
"name": "Like 1",
"id": "123",
"time": "2014-10-31T23:52:53.0000000Z",
"category": "TV",
"timestamp": "1414799573"
},
{
"name": "Like 2",
"id": "123",
"time": "2014-09-16T08:11:35.0000000Z",
"category": "Music",
"timestamp": "1410855095"
}
],
"locale": "en_US",
"name": "John Doe",
"photoURL": "https://graph.facebook.com/123/picture?type=large",
"timezone": "-8",
"thumbnailURL": "https://graph.facebook.com/123/picture?type=square",
"username": "john.doe",
"verified": "true",
"work": [
{
"companyID": null,
"isCurrent": null,
"endDate": null,
"company": "Company Name",
"industry": null,
"title": "Company Title",
"companySize": null,
"startDate": "2010-12-31T00:00:00"
}
]
}
],
"isActive": true,
"isLockedOut": false,
"isRegistered": true,
"isVerified": false,
"lastLogin": "2014-12-04T22:26:33.002Z",
"lastLoginTimestamp": 1417731993000,
"lastUpdated": "2014-12-04T22:19:42.769Z",
"lastUpdatedTimestamp": 1417731582769,
"loginProvider": "facebook",
"loginIDs": {
"emails": [],
"unverifiedEmails": []
},
"rbaPolicy": {
"riskPolicyLocked": false
},
"oldestDataUpdated": "2014-12-04T22:19:42.894Z",
"oldestDataUpdatedTimestamp": 1417731582894,
"registered": "2014-12-04T22:19:42.956Z",
"regSource": "",
"socialProviders": "facebook"
}
]
I want to extract data from created and identities but ignore identities.favorites and identities.likes as well as their data underneath it.
This is what I have so far, below. I defined the JSON keys that I want to extract in the key_list variable:
Current Code
import json, pandas
from flatten_json import flatten
# Enter the path to the JSON and the filename without appending '.json'
file_path = r'C:\Path\To\file_name'
# Open and load the JSON file
json_list = json.load(open(file_path + '.json', 'r', encoding='utf-8', errors='ignore'))
# Extract data from the defined key names
key_list = ['created', 'identities']
json_list = [{k:d[k] for k in key_list} for d in json_list]
# Flatten and convert to a data frame
json_list_flattened = (flatten(d, '.') for d in json_list)
df = pandas.DataFrame(json_list_flattened)
# Export to CSV in the same directory with the original file name
export_csv = df.to_csv (file_path + r'.csv', sep=',', encoding='utf-8', index=None, header=True)
Similar to the key_list, I suspect that I would make an ignore list and factor that in the json_list for loop that I have? Something like:
key_ignore = ['identities.favorites', 'identities.likes']`
Then utilize the dict.pop() which looks like it will remove the unwanted sub-keys if it matches? Just not sure how to implement that correctly.
Expected Output
As a result, the code should extract data from the defined keys in key_list and ignore the sub keys defined in key_ignore, which is identities.favorites and identities.likes. Then the rest of the code will continue to convert it into a CSV:
created
identities.0.provider
identities.0.providerUID
identities...
2014-12-04T19:23:05.191Z
site
cb8168b0cf734b70ad541f0132763761
...
If the keys are always there, you can use
del d[0]['identities'][0]['likes']
del d[0]['identities'][0]['favorites']
or if you want to remove the columns from the dataframe after reading all the json data in you can use
df.drop(df.filter(regex='identities.0.favorites|identities.0.likes').columns, axis=1, inplace=True)

Multiple Nested Dictionaries with Pandas

Is there a way to import this kind of JSON response into Pandas? Ive been trying to get it usable with json_normalize but I can't seem to get more than one level to work at a time ( I can get notes but can't being in custom_fields). I also cannot figure out how to call out something like ['reporter']['name'] (which should be jdoe). This is from Mantis and its the JSON output of a requests response. Im now wondering if it needs to br broken up into multiple frames and put back together, or should I use a for loop and put the data I want into a better format for PD to import?
In my head each item should be a column in the series all tied to the id column like this.
id | summary | project.name | reporter.name ..|.. custom.fields.Project_Stage | ... notes1.reporter.name | notes1.text ... notes2.reporter.name | notes2.text
{
"issues": [
{
"id": 1234,
"summary": "Some text",
"project": {
"id": 1,
"name": "North America"
},
"category": {
"id": 11,
"name": "Retail"
},
"reporter": {
"id": 1099,
"name": "jdoe"
},
"custom_fields": [
{
"field": {
"id": 107,
"name": "Product Escalations"
},
"value": ""
},
{
"field": {
"id": 1,
"name": "Project_Stage"
},
"value": "Pending"
}
],
"notes": [
{
"id": 214288,
"reporter": {
"id": 9999,
"name": "jdoe"
},
"text": "Worked with Mark over e-mail",
"view_state": {
"id": 10,
"name": "public",
"label": "public"
},
"type": "note",
"created_at": "2020-12-04T15:55:02-08:00",
"updated_at": "2020-12-04T15:55:02-08:00"
},
{
"id": 214289,
"reporter": {
"id": 9999,
"name": "jdoe"
},
"text": "I attempted on numerous occasions to setup a meeting with him to set it up for him.",
"view_state": {
"id": 10,
"name": "public",
"label": "public"
},
"type": "note",
"created_at": "2020-12-04T15:57:02-08:00",
"updated_at": "2020-12-04T15:57:02-08:00"
}
]
}
]
}
Here is what the DF would look like in my head. All the data for one ticket on one line/series.
Those structures with many lists at the same level are tricky. Try flatten_json. https://github.com/amirziai/flatten
If you're response is called 'dic', you can use this.
from flatten_json import flatten
dic_flattened = (flatten(d, '.') for d in dic['issues'])
df = pd.DataFrame(dic_flattened)
Output
id summary project.id project.name category.id category.name ... notes.1.view_state.id notes.1.view_state.name notes.1.view_state.label notes.1.type notes.1.created_at notes.1.updated_at
0 1234 Some text 1 North America 11 Retail ... 10 public public note 2020-12-04T15:57:02-08:00 2020-12-04T15:57:02-08:00
In [101]: df.columns
Out[101]:
Index(['id', 'summary', 'project.id', 'project.name', 'category.id',
'category.name', 'reporter.id', 'reporter.name',
'custom_fields.0.field.id', 'custom_fields.0.field.name',
'custom_fields.0.value', 'custom_fields.1.field.id',
'custom_fields.1.field.name', 'custom_fields.1.value', 'notes.0.id',
'notes.0.reporter.id', 'notes.0.reporter.name', 'notes.0.text',
'notes.0.view_state.id', 'notes.0.view_state.name',
'notes.0.view_state.label', 'notes.0.type', 'notes.0.created_at',
'notes.0.updated_at', 'notes.1.id', 'notes.1.reporter.id',
'notes.1.reporter.name', 'notes.1.text', 'notes.1.view_state.id',
'notes.1.view_state.name', 'notes.1.view_state.label', 'notes.1.type',
'notes.1.created_at', 'notes.1.updated_at'],
dtype='object')

3 levels json count in python

I am new at python, I´ve worked with other languages... I´ve made this code with Java and works, but now, I must do it in python. I have a json of 3 levels, the first two are: resources, usages, and I want to count the names on the third level. I´ve seen several examples but I cant get it done
import json
data = {
"startDate": "2019-06-23T16:07:21.205Z",
"endDate": "2019-07-24T16:07:21.205Z",
"status": "Complete",
"usages": [
{
"name": "PureCloud Edge Virtual Usage",
"resources": [
{
"name": "Edge01-VM-GNS-DemoSite01 (1f279086-a6be-4a21-ab7a-2bb1ae703fa0)",
"date": "2019-07-24T09:00:28.034Z"
},
{
"name": "329ad5ae-e3a3-4371-9684-13dcb6542e11",
"date": "2019-07-24T09:00:28.034Z"
},
{
"name": "e5796741-bd63-4b8e-9837-4afb95bb0c09",
"date": "2019-07-24T09:00:28.034Z"
}
]
},
{
"name": "PureCloud for SmartVideo Add-On Concurrent",
"resources": [
{
"name": "jpizarro#gns.com.co",
"date": "2019-06-25T04:54:17.662Z"
},
{
"name": "jaguilera#gns.com.co",
"date": "2019-06-25T04:54:17.662Z"
},
{
"name": "dcortes#gns.com.co",
"date": "2019-07-15T15:06:09.203Z"
}
]
},
{
"name": "PureCloud 3 Concurrent User Usage",
"resources": [
{
"name": "jpizarro#gns.com.co",
"date": "2019-06-25T04:54:17.662Z"
},
{
"name": "jaguilera#gns.com.co",
"date": "2019-06-25T04:54:17.662Z"
},
{
"name": "dcortes#gns.com.co",
"date": "2019-07-15T15:06:09.203Z"
}
]
},
{
"name": "PureCloud Skype for Business WebSDK",
"resources": [
{
"name": "jpizarro#gns.com.co",
"date": "2019-06-25T04:54:17.662Z"
},
{
"name": "jaguilera#gns.com.co",
"date": "2019-06-25T04:54:17.662Z"
},
{
"name": "dcortes#gns.com.co",
"date": "2019-07-15T15:06:09.203Z"
}
]
}
],
"selfUri": "/api/v2/billing/reports/billableusage"
}
cantidadDeLicencias = 0
cantidadDeUsages = len(data['usages'])
for x in range(cantidadDeUsages):
temporal = data[x]
cantidadDeResources = len(temporal['resource'])
for z in range(cantidadDeResources):
print(x)
What changes I have to make? Maybe I have to do it on another approach? Thanks in advance
Update
Code that works
cantidadDeLicencias = 0
for usage in data['usages']:
cantidadDeLicencias = cantidadDeLicencias + len(usage['resources'])
print(cantidadDeLicencias)
You can do this :
for usage in data['usages']:
print(len(usage['resources']))
If you want to know the number of names in each of the resources level, counting the duplicated names (e.g. "jaguilera#gns.com.co" appears more than one time in your data), then just do iterate over the first-level (usages) and sum the size of each array
cantidadDeLicencias = 0
for usage in data['usages']:
cantidadDeLicencias += len(usage['resources'])
print(cantidadDeLicencias)
If you don't want to count duplicates, then use a set and iterate over each resources array
cantidadDeLicencias_set = {}
for usage in data['usages']:
for resource in usage['resources']:
cantidadDeLicencias_set.add(resource['name'])
print(len(cantidadDeLicencias_set ))

Flatten nested json to csv with nested column names

I have rather very weird requirement now. I have below json and somehow I have to convert it into flat csv.
[
{
"authorizationQualifier": "SDA",
"authorizationInformation": " ",
"securityQualifier": "ASD",
"securityInformation": " ",
"senderQualifier": "ASDAD",
"senderId": "FADA ",
"receiverQualifier": "ADSAS",
"receiverId": "ADAD ",
"date": "140101",
"time": "0730",
"standardsId": null,
"version": "00501",
"interchangeControlNumber": "123456789",
"acknowledgmentRequested": "0",
"testIndicator": "T",
"functionalGroups": [
{
"functionalIdentifierCode": "ADSAD",
"applicationSenderCode": "ASDAD",
"applicationReceiverCode": "ADSADS",
"date": "20140101",
"time": "07294900",
"groupControlNumber": "123456789",
"responsibleAgencyCode": "X",
"version": "005010X221A1",
"transactions": [
{
"name": "ASDADAD",
"transactionSetIdentifierCode": "adADS",
"transactionSetControlNumber": "123456789",
"implementationConventionReference": null,
"segments": [
{
"BPR03": "ad",
"BPR14": "QWQWDQ",
"BPR02": "1.57",
"BPR13": "23223",
"BPR01": "sad",
"BPR12": "56",
"BPR10": "32424",
"BPR09": "12313",
"BPR08": "DA",
"BPR07": "123456789",
"BPR06": "12313",
"BPR05": "ASDADSAD",
"BPR16": "21313",
"BPR04": "SDADSAS",
"BPR15": "11212",
"id": "aDSASD"
},
{
"TRN02": "2424",
"TRN03": "35435345",
"TRN01": "3435345",
"id": "FSDF"
},
{
"REF02": "fdsffs",
"REF01": "sfsfs",
"id": "fsfdsfd"
},
{
"DTM02": "2432424",
"id": "sfsfd",
"DTM01": "234243"
}
],
"loops": [
{
"id": "24324234234",
"segments": [
{
"N101": "sfsfsdf",
"N102": "sfsf",
"id": "dgfdgf"
},
{
"N301": "sfdssfdsfsf",
"N302": "effdssf",
"id": "fdssf"
},
{
"N401": "sdffssf",
"id": "sfds",
"N402": "sfdsf",
"N403": "23424"
},
{
"PER06": "Wsfsfdsfsf",
"PER05": "sfsf",
"PER04": "23424",
"PER03": "fdfbvcb",
"PER02": "Pedsdsf",
"PER01": "sfsfsf",
"id": "fdsdf"
}
]
},
{
"id": "2342",
"segments": [
{
"N101": "sdfsfds",
"N102": "vcbvcb",
"N103": "dsfsdfs",
"N104": "343443",
"id": "fdgfdg"
},
{
"N401": "dfsgdfg",
"id": "dfgdgdf",
"N402": "dgdgdg",
"N403": "234244"
},
{
"REF02": "23423342",
"REF01": "fsdfs",
"id": "sfdsfds"
}
]
}
]
}
]
}
]
}
]
The column header name corresponding to deeper key-value make take nested form, like functionalGroups[0].transactions[0].segments[0].BPR15.
I am able to do this in java using this github project (here you can find the output format I desire in the explanation) in one line:
flatJson = JSONFlattener.parseJson(new File("files/simple.json"), "UTF-8");
The output was:
date,securityQualifier,testIndicator,functionalGroups[1].functionalIdentifierCode,functionalGroups[1].date,functionalGroups[1].applicationReceiverCode, ...
140101,00,T,HP,20140101,ETIN,...
But I want to do this in python. I tried as suggested in this answer:
with open('data.json') as data_file:
data = json.load(data_file)
df = json_normalize(data, record_prefix=True)
with open('temp2.csv', "w", newline='\n') as csv_file:
csv_file.write(df.to_csv())
However, for column functionalGroups, it dumps json as a cell value.
I also tried as suggested in this answer:
with open('data.json') as f: # this ensures opening and closing file
a = json.loads(f.read())
df = pandas.DataFrame(a)
print(df.transpose())
But this also seem to do the same:
0
acknowledgmentRequested 0
authorizationInformation
authorizationQualifier SDA
date 140101
functionalGroups [{'functionalIdentifierCode': 'ADSAD', 'applic...
interchangeControlNumber 123456789
receiverId ADAD
receiverQualifier ADSAS
securityInformation
securityQualifier ASD
senderId FADA
senderQualifier ASDAD
standardsId None
testIndicator T
time 0730
version 00501
Is it possible to do what I desire in python?

Categories