Record leaf nodes support

[vigbk]

Python version - 2.7.13
Google-cloud-bigquery - 0.22.1

I am trying to insert data using the big query python client library. I am getting an error that record leaf types are not supported. The schema contains nested records that are repeatable. Is there a way using the client to insert nested repeatable records ?

my schema is as follows:

[daspecster]

Thanks for reporting @vigbk!

Do you have some example code/stacktrace?

@tseaver is this kind of related to #2354?

[vigbk]

@daspecster Here is an excerpt of the code I used

from google.cloud import bigquery
import os
import json
from datetime import datetime


def transform(item):
    newItem = {"insertId": getUUID(), "json": item}
    return newItem

def getTemplateSuffix():

    utcDate =  datetime.utcnow().strftime("%Y%m%d")
    templateSuffix = "_" + utcDate
    return templateSuffix

def uploadToBigQuery(data):

     rowData = json.loads(data)



     os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "credentials.json"

     # Instantiates a client
     bigquery_client = bigquery.Client("logs")

     dataset = bigquery_client.dataset("eventlogs")
     table = dataset.table("events")

     # Reload the table to get the schema.
     table.reload()

     rows = [rowData]
     print(rows)
     errors = table.insert_data(rows=rows, template_suffix=getTemplateSuffix())

     if not errors:
         print('Loaded 1 row into {}:{}'.format(dataset, table))
     else:
         print('Errors:')
         print(errors)


data = '{"user_dim":{"user_id":"12345","user_properties":[{"key":"subdomain_id","value":{"string_value":"custom.zendesk.com"}}]},"event_dim":{"date":"20170116","category":"ALERT","name":"NUMBER_NOT_A_MOBILE","timestamp_micros":1484580530384000,"params":[{"key":"network","value":{"string_value":"Singapore"}},{"key":"country","value":{"string_value":"Singapore"}}]}}'

uploadToBigQuery(data) 

[lukesneeringer]

Is this a duplicate of #2951? (It seems like it is, but I am hesitant to close without being sure.)

[dhermes]

@tswast Can you confirm / deny the duplicateness?

[tswast]

These two issues look like duplicates to me, too.

[lukesneeringer]

Okay. I marked the other one as important, and am going to close this one for personal sanity / tracking. :-)